THE AUDIO DICTIONARY

THE
AUDIO
D I C T I O N A RY

Glenn D. White
and
Gary J. Louie

Third Edition, Revised and Expanded

University of Washington Press

Seattle and London
Third edition, revised and expanded 2005 by the University of Washington Press
Printed in the United States of America
12 11 10 09 08 07 06 05 5 4 3 2 1

All rights reserved. No part of this publication may be reproduced or transmitted in

any form or by any means, electronic or mechanical, including photocopy, recording,
or any information storage or retrieval system, without permission in writing from
the publisher.

University of Washington Press

PO Box 50096
Seattle, WA 98145-5096
www.washington.edu/uwpress

Library of Congress Cataloging-in-Publication Data

White, Glenn D., 1933

The audio dictionary / by Glenn D. White and Gary J. Louie.3rd ed., rev. and
expanded.
p. cm.
Includes bibliographical references.
ISBN 0-295-98498-8 (pbk. : alk. paper)
1. SoundRecording and reproducingDictionaries. I. Louie, Gary J.
II. Title.
TK7881.4.W48 2004
621.389'3dc22 2004029425

The paper used in this publication meets the minimum requirements of American
National Standard for Information SciencesPermanence of Paper for Printed
Library Materials, ANSI Z39.48-1984.

This edition contains several entries from Larry Blakes Glossary of Film Sound
terms. The complete version of which is only available on his Swelltonelabs website
<www.swelltonelabs.com>.
 To Naomi Pascal,
who rst suggested that I write this book,
and who, with characteristic enthusiasm and good humor,
inspired me to nish it.
G.W.
Contents

Preface to the Third Edition ix

Introduction xi

The Audio Dictionary 1

Appendix 1. The Art and Science of Good Acoustics 429

Appendix 2. Some Frequently Used Symbols and Units 438
Appendix 3. How to Subdue a Hi-Fi Salesperson 441
Appendix 4. Good Acoustics in Small Rooms
and Auditoria 445
Appendix 5. Some Notes on Audio Measurements 464
Appendix 6. The Balanced Line 477
Appendix 7. Some Notes on Digital Audio 480
Appendix 8. Musical Scales and the Tuning of Musical
Instruments 482
Appendix 9. Some Notes on the History of High Fidelity 487
Appendix 10. Some Notes on Decimal and Binary
Arithmetic as Used in Computer
Terminology 491
Appendix 11. Some Notes on Impedance and Frequency
Response 493
Bibliography 497
Preface to the Third Edition

Although the second edition of The Audio Dictionary is enjoying healthy sales,
friends, acquaintances, and strangers have asked me now and then when a
third edition would be published. When my good friend Naomi Pascal, now
Editor-at-Large at the University of Washington Press, who inveigled me to
write the book in the rst place, asked me the same question, I decided it
was time. I realized at once that putting a new edition together would be a
daunting task to tackle because of the new audio-related technologies such
as the explosion of the home theater systems market, and the upgrades in
motion picture sound. I asked Gary Louie, who gave me many suggestions
and lots of information when I was preparing the rst two editions, if he
would like to cowrite this new edition. Gary, a longtime friend and crack
audio/video electronics technician, experienced recording engineer, and avid
collector of audio memorabilia and arcana, agreed, and the present volume
is the result.
Gary and I were amazed at the magnitude of changes the audio industry
has undergone in the last ten years. This third edition contains over 400 new
entries, the second edition entries and appendices have been thoroughly gone
over and in some cases enlarged and brought up to date, and two new appen-
dixes have been added.
Glenn White
2005

ix
Introduction

Lexicographer, n. A pestilent fellow who, under the pretense of recording

some particular stage in the development of a language, does what he can to
arrest its growth, stiffen its exibility, and mechanize its methods.From
Ambrose Bierce, The Devils Dictionary

Throughout the years of teaching acoustics and audio-related courses at the

university level, we have felt the need for a book covering the terminology
and basic concepts in the elds of sound recording, sound reinforcement, and
musical acoustics. The marginal emphasis on audio education by our schools
and universities has permitted a large body of folklore to exist. This book is
aimed at dispelling some misinformation and is intended to supplement text-
books for courses in these areas and also to serve as a source of information
for the general reader who has an interest in reproduced music and/or musi-
cal acoustics. The dictionary format was chosen primarily for ease of cross-
referencing as well as ease of locating specic information. Denitions of
fundamental principles and standards are included, as well as entries on more
abstract and subjective concepts. The areas of basic acoustics, elementary elec-
troacoustics, digital audio, electronics as applied to audio, and some topics
within the areas of psychoacoustics are covered.
A diligent attempt was made to keep the text free of mathematics and to
present the material in a framework of familiar and comfortable territory to
the nonspecialist. At the same time, oversimplication was avoided as much
as possible. It might be argued that some concepts, such as decibels, cannot
be adequately described without mathematics, and this is no doubt true; how-
ever, the underlying concept of logarithmic ratios can be explained without
resorting to mathematical symbols. This approach to the subject via intuitive
principles is used consistently.
There is a certain amount of repetition of some concepts as they are dis-
cussed in various contexts. In these cases, we have tried to make the approaches
different enough to increase the understanding by the nonspecialist. There is
included in the denitions a signicant amount of information in anecdotal
form, but we have tried to keep personal opinions out of the dictionary proper.
In the back of the book are eleven appendixes that discuss certain topics in
more detail, and we confess that here a certain amount of personal bias may
have crept in.
It is expected that this treatment will allow the book to appeal to a broad

xi
Introduction

audience, from the nontechnical stereo set owner to the semiprofessional who
needs a ready source of information. We hope that the former may be stim-
ulated to dig deeper into the subject, and the latter may nd that the intuitive
approach will increase understanding. The appearance of a capitalized word
within an entry signies another entry with that heading. Thus, one may be
led in a zigzag path through the dictionary by looking up unfamiliar words
as they are encountered.

xii
THE AUDIO DICTIONARY
 A
AAC, Advanced Audio Coding A high-quality perceptual audio coding
technology appropriate for many broadcast and electronic music distribu-
tion applications. Coding efciency is superior to that of MP3, providing
higher-quality audio at lower bit rates. Developed and standardized as an
ISO/IEC specication by four industry leaders (AT&T, Dolby Laboratories,
The Fraunhofer Institute for Integrated Circuits, and Sony Corporation),
AAC is supported by a growing number of hardware and software man-
ufacturers as the logical successor to MP3. It was declared an international
standard by MPEG in April 1997. It is used by Apple Computer for their
music downloading service begun on April 28, 2003.
A Battery The low-voltage battery that supplied current for the tube
heaters of early radios. One of the earliest types of A battery was the No. 6
dry cell, which gave 1.5 volts. Later, 6-V lead-acid storage batteries were
used, which of course were rechargeable. Standard automobile batteries of
the day would work, but special radio A batteries were said to provide
smoother current! The rst power supplies to provide heater current from
the ac line were called A eliminators and had to have very effective lter-
ing to avoid 60-hertz hum.
ABS, ABS-Time, A-Time, Absolute Time Most digital recorders provide
a means for recording a time count along with the audio. Recorders such
as DAT and ADAT will start the time count at 00:00:00 at the beginning of
the tape, and advance in real time as the recording is made. CD-Rs have a
similar method of timing using a pre-made signal groove, known as ATIP,
for Absolute Time In Pregroove. The rst generation of DAT recorders did
not have A-Time implemented, thus a tape recorded on such a machine will
not show A-Time. A-Time is not the same as SMPTE time code, although
some machines can work with A-Time in an SMPTE environment.
Absolute Phase A sound system that preserves the actual polarity of the
original sound waves being recorded. In other words, care is taken that there
is no polarity reversal anywhere in the recording/reproducing chain. Thus,
a compression wave at the recording session is reproduced as a compres-
sion wave at the listeners location, rather than a rarefaction. There are those
who claim there is no audible difference if the polarity of both channels is
reversed and others who claim there is a vast difference. In any case, the
term should be absolute polarity.
Absolute Pitch Some people, not necessarily musicians, on hearing a
sound have the ability to accurately identify its position or note on the musi-

3
Absolute Polarity

cal scale. Absolute pitch is simply an accurate pitch memory, and persons
who have it almost always acquire it at an early age. Nobody is born with
it, but there may be a genetic factor that facilitates its learning. It is also some-
times called perfect pitch.
Persons with absolute pitch vary in their precision, and some have been
shown to be remarkably accurate. Such a person can tell immediately what
key a particular piece is being performed in, and whether it is sharp or at.
These people usually insist on having a phonograph turntable with vari-
able speed so they can adjust it to the exact pitch desired.
Many musicians have the ability to recognize musical intervals very
quickly, and this is called relative pitch. It does not seem to be related to
absolute pitch.
Absolute Polarity A condition where the polarity of a musical signal is
not reversed no matter what devices are used in the signal path. When a
musical sound reaches the ear, the rst sensation the ear will feel will be
either a compression or a rarefaction of the sound wave. An example is a
bass drum. If you are situated facing the side of the drum that is struck by
the heavy drumstick, the onset of the audible sound will be a rarefaction
of the air pressure since the drum head initially moves away from you when
struck. This is called a negative polarity at the instant of the sounds arrival
at the ear. However, if you were on the other side of the drum, your rst
arriving sound will be a compression, or a positive polarity. There is some
controversy about whether the human hearing mechanism can tell the dif-
ference between positive and negative initial polarity, and under what con-
ditions. One faction says the initial polarity of music must be preserved in
music reproduction systems. This means the polarity of the audio signal
would have to be maintained throughout the path from the microphone
through preampliers, cables, signal processing equipment, the delivery
medium, the playback amplier, loudspeakers, and the loudspeaker wires.
Even more important, the relative polarity between the stereo channels must
be the same in order to maintain directional realism (a condition sometimes
referred to as being in phase). It is generally agreed that a positive pres-
sure into a microphone will produce a positive voltage at its output con-
nector, but there is no guarantee that all audio devices maintain absolute
polarity in themselves.
Absorption Coefcient A measure of the relative amount of sound energy
that will be absorbed when a sound strikes a surface. It is a pure dimen-
sionless number ranging from 0 to 1, and is represented by the Greek let-
ter (alpha). The actual percentage of sound absorbed by a surface is equal
to the surface area multiplied by the absorption coefcient and is expressed
in units called sabins, after the well-known Harvard professor and acousti-
cian Wallace Sabine.
An absorption coefcient of one means that all the energy striking the
absorber will be absorbed and none reected. An open window has an
absorption coefcient of one. If the absorption coefcient is equal to zero,
then all the sound would be reected and none absorbed. This is never pos-

4
 A-B Testing

sible in practice, for all materials absorb some sound energy. The absorp-
tion of any material will vary with frequency; that is, a material will have
a different absorption coefcient for different frequencies.
In order to predict the reverberation time of a room, the engineer
must know the total absorption of all the inside surfaces of the room, and
for this reason, many common materials have been measured and the
absorption coefcients have been published for a wide frequency range.
The measurement of alpha is not simple, and there are several techniques
recognized by various standards organizations and carried out by approved
testing laboratories.
A-B Stereo A-B stereo is a method of recording whereby two omnidi-
rectional microphones are spaced several feet apart and placed in front
of the sound source. This was the system used by Harvey Fletcher in his
historic demonstration of stereophonic reproduction of a symphony
orchestra in 1933. It is also known as spaced microphone stereo or spaced
omnidirectional microphones.
Two omnidirectional microphones placed in front of an orchestra pro-
vide the stereo listener with differences in loudness, differences in arrival
time (or phase), and to a certain extent, differences in spectral content, or
timbre, between the two channels. These cues are used by our hearing
mechanism to detect the locations of sources of sounds.
The exact placement of the microphones in an auditorium is very
important to obtaining the best balance of direct and reverberant sound.
No amount of equalization or other signal processing can correct
for poor microphone placement.
A-B stereo recording generally has more ambience, or room sound, than
other methods of recording, and if well done, has a musically satisfying
perspective. It is especially successful when done in an auditorium with
excellent acoustics. See also stereophonic.
A-B Testing A-B testing is the comparison of one audio device with another
by switching between the two while listening to the same signal being
processed by the devices.
A-B testing is thought by many to be a valid way of evaluating audio
components, but it is fraught with pitfalls. One problem in comparing one
device with another is that the test can say nothing about how accurate either
device actually is. In comparing two loudspeakers, for instance, one should
compare each with reality, i.e., with live music. This is not completely pos-
sible because the loudspeaker must be used in conjunction with other
components (amplifiers, recorders, microphones) in order to reproduce
music, and it is impossible to isolate the effects of each link in the chain.
Over the years, various live-vs.-recorded comparisons have been staged,
with interesting if not conclusive results. A great deal of effort is required
to make such a comparison useful.
Another problem with A-B tests is that the volume levels of the two sounds
being compared must be identical. Otherwise, the louder one will generally
be judged to be better, or at least as having a wider frequency range.

5
ABX Test

ABX Test A type of listening test where a person attempts to differentiate

between two or more audio devices such as compact disc players or
amplifiers. The test involves use of the so-called ABX box, a device from
ABX Corporation, rst invented in 1977 by Arnold Krueger and Bern Muller
(of the famous Southeastern Michigan Woofer and Tweeter Marching Soci-
ety or SMWTMS), later rened and marketed by David Clark and his ABX
Corp. [For complete details see David L. Clark, High-Resolution Subjec-
tive Testing Using a Double-Blind Comparator, J. Audio Eng. Soc., Vol. 30,
No. 5, (May 1982): 33038.]1 The device assigns the letters A and B to the
two devices to be compared and the letter X to one of the devices. In a series
of trials, the box randomly assigns X to be either A or B. The subject then
listens to A, B, and X and tries to determine whether X is the same as A or
B; the box keeps track of the answers. During each trial, the subject can lis-
ten to A, B, and X as many times as needed before answering. Over a long
series of trials, the listener can be expected to achieve a 50% correct score
simply by guessing. If a score is signicantly better than 50%, then one can
say that the listener is in fact able to tell the two components A and B apart
for a certain percentage of trials.
The ABX test is quite popular and is generally a trusted way to compare
audio products, but the statistical interpretation of the results of any given
series of tests is not as simple and straightforward as might be imagined.
It is not easy to distinguish between known correct responses (i.e., correct
responses based on actually hearing a difference) and correct guesses, espe-
cially if the percentage of correct answers is not too far from 50%. There-
fore, care must be exercised in the interpretation of the results. This is an
example of double blind testing, since the person giving the test also does
not know which device is X.
AC See alternating current.
AC-1 The rst Dolby digital audio coding system was AC-1 and was intro-
duced in 1984. A rened form of adaptive delta-sigma modulation, it
was intended for applications such as direct broadcast satellite systems
at a time when digital signal processing horsepower was very costly.
AC-1 featured a comparatively complex encoder, relatively few of which
would be needed by broadcasters, and simple, inexpensive decoders that
could be readily incorporated into mass-produced consumer receivers.2
AC-2 A Dolby adaptive audio coder that combines professional audio qual-
ity with a low bit rate, substantially reducing the data capacity required in
such applications as satellite and terrestrial links. The algorithm uses a multi-
band approach to take advantage of psychoacoustic masking, while a par-
tially adaptive bit-allocation scheme keeps codec complexity relatively low.
It is used by recording studios and postproduction facilities for such pur-
poses as remote monitoring and dialogue rerecording (looping) sessions.2

1. Thanks to Dennis Bohn, Rane Corp.

2. From Dolby Laboratories literature.

6
 Acetate

AC-3 The audio coder capable of a large amount of data compression used
in Dolby Digital audio was called Audio Coder 3, or AC-3. Dolby Labo-
ratories denes AC-3 thus: AC-3 is the technical name for the digital decod-
ing technology upon which Dolby Digital is based. The terms AC-3 and
Dolby Digital are generally used interchangeably, and the company now
usually refers to AC-3 by its trademarked name, Dolby Digital.
It is used in NTSC DVDs, some laserdiscs, and certain special CDs for
5.1 multichannel home theater use and has been adopted for atsc hdtv
broadcasting. AC-3 competes with dts Consumer.
The AC-3 codec has been designed to take maximum advantage of
human auditory masking in that it divides the audio spectrum of each chan-
nel into narrow frequency bands of different sizes optimized with respect
to the frequency selectivity of human hearing. This makes it possible to
sharply lter coding noise so that it is forced to stay very close in frequency
to the frequency components of the audio signal being coded. By reducing
or eliminating coding noise wherever there are no audio signals to mask
it, the sound quality of the original signal can be subjectively preserved. In
this key respect, a coding system like AC-3 is essentially a form of very selec-
tive and powerful noise reduction. Additional features include the trans-
mission of metadata, that can control playback parameters such as
dialogue-keyed playback level, and the ability of two-channel decoders to
down-mix multichannel soundtracks.
AC Bias See bias.
Accidental In a musical scale, the accidentals are the extra sharp and at
notes that are not part of the diatonic series. In the key of C on a piano, the
accidentals are the black keys.
It is surprising that these notes are called accidentals, so important
are they to music. In no way can one say that they exist simply by accident.
See also Appendix 8.
Acetate In the making of phonograph records, the rst step is called mas-
tering, and this consists of recording the audio signal onto a metal disc
coated with a thick lacquer made of cellulose nitrate with some castor oil
added as a plasticizer. This disc, either before or after recording, is often
called an acetate, which is a misnomer since it is not made of acetate. Sub-
sequent discs, or pressings, are replicated from the master record. It con-
sists of a 14-inch diameter aluminum disc, which is coated with a special
formulation of cellulose nitrate. The master tape is played while con-
nected to a recording lathe, which has a cutterhead that cuts the
groove in the nitrate material. The disc is then electroplated with several
layers of silver and pure nickel.
Finally, the nickel plating is carefully stripped away from the disc, a pro-
cedure requiring considerable skill and experience, and becomes a nega-
tive replica of the original disc. It can be used as a stamper to press records,
or it can be subjected to further plating in three-step processing, if many
hundreds or thousands of records are to be pressed. The acetate is also com-
monly referred to as the lacquer master.

7
Acousta-Voice

Cellulose acetate is one of the older magnetic tape base materials, and
has been largely replaced by polyester material with the trade name mylar
in professional applications. Cellulose acetate is a little unstable under
changing humidity, becoming brittle when very dry.
Acousta-Voice A trademark owned by the Altec Lansing company for
the rst commercial one-third octave equalizer designed for sound sys-
tem equalization. It was a graphic equalizer and was introduced in about
1967. Since then, many other similar units have appeared on the market.
Acoustic Feedback Acoustic feedback refers to the introduction of acoustic
energy from a sound system back into the same system.
One form of acoustic feedback is the disturbance of a record player sys-
tem by the sound produced by the system. For example, if a turntable play-
ing a record is placed directly on top of a loudspeaker cabinet, vibration
from the cabinet caused by the sound being reproduced will be picked up
by the cartridge and amplied along with the music. If the gain is suf-
ciently high, the level of the feedback sound will increase, nally causing
the amplifier, loudspeaker, or both to be overloaded. The result is a
loud rumble, or low-frequency tone.
Even before the feedback signal becomes regenerative, that is, before it
begins to increase in amplitude, the reproduced sound will be colored by
the feedback, causing muddiness in the low-frequency range. For this
reason, turntables must be very well isolated from vibration and should be
a relatively long way from the loudspeakers.
Another form of acoustic feedback is familiar to most people as the
howl of a sound reinforcement system when the volume control is set
too high. The amplied sound enters the microphone where it is amplied
again, making repeated trips around the loop and becoming louder and
louder. Acoustic feedback is the limiting factor in the gain of a sound sys-
tem, and much effort has been expended in developing techniques to reduce
it. Sound system equalization is one such technique.
Even if the gain is not high enough to cause a continuous howl, the fre-
quency response of the sound system will be distorted when the gain
nears the feedback point, and the apparent reverberation will be
increased also.
Acoustic Labyrinth A type of loudspeaker baffle popular in the 1950s in
which the rear of the woofer cone is coupled to a folded duct about 6 to
8 feet long. The inside of the duct is lined with sound-absorbing material
to reduce the tendency for the duct to resonate, and to absorb the energy
radiated from the back of the cone. The labyrinth acts as a load on the loud-
speaker, providing some control over the cone motion at the low-frequency
resonance. Reasonably good low-frequency performance can be obtained
from a well-designed system of this type, but it will be rather more bulky
than an infinite baffle of comparable performance.
Acoustic Lens An acoustic lens is a device sometimes placed in front of a
high-frequency horn loudspeaker to spread out the sound wave in order
to make it less directional. In this way, the angular coverage of the loud-

8
 Acoustics

speaker is increased. Acoustic lenses sometimes cause problems, such as

reecting some energy at particular frequencies back into the loudspeaker
and causing a somewhat rough frequency response.
Acoustic Radiation Pressure The steady state pressure exerted upon a
surface exposed to a sound wave. This pressure is very small, but lord
rayleigh, a prominent investigator in the nineteenth century, used the phe-
nomenon to construct an instrument to measure sound level. It consisted
of a small, extremely lightweight disc suspended on a ne lament, and
the sound radiation pressure would cause the disc to twist the lament in
proportion to the magnitude of the pressure.
Acoustic Recording An early recording where no electrical energy is used
to cut the record grooves is known as an acoustic recording.
Edisons phonograph was a purely acoustic device, i.e., all the energy
needed to emboss the tin foil on the cylinder was supplied by the sound
waves entering the recording horn. All recordings made before 1925
were acoustically powered. Many acoustic recordings sound surprisingly
good, and the rst electrical recordings were thought by many to be infe-
rior. This is probably true, considering that the acoustic method had been
highly perfected by this time and the rst electrical apparatus was by com-
parison somewhat crude.
Acoustics Acoustics is the study of sound and its interaction with the human
hearing mechanism.
Physical acoustics is the scientic discipline dealing with measurable
objective parameters of sound and its behavior in any medium. Audio
people are interested in the acoustics of air, but underwater acoustics is a
very large eld under abundant study. One of the most famous of the phys-
ical acousticians was lord rayleigh, the nineteenth-century British sci-
entist who wrote a monumental book on sound (The Theory of Sound).
psychoacoustics is the second major branch of acoustics; it is con-
cerned with the interaction of sound with the human hearing system.
Because mental impressions are difcult or impossible to measure objec-
tively, psychoacoustics deals with subjective responses to objectively
measured phenomena. Psychoacoustics had its beginnings in the work
of Hermann von Helmholtz, a well-known nineteenth-century German
physician and scientist, who wrote a book titled On the Sensations of Tone,
which covers the beginnings of psychoacoustics and also goes deeply into
musical acoustics. Psychoacoustics is a rich eld for study. A great deal
of the twentieth-century work in psychoacoustics was done by the Bell
Telephone Laboratories. Musical acoustics is a subset of both physical
acoustics and psychoacoustics and deals with how musical instruments
work, how and why music is perceived, performance techniques of musi-
cal instruments, and the acoustics of rooms. The oldest surviving work
on acoustics is contained in the Ten Books on Architecture, by Vitruvius, a
Roman architect who lived about 79 b.c. Vitruvius discusses in some detail
the acoustics of Greek and Roman theaters. He makes the curious sug-
gestion that certain large urns, or sounding vessels, will, if placed in

9
Acoustic Suspension

strategic locations about the audience area of a classical theater, improve

the acoustics.
Acoustics has been called the most basic of the sciences, everything else
being a branch of it. This may be an exaggeration, but it is true that energy
is transferred from the interior of the sun to its outer surface by acoustic
waves. Without them, we would not exist. Also, it has been postulated that
the most basic of the particles of which matter is made may be in the form
of tiny super strings that vibrate at their natural or resonant frequencies.
The strings vibrating at different frequencies correspond to the different
energy levels that the various subatomic particles are observed to possess.
This theory is a simplication of the quantum theory of physics. More basic
than this, it is difcult to get.
Acoustic Suspension A design for a low-frequency loudspeaker baf-
fle where most of the restoring force on the cone is the result of the elas-
ticity of the air in the sealed cabinet. See also woofer.
Active An audio device that requires a power source, such as from the
AC line or a battery, is called active, as opposed to a passive device. Some-
times amplifying components such as transistors or integrated cir-
cuits are called active circuit elements.
Active Crossover An active crossover is a crossover network that is
designed to be placed before the power ampliers in a biamplified or tri-
amplified loudspeaker system. Active crossovers are often designed so
the crossover frequencies can be varied, and they have advantages
such as better control of parameters and eliminating passive crossovers that
rob some of the power from the amplier.
Active Noise Canceling A method of reducing acoustic noise by adding a
polarity-inverted noise signal to cancel it. Some headphones are available
that work well in noisy environments such as airplanes. A small microphone
on the headphone receives the noise, inverts it, and adds it to the head-
phone signal that partially cancels or reduces the level of the noise heard
in the headphones. It is also possible to perform noise cancellation at very
low frequencies in small, enclosed spaces using loudspeakers to add the
inverted noise, obviating the need for headphones. An example is helicopter
cockpits.
Adaptive Decorrelation A digital signal processing (dsp) circuit
found in Dolby Digital 5.1 decoders for use in home theater sound systems
that takes the surround signal, which may be monaural, and derives two
signals to feed the surround speakers. The decorrelation process adds dif-
fusion to the sound eld to prevent its being localized at the speaker loca-
tions, which would occur if the monaural signal were fed directly to the
surround speakers. If the surround channels are in stereo, the decorrela-
tion circuit is automatically disabled. Adaptive Decorrelation is a registered
trademark of THX LTD.
ADAT, Alesis Digital Audio Tape Recorder A digital audio recording sys-
tem developed by Alesis that allows 8 independent CD-quality tracks to
be recorded on S-VHS video cassettes.

10
 Ader, Clement

In the ADAT system, the tape is formatted (an operation that can be
done before or during recording) with a proprietary Alesis time code that
is much more accurate than SMPTE, and it time-stamps the tape to the
single-sample accuracy; that is, 1/48,000th of a second. Because of this tight
synch, multiple ADATs are virtually free of incoherent phase between tracks
playing on different machines. This means that if you lock 2 or more ADATs
together, you create the equivalent of one large digital tape recorder and
one very wide, seamless piece of tape.
The ADAT was somewhat of a landmark product; it accelerated the home
and project studio market with affordable digital multitrack recording capa-
bility. It and others like it are sometimes described as Modular Digital Mul-
titracks (mdm).
ADC, or A/D Converter Abbreviation for analog-to-digital
converter.
Additive Synthesis The generation of complex musical waveforms in
electronic music synthesizers by the linear addition of sine wave com-
ponents whose frequency relationship is a harmonic series.
The rst musical instrument to use additive synthesis was the ill-fated
telharmonium of Thaddeus Cahill. The idea was very successfully
exploited by Laurens Hammond with the introduction of the rst Ham-
mond Organ in the early 1930s. A limitation of the Hammond scheme was
that the sine wave components added together were taken from an equally
tempered musical scale and, other than the octave components, are not pre-
cisely in tune with a harmonic series. This gives the Hammond Organ a
particular timbre that is easily recognized and, in some cases, highly prized.
The advent of the digital computer has greatly expanded the possibili-
ties of using additive synthesis with precisely controlled frequencies of the
components. However, fm synthesis is more economical of hardware, and
although somewhat more limited in the range of waveforms generated, it
is much more commonly used than additive synthesis in electronic music
synthesizers.
Ader, Clement A Parisian engineer who, in 1881, led a patent in the Ger-
man Imperial Patent Ofce covering Improvements of Telephone Equip-
ment for Theaters. This is no doubt the worlds rst patent having to do
with two-channel transmission of sound. In fact, it essentially covers a bin-
aural sound system for listening from the comfort of ones home to a musi-
cal performance taking place in a remote auditorium. Ader states in his
patent: The transmitters are distributed in two groups on the stage, a left
and a right one. The subscriber has likewise two receivers, one of them con-
nected to the right group of microphones and the other to the left. Thus,
the listener is able to follow the variations in intensity and intonation cor-
responding to the movements of the actors on the stage. This double lis-
tening to sound, received and transmitted by two different sets of apparatus,
produces the same effects on the ear that the stereoscope produces on the
eye.
The apparatus was demonstrated in the Paris opera house as a part of

11
ADR

the 1881 Paris Exposition. This was a remarkable achievement, for it

occurred only ve years after the invention of the telephone and almost
twenty years before the invention of the audion amplier.
ADR Short for Automatic Dialog Replacement, or Additional Dialog
Replacement. This is the technique of adding or replacing dialogue in a
motion picture by recording the appropriate voice in the studio after the
lm is shot and adding the resulting dialog recording to the sound track
of the lm. It is often used when the original dialogue recording cannot be
used due to noisy conditions on the set, etc. There are many techniques used,
including devices that operate automatically and can space the recorded
words afterward to t the lip synch.
ADR, Astra Digital Radio A European satellite system that is the largest
provider of satellite broadcasting of television and radio. It is said to offer
near CD-quality music.
ADSR Short for Attack Decay Sustain Release. It is the time constants asso-
ciated with signals generated by electronic music synthesizers. The attack
time is the time it takes the signal level to rise from zero to its maximum
value. The decay time is the time required for the level to fall to the sustain
value, and the sustain time is the time it remains at this value. The release
time is the time it takes for the level to fall to zero after the sustain time is
elapsed. These times are all adjustable to allow a wide variety of dynamic
possibilities in synthesized signals. The ADSR actually denes the enve-
lope of the generated signal.
In a typical keyboard-controlled synthesizer, the ADSR sequence is ini-
tiated each time a key is pressed, and that particular envelope will be

ADSR

12
 AGC

impressed on the generated note. The sustain part of the envelope can be
set to be as long as the key is held down, providing an organ-like timbre
to the signal. The shape of the envelope has a great deal to do with the tim-
bre of the signal, even if the waveform of the signal is the same. The ADSR
is also called an envelope generator.
Advance Ball A small highly polished ball, usually made of sapphire, which
is mounted on a cutterhead assembly of a cutting lathe used in the cut-
ting, or mastering, of vinyl records. The advance ball rests on the uncut
surface of the rotating acetate disc during the cutting process. Its pur-
pose is to control the depth of cut of the stylus. It must be kept scrupulously
clean to avoid damaging the disc surface, and it is usually polished with a
small leather pad before each cut is begun.
Not all cutterheads use an advance ball and instead control the depth
of cut with an electronic servomechanism.
Advance Head See preview head.
AES The Audio Engineering Society is a worldwide professional society
of audio people somewhat analogous to the Society of Motion Picture and
Television Engineers in the U.S. It was formed in 1947 in the U.S. and its
headquarters are in New York City. The AES is active in the setting of stan-
dards, research, and education. The riaa de-emphasis curve was origi-
nally formulated by the AES, and was known as the AES curve before record
manufacturers agreed to standardize. The standard de-emphasis curve for
30 ips analog tape recorders is from the AES.
AES3 Interface A digital audio transmission standard created by the AES
and standardized by the American National Standards Institute (ANSI
S4.40-1985) and the iec (IEC 60958-4). AES3 transmits two channels of dig-
ital audio data on a single twisted-pair 110-ohm balanced cable using XLR-
3 (IEC 268-12) connectors. An unbalanced conguration also exists.
Previously known as the aes/ebu digital interface, a term which is still com-
monly used. The consumer version is called S/PDIF.
AES/EBU, Audio Engineering Society / European Broadcast Union. See
aes3 interface.
AF, or A.F. Audio frequency (AF) means having frequencies within the
audible range, usually taken as 20 hertz to 20 kilohertz. This frequency
range is only an average for a large number of people. Many people hear
tones below 20 Hz, and some people hear above 20 kHz, although most
people are virtually deaf above 15 or 16 kHz. The audibility of a given fre-
quency depends on its level. See also fletcher-munson effect.
AFL, After Fader Listen A pushbutton sometimes found on older audio
mixing consoles. When the AFL button is pressed, the monitor output is
affected by the volume control, or fader.
AGC An Automatic Gain Control (AGC) circuit adjusts the gain of an
audio device in inverse proportion to the signal level entering the device.
An example is a portable tape recorder that is designed for speech record-
ing. When the talker is close to the microphone, the gain is reduced so
as not to overload the tape. As the level from the talker decreases, for instance

13
AIM

because of a greater distance, the gain increases to keep the recorded level
the same.
This type of machine is often used for radio interviews, and usually the
gain changes can be plainly heard as the background noise rises each time
the speaker pauses for a few seconds, only to suddenly fall the moment the
next syllable is uttered.
Another use of AGC is in radio receivers, where the gain is adjusted to
keep all stations at about the same level even though the signal strength
from the transmitter varies widely depending on distance.
AGC is also called AVC, for automatic volume control, or ALC, for auto-
matic level control.
AIM See amplitude intermodulation distortion.
Air Coupler A type of woofer system that consisted of a woofer mounted
on a wall near the oor or ceiling facing into the wall and centered between
two wall studs. About 8 feet below or above the speaker, a rectangular hole
was cut into the wall to let the sound out. The system was like a speaker
in an 8-foot-long organ pipe, which is resonant at about 30 Hz, giving a
pronounced peak in the response at this frequency. The air coupler was pop-
ular among some do-it-yourself audiophiles in the late 1950s, but its appeal
has mercifully waned.
Air-Motion Transformer A proprietary type of loudspeaker developed by
Oscar Heil. The air-motion transformer resembles a miniature accordion-
fold door in a strong magnetic eld. The folds contain conductors that carry
the amplied audio current, and the interaction with the magnetic eld
causes the folds to alternately squeeze together and move apart, generat-
ing a radiated sound wave. The system is said to be quite linear and low
in distortion.
Aliasing Aliasing refers to the production of spurious frequency com-
ponents in a digital audio system due to the presence of frequencies in
the signal that are higher than one-half the sampling frequency.
Visual aliasing can sometimes be seen in motion pictures, especially
westerns. If a wagon is moving at a speed where the spokes in a wheel are
passing by faster than the frame rate of the camera (24 frames per second),
the spokes will be seen to move backward, or forward at an incorrect speed.
This is also the principle of the stroboscope, which is a quickly ashing light
used to examine rotating machinery. When the lamp ashes at the correct
rate, corresponding to the rpm of the machine, the moving parts will appear
stationary.
Alignment Adjustments of various characteristics of audio devices are
loosely grouped together and called alignment procedures. Alignment is
somewhat analogous to tuning a car engine. Some of the steps in the align-
ment of an analog tape recorder are properly adjusting the head azimuth
angle and adjusting the playback de-emphasis and record pre-emphasis
and bias level circuits for optimum response.
Alignment Tapes The adjustment of an analog tape recorders many cal-
ibration controls is a painstaking and delicate job if the best results from

14
 Alkaline Cell

the machine are to be realized. To assist the technician in this procedure,

specially recorded and standardized tapes are used. Not only does this allow
optimum results from the machine, but it also ensures that tapes can be freely
traded between properly aligned machines with similar results on each.
Some of the parameters that have been standardized are operating level
and playback equalization, although there are many standards in
existence.
The alignment tape is made under carefully controlled laboratory con-
ditions to ensure accuracy. One of the tones on the tape is a high-frequency
azimuth alignment tone, usually at 16 kilohertz. While this tone is
replayed, the head azimuth is adjusted so that a maximum signal level
at the high frequency is attained, and the signals from any two tracks are
as nearly as possible in phase. For stereo recording, the phase response
of the two channels should be the same.
There will then be a series of tones at various frequencies, all at stan-
dardized levels. The playback equalization is adjusted until all the tones
are at the same level as the tape is played. The tape machine is now said to
have flat frequency response in the playback mode.
Next, there will be a tone of 700 or 1,000 hertz (usually 400 Hz in the
case of cassettes) recorded at the standardized reference level. The play-
back gain control is then adjusted to give the standardized output level,
usually 0 decibels on the tape recorder meter.
When the rst American tape recorder standards were being formulated
by manufacturers in the late 1940s, the state of the art of tape manufacture
was not very advanced. The reference level was dened as that record level
that produced 1% third-harmonic distortion or sometimes, total har-
monic distortion (thd). The maximum level used for signal to noise meas-
urements was typically dened as the level that produces 3% distortion.
This meant that the reference level was not the same for all tapes because
of variations in the tape manufacture. For practical reasons this level was
carried on as a fluxivity, regardless of the distortion produced. This
standard ux level was measured as 185 nanowebers per meter (nW/m)
at 700 Hz, which became known as Ampex Operating Level. Note that
uxivity changes at different frequencies.
By the 1970s and 80s, tapes were available with much greater capacity
for high recorded levels at low distortion, and the 185 nW/m reference level
is seldom used. The European din standard uses 320 nW/m, which is 4.7
dB higher than 185 nW/m. The method of ux measurement prescribed by
din has traditionally been thought to result in about 1 dB less output than
the American (ansi) standard. Thus, a tape played on a DIN-aligned
machine will be about 1 dB higher in level than when played on an ANSI-
aligned machine. This discrepancy has been refuted recently, and for prac-
tical reasons it is difcult to attempt to adjust most uxivities with excessive
precision. See also operating level.
Alkaline Cell A type of cell used for powering small tape recorders,
radios, etc., similar in appearance to the leclanche cell but using a zinc

15
Alligator Clip

anode and a steel cathode with an alkaline electrolyte. The alkaline

cell can be recharged if done slowly and has almost twice the capacity of
a similar-sized LeClanche cell. It also has better regulation than the
LeClanche cell, delivering most of its current capacity at over 1 volt.
Alligator Clip An alligator clip is a terminal for the end of a wire that is
designed like a small spring-loaded clamp with serrations along the edges.
It resembles the jaws of an alligator, hence the name. Alligator clips are used
to connect test equipment such as oscilloscopes and voltmeters, etc., to
various places in a circuit.
All-Pass Filter An all-pass lter, or all-pass network, as it is often called,
is an electrical circuit with a uniform amplitude response versus fre-
quency (also called frequency response), but it has a phase shift that
does not vary in a linear relationship with frequency. (A pure time delay
device will have a phase shift that is directly proportional to frequency, i.e.,
its phase shift increases at a constant rate with frequency.)
Complex filters often have signicant phase distortion because they
are not phase linear, and an all-pass network can be designed to correct the
phase anomalies without affecting the amplitude response. Sometimes all-
pass networks are used in psychoacoustic research to investigate the
audibility of phase distortions of various types. See also group delay and
group velocity.
Alternating Current An alternating current, or AC, is an electric current
that periodically reverses its direction, as opposed to a direct current,
or DC, which maintains a single direction. The rate of alternation is called
the frequency and is measured in cycles per second, or hertz (Hz).
We are all familiar with normal house current, which alternates at 60 Hz
in the U.S., Canada, and Mexico. audio signals are, of course, always alter-
nating, the frequency corresponding to the frequencies of the sounds the
signals represent.
AM See amplitude modulation.
Ambience Ambience refers to the acoustical qualities of a listening space.
reverberation, echoes, background noise, etc., are components of
ambience. On most music recordings, some of the acoustical characteris-
tics are recorded along with the music and are to a certain extent repro-
duced in the listening environment. For instance, an organ in a cathedral
will be recorded with a good deal of reverberation included, and this helps
to provide the proper acoustic atmosphere when heard in playback.
The true acoustical qualities of the source cannot, however, be very effec-
tively recorded and reproduced, mostly because the reverberation in
the source room is diffuse, coming from all directions. This helps to give
the concertgoer a sense of envelopment. This directional realism is not
preserved in a stereo recording, regardless of any advertising to the con-
trary. One of the promises of quadraphonic sound was the preserva-
tion and reproduction of the ambience of the hall where the original
recording was made, but it fell far short, a fact that contributed to its rapid
demise.

16
 Ambisonics

Ambisonics Ambisonics is a system for the reproduction of a three-dimen-

sional sound eld, something that quadraphonic sound systems prom-
ised and failed to provide. It uses two or more transmission channels and
four or more loudspeakers. Ambisonics is the result of research efforts
of the British Technology Group, formerly the National Research Devel-
opment Corporation, in England. The late Michael Gerzon (19451996), of
Oxford University, was responsible for most of the theoretical work that
led to its development.
According to Gerzon, all the information to re-create a sound eld in three
dimensions can be transmitted by four independent audio channels, and
only three channels are needed to reproduce surround sound in two
dimensions. These channels are not simply signals from microphones, but
are derived by suitable mixing, or matrixing, and phase shifting of sev-
eral microphone signals. Imagine three figure eight microphones and one
omnidirectional microphone all occupying the same point in space. The
gure eights are orthogonal, i.e., they are angled 90 degrees from each other.
One is oriented front-back, one is left-right, and the third is up-
down. The up-down signal is only needed if vertical localization is desired
in the reproduced sound eld. These four signals are called B-format sig-
nals, and there is one compound microphone made by Calrec which syn-
thesizes them directly from four closely spaced transducer elements.
These signals are fed into an Ambisonics decoder which suitably
matrixes and phase shifts them to obtain signals that are fed to amplifiers
and then to loudspeakers. The loudspeaker array may be of several different
congurations, with four or six loudspeakers. The Ambisonics decoder can
be adjusted to supply these different signals, depending on the requirements
and bank account of the listener.
In order to record or broadcast these B-format signals, they can be
encoded into two channels in several ways similar to techniques formerly
used for quadraphonic recordings. The one chosen by the Ambisonics
design team is the Universal HJ system designed by the BBC. The UHJ
composite signal is of either two or three channels. The two-channel ver-
sion is suited to recording on stereo records or cassettes, and the three-
channel version could be easily accommodated by compact disc with
suitable modications.
Once transmitted and received (or recorded and reproduced), these sig-
nals are able to feed loudspeakers for conventional monaural or stereo
reproduction. To recover the surround-sound information, a UHJ decoder
is used to recover the B-format signals, which are fed to the Ambisonics
decoder. In other words, the UHJ encoding system converts the B-format
signals into a form compatible with mono and stereo playback.
All this sounds very impressive on paper, but it is quite another thing
to bring it to commercial fruition. Unconventional mixing and recording
techniques are required for initial production, and a good deal of added
equipment is needed.
Ambisonics research can be said to have started with Alan blumlein,

17
Ampere

chief engineer for the BBC, in the 1930s. Interestingly enough, nearly all
the subsequent research and development has also occurred in Britain.
Ampere, abbr. Amp or A The ampere is the si unit of electric current; it
is dened as that current that, if existing in two parallel wires 1 meter apart,
will cause an attractive force between them amounting to 0.2 millionths of
a newton per meter of length. The ampere is also equal to a current of 1
coulomb of charge per second, or 6.2414 million million million electrons
owing past a point in 1 second. Incidentally, one should not speak of the
ow of current. The current exists; the charge ows. This is analogous
to the current in a river, which consists of the ow of water.
The mathematical symbol for current is I. The relationship between cur-
rent, voltage, and impedance is quite complex for alternating cur-
rent, but for direct current it is a simple proportion. The current I is
equal to the voltage V divided by the resistance R. See ohms law.
Amplier A device for increasing the amplitude of the voltage, current,
impedance, or power of a signal. An amplier is an active device and
strictly speaking should always increase the power of a signal (some ampli-
ers, such as certain distribution ampliers, may only reduce the imped-
ance level of the signal for the purpose of driving long lines).
The amount of amplication that an amplier provides is called its gain.
The gain is a ratio of its input signal level to its output signal level and is
simply a multiplier or a pure number. For instance, an amplier that dou-
bles the voltage of its input signal is said to have a voltage gain of 2. If its
output current is 10 times its input current, its current gain is 10, etc. An
amplier may have almost any combination of voltage gain, current gain,
or power gain, and some of these quantities may be negativesome
power ampliers can have a negative voltage gain but a large current gain
and thus a large power gain.
Amplier gain is also commonly expressed in decibels, or dB, and this
causes some potential (and actual) confusion because dB always expresses
a power ratio by denition, and one should not express the voltage gain of
an amplier in dB unless the input and output impedances are equal. The
confusion arises because the voltage gain of a typical amplier is not related
to its power output capability. For instance, if an amplier has a voltage
gain of 10, one is tempted to say the amplier has a gain of 20 dB because
it actually would raise the power level of a signal by 20 dB if the input and
output impedances were the same. In practice, however, this is very sel-
dom the case, and the true power gain is usually very much different from
what would be predicted by the voltage gain. For instance, a power
amplier could have a 1.6-megohm (M) input impedance and could
deliver 400 watts (W) to a 4-ohm load. Its output and input voltages could
be the same, leading one to say it had a voltage gain of 1, or 0 dB. But its
voltage output is 40 volts (V) when delivering 400 W into 4-ohm, and its
input voltage is also 40 V. The 40 V of input delivers only 1 milliwatt (mW)
to the input (402 divided by 1,600,000), so the power gain is actually 400
divided by .001, or 400,000. This is a true gain of 56 dB.

18
 Amplitude Modulation

For another example, consider a microphone preamplifier with a volt-

age gain of 1,000, an input impedance of 100 ohms, and a 1-mV signal. The
input power is .0012 divided by 100, or one ten-millionth of a watt. Its out-
put will be 1 V, and when connected to a 10,000-ohm load, will deliver .1
mW. Its true power gain is .0001 divided by .00000001, or 10,000, which is
40 dB, and its voltage gain of 1,000 would imply 60 dB of gain.
To reduce the confusion in practice, one says the voltage gain of the
rst amplier is 0 dB and the voltage gain of the second one is 60 dB, but
we believe it would be preferable to express voltage gains simply as pure
numbers and let the original denition of the much-beleaguered decibel
stand.
Amplitude Amplitude is the strength of a signal or a sound, without
regard to its frequency content. Amplitude measurements of audio sig-
nals generally refer to signal voltage rather than power, which is pro-
portional to voltage squared. The amplitude of a sound is usually measured
in decibels of sound pressure level. It is possible to dene and mea-
sure the sound power, but this is seldom done in audio work.
It is important not to confuse sound pressure level with sound
intensity, which has a very specic denition.
Amplitude Intermodulation Distortion This is the conventional inter-
modulation distortion as distinguished from phase intermodu-
lation distortion.
Amplitude Modulation, abbr. AM In AM radio transmission, the audio
signal is combined with a very high frequency sine wave called a car-
rier, in such a way that the strength or amplitude of the carrier is varied
in exact response to the amplitude and frequency of the signal. This is
called amplitude modulation of the carrier. The modulated carrier is then
transmitted at high power by the transmitter and the broadcast antenna,
where it is later received by radio sets that are tuned to the carrier frequency.
The modulated carrier is then demodulated by a process called detection,
and this recovers the original signal. The process of modulation and
demodulation of the carrier permits the transmission of the signal infor-
mation over long distances with relatively high efciency, and allows many

Amplitude Modulation

19
AM Stereo

radio stations to broadcast different programs simultaneously because the

carrier frequencies are different. The different carriers are easily separated
by the tuning circuits in the receiver set.
Strictly speaking, the carrier wave is not actually varying in strength,
but new frequencies, above and below the carrier frequency, are generated
in the modulation process. In typical AM broadcasting, these sidebands
extend in frequency 10 kilohertz on either side of the carrier frequency,
while the carrier frequency will be between about 500 kHz and 1.5 MHz.
AM transmission is capable of high quality if the received signal strength
is adequate, but interference from lightning and other electrical disturbances
can cause noise in long-distance reception. See also am stereo
AM Stereo AM Stereo was actually the rst stereophonic broadcast tech-
nology to be ready for widespread use. Engineers such as Leonard Kahn
developed and tested AM Stereo systems in the late 1950s that worked well
and might have made a very satisfactory solution for AM Stereo broad-
casting at the time. The Federal Communications Commission (FCC)
approved use of FM Stereo, but did not allow AM Stereo to be broadcast,
presumably because the proposed systems required greater bandwidth than
the existing 10 kilohertz allowed for each AM station.
Meanwhile, AM Stereo had not been forgotten. Engineers continued to
work on it and perfect its performance and sound quality. Various com-
panies, including Motorola, developed their own systems for transmitting
AM Stereo.
As the 1980s approached, the FCC decided to take another look at AM
Stereo. There were ve different AM Stereo systems being proposed by dif-
ferent suppliers, and the FCC struggled with the decision of which one to
choose as the standard.
At rst, the FCC decided to select one of the systems, designed by Mag-
navox, as the standard. As you can imagine, the proponents of the four other
systems complained, saying that their systems worked better than Mag-
navoxs and that the FCC should do some more research before trying to
decide a single standard again.
In 1993, the FCC chose the Motorola C-Quam system as the standard,
which was already the standard for all other countries in the world that
were broadcasting AM Stereo. C-Quam stands for Compatible Quadrature
Amplitude Modulation. The AM carrier being broadcast is accompanied
by another AM carrier of the same frequency, but it is 90 degrees out of
phase with the regular carrier. This is called a quadrature carrier, and its
main advantage is that it does not increase the bandwidth the station uses
to transmit its signal. The stereo information in the music signal is used to
modulate the quadrature carrier, while the normal carrier is modulated by
the sum of the two stereo channels. The system is compatible with mon-
aural receivers because the receiver does not sense the stereo carrier because
of the 90-degree phase difference.
Today (2003) there are several hundred AM Stereo stations in operation
in the United States, and thousands more in the rest of the world.

20
 Analog-to-Digital Converter

AM Suppression The ability of an FM tuner or receiver to reject ampli-

tude modulation of the received signal. Ideally, the receiver would only
be sensitive to frequency modulation, but some FM detectors are
also sensitive to am. Much of the interference and noise in broadcasting
appears as amplitude modulation, so a tuner with good AM suppression
will have less distortion and noise than a tuner with poorer suppression.
Analog An audio signal is an electrical replica, or analog, of the wave-
form of the sound it represents. The voltage of the signal varies up and
down (negatively and positively, in electrical terminology) the same way
as the sound pressure varies up and down at the microphone.
As long as the signal is in this form, i.e., a voltage that varies directly
with the sound pressure, it is an analog, and audio devices that use such
signals are analog devices. The majority of audio devices are analog in
nature, though digital devices are rapidly catching up in popularity.
An analog audio device need not be electrical; the Edison mechanical
phonograph was an analog device, the groove depth being an analog of
the sound pressure at the recording diaphragm.
Analog-to-Digital Converter, ADC or A/D In digital audio systems, the
audio signal (analog) must rst be converted to digital form before it
can be further processed. This entails sampling the signal at very short
successive time intervals, and converting the height of each sample to a dig-
ital word, which is simply a binary number indicating the amplitude of
the waveform at that instant.
The output of the A/D converter is a series of digital words, or sam-
ples, expressed in binary form, at a rate of 44,100 samples per second for
linear PCM such as on compact discs. Each word in CD digital audio is 16
bits, or binary digits. Other digital audio formats may have different word
sizes and sample rates. Before the signal can be fed to a common analog
amplier so it can be heard, it must undergo digital-to-analog conversion.
This recovers a replica of the original audio signal from the digital words.
ADCs are subject to certain errors, as follows:
1. Absolute accuracy: The difference between the ADCs actual input
voltage and the equivalent voltage value of the resultant output digital
code.
2. Conversion rate: The maximum rate at which the ADC can perform
valid conversions.
3. Conversion time: The time interval between the application of the con-
vert command and the output of a valid digital code.
4. Differential linearity: Any deviation in the ADCs measured step
width, which is ideally exactly one least signicant bit (lsb).
5. Linearity: A measure of any departure from a linear transfer func-
tion, expressed in percent of full-scale range, or as a fraction of one LSB.
6. Offset error: The mean value of the input voltage to set the output
code to zero.
7. Resolution: The amount of input voltage change required to change
the output by one LSB.

21
Analytic Signal

Analog-to-Digital Conversion

8. Scale error: The departure of the actual input voltage required to attain
a full-scale output code from its design value.
Analytic Signal An analytic signal is by denition a signal that has two
partsthe so-called real part, that is the same as a conventional signal, and
an imaginary or quadrature part that is always 90 degrees out of phase
with the real part. The analytic signal is very useful in several types of sig-
nal analysis. For instance, a conventional signal is represented by its wave-
form, which is simply a graph of the signal amplitude vs. time. The
waveform (and the signal itself) has positive and negative parts, and for
this reason it cannot be graphed on a logarithmic or decibel scale since log-
arithms of negative numbers are not dened. The waveform can be viewed
on an oscilloscope to see its overall shape, but it is difcult to see very small
components of the signal. However, the analytic signal can be transformed
by squaring the real and imaginary amplitudes, adding the squared val-
ues together, and then taking the square root of the result. This is the RMS
value of the signal vs. time, and is always positive. It is called the magni-
tude of the signal, and can be expressed in a decibel scale, where extremely

22
 Anti-Aliasing Filter

small components and variations in the signal can be easily seen. The ana-
lytic signal is also used in plotting the so-called heyser spiral.
Anechoic Literally, without echo. An anechoic room, or chamber, as it is
usually called, is designed to have as short a reverberation time as pos-
sible. This means the sound in the room will disappear as soon as the source
of sound is stopped. It is impossible to build a completely anechoic cham-
ber because there is no perfect sound absorber, but at high frequencies,
nearly anechoic conditions can be obtained. Low frequencies present a much
greater problem because the absorption of a surface is wavelength
dependent. The absorber must be at least as thick as one-half wavelength
to be effective. The wavelength at 100 hertz is about ten feet, so the problem
is obvious: it takes a very large chamber to accommodate low-frequency
absorbers.
Anechoic chambers are useful for the testing of loudspeakers and
microphones where reverberation would confuse the measurement. For
low frequencies, a good pseudo-anechoic environment is a parking lot on
a Sunday morning, provided no buildings or trees are nearby. This is actu-
ally a semi-anechoic space because of the reecting surface of the pavement,
but it can still be useful if the test object is placed at the pavement level so
no delayed reections will be measured. The effect of the so-called ground
plane is predictable and can be subtracted from the data.
Anhysteretic Literally, without hysteresis. In analog magnetic tape
recording, the hysteresis inherent in the process of magnetizing the tape
represents a large nonlinearity, and this causes harmonic distortion. The
use of bias in the recording process reduces the effect of the hysteresis
(makes the magnetization anhysteretic) and reduces the distortion.
Anode The anode in any electronic component, such as a silicon diode or
a vacuum tube, is the electrode normally connected to the positive volt-
age or, in a battery, the positive terminal. In vacuum tubes, the anode is
often called the plate, and its positive charge attracts electrons emitted by
the cathode or negatively charged element.
ANSI The American National Standards Institute (ANSI) is an organiza-
tion engaged in the establishment of standards.
Anti-Aliasing Filter Before a signal is subjected to the process of a/d con-
version, it must be passed through a low-pass filter to remove any com-
ponents that are higher in frequency than one-half the sampling
frequency. This is because it requires at least two samples per cycle to deter-
mine the existence and strength of a frequency component, that is, it would
require at least one hundred samples per second to encode a tone of 50
hertz. The A/D process will create spurious signals, called aliased com-
ponents, if this rule is not followed.
The frequency corresponding to one-half the sampling rate is some-
times called the nyquist frequency, and the requirement to limit sig-
nals into an A/D to below this frequency is sometimes called the nyquist
criterion.
In order to affect the audible signal as little as possible, an anti-aliasing

23
Anti-Imaging Filter

lter is designed to be very steep, having an extremely rapid fall-off above

the upper frequency limit. Such lters are complex, and they of necessity
suffer from phase shift, which under some conditions may be audible.
There are those who believe they can detect a certain harshness in dig-
ital audio systems that might be caused by this phase shift, and some
experiments have been reported where signals with and without phase shift
were directly compared. Most people are able to tell a difference in the
sounds, but it is another matter to tell reliably which one is the phase-shifted
and which one is not.
Anti-Imaging Filter In a digital audio system, in order to recover the
signal from the digital words, a d/a converter is used. The output of this
is a stair-step type of waveform which contains a great deal of high-
frequency energy called images. To reconstruct a smooth replica of the
original signal, the stair-step is passed through a steep low-pass lter called
an anti-imaging lter. It is similar, or even identical, to the anti-aliasing
filter found at the input of the A/D converter, but its purpose is quite
different.
Antinode A place of minimum sound pressure level in a standing
wave, as opposed to a node, which is a maximum in level.
Antiphase Signals are said to be in antiphase if they are of opposite polar-
ity, or 180 degrees out of phase with each other.
Antiskating Device The antiskating device on a record player tonearm
tends to force the arm toward the outer edge of the record to counteract the
skating force, which forces the arm toward the center.
The magnitude of the antiskating force is dependent on the stylus ver-
tical tracking force and must be adjusted accordingly.
Aperiodic Aperiodic means without period and refers to a signal or
waveform that does not repeat at a xed time interval. An example of an
aperiodic waveform is random noise. See also periodic.
In reference to mechanics the term is sometimes applied to a system with
a high degree of damping such that it no longer oscillates, or resonates.
For instance, a loudspeaker is said to be aperiodic if its mechanical reso-
nances are at least critically damped. Advertising claims aside, true aperi-
odic behavior in a loudspeaker is rare, if not nonexistent.
Aperture Time Errors In an analog-to-digital converter, the sam-
ple and hold circuit would ideally take zero time to determine the level
of the signal waveform and to hold this level until the next sample is called
for. However, it takes a nite time to charge the holding capacitor in
the sample and hold, and this is called the aperture time. Because the time
required to establish a new value of charge depends on the amount of change
in the signal level from one sample to the next, the aperture time will vary
with the rate of change in signal level, increasing for high-level, high-fre-
quency signals.
The starting time of the sampling aperture is also slightly uncertain, and
this is called jitter. In other words, lack of precision in the sampling time

24
 Apodize

leads to amplitude errors in rapidly changing signals. The errors involve

aperture time, uncertainty in aperture time, and jitter. The result is dis-
tortion of the sampled signal which rises with frequency.

Aperture Time Errors

Aphex Aural Exciter A trade name for a device that adds even-order har-
monic distortion to a signal for the purpose of making it more audi-
ble in the presence of other sounds. The Aphex system is used in multitrack
popular music recording, generally on only one or two signals. Voice tracks
are commonly so treated. The added even harmonics make the signal a lit-
tle brighter and crisper, with minimal change in loudness. Vocal lines stand
out more from the rest of the mix.
It is interesting that even-numbered harmonics have this effect, while
odd-numbered harmonics cause a closed-in, or covered, sound.
Apodize Literally, to remove the foot. To apodize is to remove a sharp
discontinuity in a structure, a signal, or a mathematical function.
One example of apodization is the rounding of the edges and corners
of loudspeaker cabinets to reduce the diffraction of sound waves from
them. Diffraction in this case results in a reection of the wave, and the
reected wave will interfere with the direct wave from the loudspeaker
cone, causing irregular frequency response. The rst proposed cabinet
apodization was probably by Harry F. Olson, of RCA Research Labs. Other
forms of cabinet apodization are the placement of sound-absorbing mate-
rial on the cabinet front around the speaker cone, especially helpful in the
case of tweeters.
Another example of apodization is the system of absorbing-wedges on
the walls of an anechoic chamber, which eliminate the effect of the acoustic
boundary at the walls.

25
Arc

In mathematics, the smoothing of discontinuities in a curve is called

apodization. The Hanning window, used to smooth the ends of the time
record in an fft analyzer, is a form of apodization.
Arc The phenomenon of electricity jumping across a gap is called an arc,
even though its path may not be shaped like a true arc. A spark is a very
short-duration arc.
When electric switches carrying signicant currents are opened, a
momentary arc is formed, and this accounts for the eroding of the switch
contacts. Arcing in electronic equipment is never desired, and effort is
expended on suppressing it.
One example of a useful application of an arc is the carbon arc lamp,
one of the rst types of electric light to be built. It predates the incandes-
cent electric light bulb and was used extensively for street lighting in the
nineteenth century. The arc between two carbon electrodes emits a very
bright, almost pure white light, and it is still used in some commercial
motion picture projectors. More modern projectors use xenon lamps, which
are actually arc lamps enclosed in quartz bulbs. The electrodes are of tung-
sten and the arc occurs in an atmosphere of xenon gas.
Arm The moving contact in a potentiometer is sometimes called the arm.
Armature The moving part of an electric motor or a magnetic transducer
such as a phono cartridge.
Armed When a channel of an audio recorder is set so it will record when
the machine is placed in record mode, it is said to be armed. In multi-
track recorders, the channels can be armed independently in any combi-
nation. The unarmed channels will not be recorded while the armed ones
will be. Some machines call this safe (unarmed) and ready (armed).
ASA, Acoustical Society of America An organization founded by Harvey
Fletcher of the Bell Telephone Laboratories in 1929. The ASA sponsors yearly
meetings where investigators in all areas of acoustics gather and deliver
papers on current research. The Journal of the Acoustical Society of Amer-
ica (JASA) is a very important forum for workers and students in the
extremely complex world of acoustics.
Also, the American Standards Association (ASA) was founded in 1928
by renaming and reorganizing the American Engineering Standards Com-
mittee (AESA) which had been founded in 1918. Then, in 1969, the ASA
was renamed ANSI, for American National Standards Institute. The sen-
sitivity of photographic lm was formerly rated by so-called ASA num-
bers, but ISO also published a standard for lm speed, and all lm
manufacturers conform to it. The speed numbers are almost identical with
the old ASA numbers. ASA lm speed numbers had nothing to do with the
Acoustical Society of America.
ASCII Acronym for American Standard Code for Information Interchange.
This is the standard digital code used rst by teletype technology and later
for communication of text between computers.
Asperity A small irregularity or imperfection in the surface of a magnetic
tape is called an asperity. Some gross asperities are visible on inspection

26
 ATRAC

of the tape, but most are not. Asperities are numerous in all tapes and pro-
duce asperity noise in tape-recorded signals. The author, in a previous
life working in a data analysis lab at a well-known airplane company, heard
a loud low frequency burp when listening to a reel of 12-inch-wide tape
with vibration data recorded on it. There was a relatively well-preserved
and well-attened mosquito embedded in the oxide surface! The name of
the tape manufacturer is withheld at the request of the American Anti-
Agony Society.
Asperity Noise Low-frequency noise in analog magnetic tape recordings
caused by asperities in the surface of the tape. Asperity noise is a type
of modulation noise.
Assemble Editing Editing of an audio or video program by making a mas-
ter copy of the various takes, rather than physically splicing the pieces of
tape together. Virtually all digital editing is done this way.
Assigns Push buttons on the input modules of a control console (audio
mixer), that connect, or assign, that particular input to any of the output
busses of the console.
Asymmetrical Limiter A special type of limiter that is used mostly in am
broadcasting to reduce the distortion caused by overmodulation. The
asymmetrical limiter limits the negative-going peaks of the audio wave-
form more than the positive-going peaks, and this helps to prevent the
degree of modulation from going below zero. In amplitude modulation,
there is no theoretical limit to positive-going peaks, but there can be no mod-
ulation below zero carrier amplitude. Any attempt to modulate more than
this results in hard clipping and gross distortion. This distortion is some-
times called splatter, which is quite descriptive of what it sounds like.
Asymmetrical Response This refers to the spectral shapes of the boost and
cut response curves of certain equalizers. Usually, the cut response curves
are quite a bit sharper than the boost response curves. The purpose of sym-
metrical response is to improve the subjective quality of the equalizers in
question.
ATIP, Absolute Time In Pregroove All CD-R and CD-RW discs have a pre-
embossed spiral groove that wobbles slightly. The groove keeps the laser
assembly tracking properly, and the wobble (sinusoidal with a frequency
of 22.05 kHz) provides timing information to the recorder. The wobble is
frequency-modulated with a +/-1kHz signal, which creates an absolute time
clocking signal, known as the Absolute Time In Pregroove. In addition, the
ATIP area has other encoded data, which purports to contain the CD-R man-
ufacturer, length of disc, and dye strategy.
ATRAC, Adaptive Transform Acoustic Coding A lossy, split band per-
ceptual coding and compression scheme, invented by Sony, for reduc-
ing the amount of data to be written on a MiniDisc. ATRAC offers a 5 to 1
data reduction ratio in the case of MiniDisc, employing the equivalent of
52 lter bands for spectral analysis and re-quantization. Later versions of
ATRAC vary the size of the sample blocks dynamically between 11.6ms
and 1.45ms according to the input signal to allow for temporal masking.

27
ATSC

ATRAC compresses compact disc audio to approximately 15 of the origi-

nal data rate with virtually no loss in sound quality.
ATRAC2 is a new version released by Sony that can compress digital
audio by a factor of 10 without sacricing audio quality, according to Sony.
ATRAC3 and ATRAC3plus are further developments of ATRAC2 designed
especially for audio transmission over the internet. ATRAC3plus achieves
a compression ratio of 20 to 1, again without loss of sound quality, accord-
ing to Sony.
ATSC, Advanced Television Systems Committee The Advanced Televi-
sion Systems Committee Inc. is an international, nonprot membership
organization developing voluntary standards for the entire spectrum of
advanced television systems.
ATSC is working to coordinate television standards among different com-
munications media focusing on digital television, interactive systems, and
broadband multimedia communications. ATSC is also developing digital
television implementation strategies and presenting educational seminars
on the ATSC standards.
ATSC was formed in 1982 by the member organizations of the Joint Com-
mittee on InterSociety Coordination (JCIC): the Electronic Industries Asso-
ciation (EIA), the Institute of Electrical and Electronic Engineers (IEEE), the
National Association of Broadcasters (NAB), the National Cable Television
Association (NCTA), and the Society of Motion Picture and Television Engi-
neers (SMPTE). Currently, there are approximately 170 members repre-
senting the broadcast, broadcast equipment, motion picture, consumer
electronics, computer, cable, satellite, and semiconductor industries. ATSC
incorporated on January 3, 2002.
When the dtv Standard was rst adopted by the ATSC in 1995, the ATSC
was strictly a United States organization with approximately fty mem-
bers, although Mexican and Canadian organizations played a signicant
role throughout the entire Advisory Committee process in developing what
was always expected to be a standard for all of North America, at least. In
January 1996, the ATSC modied its charter to become an international
organization, and the ATSC began to work with a variety of countries
around the world to explore the possibility of using the ATSC Standard for
their DTV services. Since that time, the ATSC DTV Standard has been
adopted by the governments of Canada (November 8, 1997), South Korea
(November 21, 1997), Taiwan (May 8, 1998), and Argentina (October 22,
1998), and many other countries are now considering the ATSC Standard
for possible use in their countries. Today, the ATSC has approximately 170
members from a variety of countries in North America, South America,
Europe, Asia, and Australia.
Attack Time Attack time is the time it takes for a compressor or limiter
to reduce its gain after a strong signal is applied to it.
transient signals that are shorter than the attack time of the device
will not be affected by the gain reduction, so it is important that the attack
time be as short as possible. In the case of compander systems, the attack

28
 Attenuator

time of the compressor and the expander must be exactly the same in order
to preserve accurate dynamics of the signal. In this case, these times are
selected for minimum audibility of noise pumping.
Attenuation Reduction in amplitude or level of a signal is called atten-
uation. It is usually measured in decibels. resistance in transmission
line wires, among other things, causes attenuation of signals. Sometimes
signal attenuation is desired, and attenuators are used to achieve it.
Attenuator An attenuator is a device for reducing, or attenuating, the
amplitude of a signal. A common example is the volume control. Many
times, attenuators are specially designed to reduce signals by various
numbers of decibels. Such attenuators may be either xed or variable. Inex-
pensive attenuators can affect the frequency response of a signal under
some conditions, and high-quality attenuators are never cheap.
Attenuators are sometimes called pads, or loss pads. We have wondered

Attenuators

29
Audio

for many years why the term pad came to be used to describe an atten-
uator, but we havent discovered the source of this usage.
Audio Literally, I hear in Latin. The term pertains to any signal, sound,
waveform, etc., that can be heard, as distinguished from ultrasonic
sound, radio-frequency signals, or video signals. In some peoples
minds, audio is nearly synonymous with stereo, or hi-.
Audio Analgesia A technique, introduced in the 1950s, for reducing sen-
sitivity to pain in dental patients. The patient wore a headset that repro-
duced a specially equalized pink noise signal at a relatively high level.
Usually, stereo music was added to the noise signal to further distract the
listeners attention from the violence being done in his mouth. The theory,
according to Dr. J. W. Gardner and Emory Cook, two of its developers, was
that the nerves carrying the pain impulses to the brain were effectively
jammed by the music and noise. The audio analgesiac, as it was called,
enjoyed some temporary popularity among dentists, but the novelty
apparently soon wore off, as did the analgesia.
Audiometer An audiometer is a device for measuring the hearing acuity
of a person. It actually measures the auditory threshold at various fre-
quencies, usually from 100 hertz to 8 kilohertz. The resulting fre-
quency response curve is called an audiogram. There are several accepted
ways to measure hearing threshold, the most common using calibrated ear-
phones. See also headphones.
Audion Audion is the name given the rst triode (three-element vacuum
tube) by its inventor, Dr. Lee DeForest in 1906, the year of the invention.
In an article on the history of the electronic age that DeForest wrote for the
ftieth-anniversary issue of Popular Mechanics magazine (January 1952),
appears the following paragraph: As a growing competitor to the tube
amplier comes now the Bell Laboratories thermistor, a three-electrode ger-
manium crystal of amazing amplication power, of wheat-grain size and
low cost. Yet its frequency limitations, a few hundred kilocycles, and its strict
power limitations will never permit its general replacement of the audion
amplier.
The thermistor is of course the transistor we know today. This
shows again the futility of trying to predict the future, even by so knowl-
edgeable an authority as Dr. DeForest. That 1950s issue of Popular Mechan-
ics makes fascinating reading today.
Audiophile Audiophile stems from Latin and Greek roots and means, lit-
erally, a lover of hearing. There are many vernacular synonyms for audio-
phile, including hi- nut, audio weenie, sound freak, etc.
The true dyed-in-the-wool audiophile is interested in the (real or imag-
ined) perfection of the reproduction of music. The music itself is of minor
importance; it serves merely as a vehicle for the apparatus. (In fairness, it
must be said that there are many music lovers and musicians who qualify
as audiophiles.) In contrast, many musicians are not extremely concerned
with the accuracy of the reproduction of the sound, but are able to listen

30
 Auto Reverse

to the music through the medium. I have been frequently amazed that
some musicians can put up with gross distortions of sound from their sound
systems.
Audio Taper A type of potentiometer designed for use as a volume con-
trol in audio equipment. Its resistance varies more in a logarithmic, rather
than linear, fashion with rotation of the knob. This gives a better correla-
tion between control rotation and subjective loudness of the signal.
Auditory Perspective A phrase used by the Bell Telephone Laboratories
engineers in the early 1930s meaning stereophony. According to John
Frayne, who was one of the engineers at the Western Electric company at
the time, the term auditory perspective was thought too cumbersome for com-
mercial use, and Western Electric held a contest among its employees to
coin a better term. Someone suggested stereophony and stereophonic, and these
names have been accepted usage ever since.
Auralization The use of computer modeling to simulate the sound eld in
a virtual space. The result is the ability to hear, via special headphones, the
acoustical properties of an auditorium that has not been built, allowing archi-
tects and sound system technicians to make changes in proposed designs
and hear the results. Some auralization programs are quite sophisticated,
and allow the modeling of different types of sound systems and different
locations in virtual rooms of varying designs. A very important feature of
these programs is the simulation of binaural sound in the earphones in order
to gain a semblance of directional realism.
Autocorrelation Function, or ACF A measure of how much a signal resem-
bles a delayed version of itself. It is a function of time delay and can vary
from 1 to -1, always having a value of 1 for zero delay.
The ACF of music is a signicant statistical parameter, especially the time
delay required for a given selection to decrease to one-tenth of its maxi-
mum value. This is called the effective duration of the ACF. Music with a
short effective duration of ACF requires a relatively short reverberation time
to maintain clarity, whereas music with a long duration of ACF (such as
organ music) sounds better with long reverberation times. Speech has a short
ACF, and this explains why intelligibility is reduced when it is heard in
very reverberant spaces.
Some years ago there was a single-pass noise reduction unit on the mar-
ket called an autocorrelator. It was actually a dynamic filter and had
nothing to do with autocorrelation. This was an unfortunate and mislead-
ing use of terminology.
Automated Mixdown, Automation See mixdown.
Auto Reverse Tape recorders that can play a tape in either direction, elim-
inating the need to ip the tape to play the other side. Virtually all recent
automobile cassette players are auto reversing. The mechanisms that
evolved could be quite interesting. Some machines used dual capstans, some
a central capstan. Separate heads were usually used for each side, although
some machines moved one head and some simply have one head with cores

31
Autotransformer

for all tracks. Some machines sensed the end of a side by the application
of a metal foil to the ends of the tape, sensed by an electrical contact. Some
machines used a photocell and looked for clear leader tape. Some cassettes
simply sensed the tension of the end of the tape. Professional tape machines
never record on 2 sides and thus are never auto reverse.
Autotransformer A transformer which has only one winding. The
winding will have taps along its length as well as terminals at each end.
Autotransformers are used in some multiple-loudspeaker sound systems
to couple the individual loudspeakers to a wire coming from the ampli-
fier. Selection from the various taps determines how much power the
speaker will accept from the line. The autotransformer does not provide
electrical isolation between the input and output signals as does a con-
ventional transformer, and cannot be used where this is required, such as
microphone preamp input transformers.

Autotransformer

Aux, Auxiliary Generally refers to an input connector in a preamplier or

integrated amplifier which is not specically designed to accept a phono
cartridge output. The auxiliary input has no de-emphasis or other special
equalization and accepts line level signals.
Aux Send See send.
Avalanche Diode See diode; zener diode.
AVC See agc.
AVI, Audio Video Interleaved A Microsoft multimedia le format, simi-
lar to MPEG, used by Video for Windows. In AVI, audio and video elements
are interleaved (stored in alternate segments) in the le.

32
 Back Coat

A-Weighting A-weighting is an equalization curve rst applied to

sound level meters in an attempt to make their measurements corre-
spond better to perceived loudness of sounds. It decreases the sensitivity
of the meter to frequencies below 1,000 hertz in a way that approximates
the Fletcher-Munson curve of equal loudness of 40 phons. (See fletcher-
munson effect.) If a sound level meter is measuring sound levels corre-
sponding to loudness levels of about 40 phons (the loudness of a 1,000
Hz tone at 40 decibels spl), then it makes sense to use the A-weighting
lter. At greater loudnesses, however, the contours of equal loudness are
more uniform with frequency, and a different curve would be required.
To build a sound level meter that approximates the sensitivity of the
human ear at all frequencies and loudnesses would be impossible due to
the complexity required.
Even though A-weighting is not suited for measuring loud sounds, it
has been used so much in the past, and there is so large a database of A-
weighted measurements, that its continued use is almost mandated. Most
standards for noise measurement, both American and European, call for
A-weighting.
There is one advantage to makers of noisy products in using A-weighted
levels: they give smaller numbers than nonweighted measurements would,
giving the impression of less subjective noise to the uninitiated.
AWG Acronym for American Wire Gauge, also known as the Brown &
Sharpe wire gauge. A numerical code, from 0000 to 40 is used to indicate
the diameter of electrical wire. The higher the gauge, the smaller the wire
diameter.
Azimuth In an analog tape recorder, the angle that the gap in the record
or playback head makes with the direction of tape travel is called the
azimuth angle, and it must be precisely 90 degrees to ensure proper high-
frequency performance. To adjust the azimuth, an alignment tape is put
on the machine and the output signal level is monitored while the
azimuth adjustment is slowly moved back and forth. The azimuth is cor-
rect when the highest level signal is obtained and the signals from the top
and bottom tracks are in phase.
The frequency recorded on the tape is usually 16 kilohertz. Of course
the accuracy of this adjustment relies on the accuracy of the alignment tape.
The narrow tracks of quarter-track machines require less precise adjust-
ment than the wider tracks of half-track or full-track tape recorders to attain
correct high-frequency response.

B
Back Coat Back coat is an electrically conductive coating on the back or
nonrecording side of recording tape, for the purpose of preventing static elec-
tricity from collecting on the tape and improving tape handling by provid-

33
Back EMF

ing a bit of friction on the normally smooth plastic. It is only seen on open
reel tapes, and always black in color from the conductive carbon particles.
Back EMF A voltage that is induced in a loudspeaker voice coil due
to the coils motion in the magnetic eld. The back EMF is in opposite
polarity to the voltage driving the coil, and so opposes that voltage. See
also motional impedance.
Back Plate The surface behind the diaphragm in a condenser micro-
phone. The exact spacing, conguration, and features of the back plate,
such as vents, determine many of the microphones characteristics.
Bafe An enclosure, or a large panel, into which a loudspeaker is
mounted in order to improve its low-frequency efciency is called a bafe.
Originally, the bafe was simply a at panel of wood that effectively
separated the radiation coming from the front of the speaker cone from
the rearward radiation, preventing cancellation at low frequencies. An ideal
at bafe would be of innite size so none of the rear wave would reach
the front regardless of the frequency. This would be a true innite bafe.
A practical innite bafe is a totally closed box that isolates the rear radi-
ation, although it also raises the low-frequency resonance of the speaker.
This effect can be used to advantage in certain loudspeaker designs.
There are a great many different designs for loudspeaker bafes, and it
seems that several new ones are patented each month. Many of these designs
are popular for a while and then rapidly fall into disfavor, but some, such
as the bass reflex, acoustic suspension, and horn, and their varia-
tions, have endured for a long time.
Balanced, Balance Balanced refers to audio lines in which the signal cur-
rent is not carried by the shield.
To transmit information electrically, whether by telegraph, telephone,
or an audio signal, two wires, or conductors, are needed. The current is
in the opposite direction in the two conductors. In most audio circuitry and
in interconnecting cables, one conductor of the signal is connected to the
chassis, or ground. This conguration is called single-ended. The chas-
sis thus carries signal current. Wires interconnecting audio devices are com-
monly shielded, with the shield grounded, or connected to the chassis. The
shield prevents electrostatic elds from inducing noise voltages in the inner
conductor. The shield carries audio current in such a cable.
The problem with this arrangement is that any noise induced in the shield
will be added to the signal. Typically such noise is 60-hertz hum induced
by magnetic elds produced by power lines, and it is troublesome in audio
cables that are more than a few feet long.
The way to prevent this type of noise interference is to use two conductors
for the signal and enclose them together in a shield, with neither conductor
connected to the shield. This is called a balanced conguration, and the
shield carries no audio current. The conductors in the shield are also twisted
together, which helps in reducing magnetically induced hum. The wires
change places every half-twist, and this reverses the direction of the induced
current, causing it to effectively cancel itself over the length of the cable.

34
 Balun

Also, any hum induced in a balanced circuit will be equal in the two
conductors, i.e., in phase, whereas the signal currents will be out of phase
in the two conductors. This in-phase noise induced in the two conductors
is called a common-mode current, while the signal current is called dif-
ferential. At the receiving end of the cable, it is not desirable to pick up the
common-mode current, so a differential amplifier is used. It is sen-
sitive only to the difference in voltage between the two conductors, and
rejects the common-mode part of the signal. This characteristic of differ-
ential circuits is called common-mode rejection.
Most long audio cables are balanced. Probably the most familiar exam-
ples are microphone cables. One way to balance an audio circuit is to use
a transformer at each end because the transformer is not sensitive to
the common-mode currents. This is the usual practice for microphones. It
is also possible, by using differential solid-state ampliers and audio
opamps, to construct balanced circuits without using transformers.
Sometimes the center tap of the input transformer primary is
grounded, which does not affect the balance. If it is not grounded, the cir-
cuit is said to be oating and balanced. See also differential ampli-
fier and Appendix 6.
Balanced Line See Appendix 6.
Balanced Power Mains AC power supplied to power audio equipment that
is congured as a transformer secondary with a center tap grounded. Such
a powering scheme can reduce power line induced noise by isolating the
audio equipment from the power line ground. While the neutral con-
ductor supplied by the power line is technically grounded, it can have a
lot of different kinds of electrical noise riding on it, and this noise can become
audible in audio systems. The U.S. National Electrical Code (NEC) recog-
nizes the scheme, but it is somewhat obscure and requires knowledgeable
installation.
Ballistics The dynamic behavior of the needle in a meter such as a VU meter
is called the meters ballistics. Such things as the time it takes for the meter
to read full scale after a 0 VU signal is applied, the distance the needle
will overshoot the 0 VU mark, and the time it takes to fall back when the
signal is removed are examples of ballistic specications.
The ballistics of a meter are very important because music and speech
are such irregular signals, having large peaks lasting only short times. Meters
with different ballistics will read quite different levels of such signals. It is
interesting that American VU meters have different ballistics from similar
meters on European equipment. Even the reaction of nonmechanical level
indicators may have their response described as ballistics. See also ppm.
Balun A transformer connecting a balanced circuit to an unbalanced circuit.
The Balun was originally used to connect radio transmitters to their anten-
nas. A common use is to connect a 75-ohm antenna coaxial cable to a 300-
ohm twinlead TV or radio receiving antenna input. They are also often
used to connect balanced audio lines to unbalanced lines or audio compo-
nents, although the term is seldom heard in this application. Nowadays,

35
Banana Plug

they mostly consist of solid-state circuitry rather than transformers,

although the transformer is often superior if complete electrical isolation
is required, such as to solve a ground problem.
Banana Plug, Banana Jack A banana plug is an electrical connector designed
to join wires such as speaker wires to binding posts, or special jacks called
banana jacks. The metal part of the plug resembles a small banana, hence
the name. A common conguration of banana plugs is two of them molded
together and spaced 34-inch apart. Such plugs allow fast engagemant and
disengagement, and can carry reasonably high currents.
This assembly is also called a GR plug, after the General Radio Cor-
poration, which introduced it many years ago. Banana jacks are often found
as terminals on older test equipment such as voms, vtvms, oscillo-
scopes, and oscillators, etc., but they are essentially obsolete for this
application. They are sometimes found as output connectors on power
ampliers.
Band An extent along the frequency dimension in which a signal exists
is the band. For instance, an octave band is a band one octave wide, while
a one-third octave band is one-third octave wide. The audio band, encom-
passing the frequency range of human hearing, is usually regarded to be
20 hertz to 20 kilohertz.
Band-Limited A signal is said to be band-limited if its frequency con-
tent is restricted to a particular frequency range. A band-limited signal can
be made with a bandpass filter, but some signals are band-limited by
their nature. For instance, the output signal of a compact disc player is
band-limited to 20 kilohertz by the anti-imaging filters built into
the player.
Bandpass Filter A bandpass lter is a lter which has a bandwidth. Band-
pass lters can be broad, having a wide bandwidth, or narrow, hav-
ing a narrow bandwidth. They may be xed in frequency and bandwidth,
or variable in either frequency and/or bandwidth. See also sound effects
filter; tone control.

Bandpass Filter

36
 Basilar Membrane

Band Reject Filter A filter that discriminates against signals in a specic

frequency band. The most common band reject lters reject a very narrow
frequency band, and they are usually called notch filters.
Bandwidth The bandwidth of a bandpass filter is the upper cutoff
frequency minus the lower cutoff frequency. It is thus the extent, in hertz,
of the frequency range, or band, passed by the lter.
Bandwidth is literally a frequency span, and is not necessarily connected
to the specication of a lter. For instance, the human voice can be trans-
mitted with good intelligibility if the frequency response of the trans-
mission chain extends from about 100 Hz to about 3,000 Hz. Thus, a
2,900-Hz bandwidth is needed to transmit voice. This is about what a stan-
dard telephone system attains. The audio bandwidth, however, is gener-
ally considered to be about 20 kilohertz.
The information-carrying capacity of any communication channel is
determined by its bandwidth, although there are many ways to encode sig-
nals within the bandwidth. Information theory is the study of how band-
width and encoding schemes affect the transfer of information. The most
signicant work in the eld of classical information theory was done by
Claude Shannon, of the Bell Telephone Laboratories.
Bar Originally dened as the average air pressure at sea level and 45 degrees
latitude, which is equivalent to 10.1325 newtons per square centimeter. The
si system has redened it as 10 newtons per square centimeter, or 10 pas-
cals, simplifying some calculations.
The reference pressure for the sound pressure level scale used to be
based on the bar, but is now based on the pascal.
Barkhausen Effect The Barkhausen effect is the tendency of the magnetic
elements, or domains as they are called, to inuence one another and to
become magnetized in one direction or another as a group rather than indi-
vidually. This means that a magnetic medium, such as recording tape, has
a graininess in its magnetic makeup, which is what causes most back-
ground noise or tape hiss in magnetic recordings.
modulation noise, which is only present in conjunction with a
recorded signal, is also caused by the Barkhausen effect, and is sometimes
called Barkhausen noise. The term comes from the name of the discov-
erer of the effect.
Barney A soft, heavy fabric zip-up cover designed to be placed over a motion
picture camera to reduce its noise output so sound can be recorded while
lming. See also blimp.
Barometer An ingenious device that indicates what kind of weather we
are having.A. Bierce
Base The control electrode in a bipolar transistor is called the base.
It is analogous to the gate in a field-effect transistor and to the grid
in a vacuum tube. Base is sometimes used to designate the plastic portion
of recording tape onto which the magnetic coating is applied.
Basilar Membrane The part of the cochlea in the inner ear that contains
the hair cells that generate the nerve impulses sent to the brain. Sound strik-

37
Bass

ing the ear drum is transmitted through the middle ear to the cochlea, where
it causes the basilar membrane to vibrate. This vibration excites the hair
cells, which emit impulses into the auditory nerve.
Bass Bass is that portion of the audible frequency which encompasses
the lower pitches. The bass range is generally considered to be from 30
hertz or so up to about 200 Hz. Sometimes frequencies around 200 to 300
Hz are called mid-bass.
Bass Intermodulation, or BIM Bass Intermodulation is a type of distor-
tion caused by the modulation of audio frequencies by subsonic
noise. The term is not universally recognized, but the effect is real.
Examples of low-frequency noises that cause BIM are flutter, wow,
and tonearm resonance. It might be thought that noises that are below
the frequency range of human hearing should be of no concern, but this
is far from the case. If an audible signal is amplitude modulated by a
subsonic waveform, the modulation will add sidebands to the signal,
and they are closely spaced in frequency to the signal. In some cases, these
sidebands can be as little as 10 decibels lower than the signal itself. This
constitutes 30% distortion.
Tonearm resonance can be a major offender, for the amplitudes of
the cartridge response due to it are frequently at levels comparable to
the actual signal level from the record. If the phono preamp is not lin-
ear in this low-frequency region, BIM will occur. The audible effect is a
general muddying of the sound and an imprecise or smeared stereo
image.
Bass Management A function built in to many surround-sound processors
in home theater systems that route the low-frequency content of the pro-
gram to the subwoofer. If the system has no subwoofer, the bass man-
agement function can be adjusted to route the bass to the full-range
speakers in the system.
Bass Reex The term bass reex was a trademark of the Jensen company
in the 1930s and referred to a type of loudspeaker enclosure that was
sealed except for a rather large hole, or port, below the speaker itself.
Such enclosures are now called vented systems, and the theory behind
their operation has been thoroughly investigated by two workers named
Neville Thiele and Richard Small. In essence, a vented enclosure is a reso-
nant box (a helmholtz resonator), even without the loudspeaker in it.
The loudspeaker will also have a low-frequency resonance, caused by
the springiness of the cone suspension and the mass of the cone and voice
coil assembly. Each one of these resonances results in a peak in the fre-
quency response. If the two resonant frequencies are made the same and
then are coupled by mounting the speaker in the box, an odd thing hap-
pens: the two resonances effectively cancel each other and two more res-
onances appear, one higher in frequency and one lower in frequency than
the originals. The lower of these two peaks helps to extend the response
of the system to a lower frequency than it would achieve without the
resonance.

38
 Baud

What actually happens is that the air in the port is moving out of phase
with the speaker at the original resonant frequency, partially canceling its
output. Below this frequency, the air in the port is in phase with the cone
motion, adding to the output and extending the lower limiting frequency
downward.
An important characteristic of such a system is the damping, because
if it is under-damped, the lower resonance will cause the response to hang
on after the signal stops. Some under-damped early Bass Reex systems
were called boom boxes for good reason. An example of a boom box
is the juke box, popular in the 1950s. See also hangover; thiele-small
parameters.
Bass Trap A bass trap is a low-frequency sound absorber specially
designed to reduce the effects of standing waves in recording studios. It
is a tuned absorber, and may have a narrow or wide range of frequencies
over which it operates. It usually consists of resonant wood panels with
absorptive material behind them, or suitably shaped slots in a wall or ceiling.
Battery A device for storing electrical energy and making it available for
powering electrical devices. Strictly speaking, a battery is two or more elec-
tric cells connected in series. Thus, a ashlight battery is really only a
cell.
There are many types of batteries available, and the variety can be bewil-
dering to the uninitiated. The least expensive type uses carbon and zinc as
the active elements, but the more common alkaline cell provides more power
and longer life at a somewhat higher cost. Lithium cells provide very high
energy density at high cost.
Many batteries are rechargeable, the most common of which are nickel-
cadmium (NiCad or NiCd) or nickel-metal hydride (NiMH). The NiCad
battery can be damaged if it is recharged when less than completely dis-
charged, although this has been disputed. To best prolong its useful life, it
should periodically be completely discharged and then fully charged. The
NiCad battery, even when fully charged, does not contain as much elec-
trical energy as the alkaline battery, so if the application requires long con-
tinuous use between replacements, the alkaline battery is a better choice.
Under high power draw conditions, the NiMH battery can deliver even
more energy. Lithium Ion batteries are rechargeable and have very high
energy capacity, but are tricky to handle safely.
Baud The speed of digital data transmission is measured in baud. One baud
roughly corresponds to one-half dot cycle per second in Morse code, or one
bit per second in a train of binary signals.
In the case of a modem, where digital data are modulated onto a car-
rier for transmission over a telephone line, the baud rate may be differ-
ent from the bit rate. In such a situation, the baud rate is equal to the
reciprocal of the shortest modulation element duration in seconds, and thus
may be faster than the actual bit rate if the bits have variable spacing.
The term comes from Baudot, a French engineer who devised a ve-level
code for telegraphs.

39
Baxandall Tone Controls

Baxandall Tone Controls A type of bass and treble tone control circuit
invented in Britain by Peter Baxandall in the 1950s. Baxandall tone controls
work by inserting variable frequency-selective elements in the negative
feedback loop of an amplier stage. To achieve treble boost, the feed-
back at high frequencies is reduced, and to achieve treble cut, the high-fre-
quency feedback is increased. The action is similar for low frequencies.
Baxandall circuits were very popular for many years, but they are not ideal.
The variable feedback causes the distortion to vary with the tone control
settings, and the inevitable noise added to the signal is higher than with
some other circuits and has a spectrum that varies with tone control settings.
Baxandall tone controls can be easily realized using circuits based on
opamps and are still used.
B Battery The relatively high voltage battery used to supply the tube plates
in the radios of the 1920s was called the B battery. At rst, B batteries were
made of a series of small 1.5-V leclanche cells, but later, rechargeable
B batteries became available. Usually these consisted of 60 or so edison
cells in test tubes connected in series. Every so often, the cells needed to
be relled with distilled water, which was no small bother. Few people today
realize the difculties involved in listening to the radio in the 1920s. It was
in 1938 that Robert Eichberg asked: Is the radio listener of today as happy
as the fan of yesteryear? I doubt it, for though programs have improved with
the equipment, the old thrill is largely gone. No more does one wonder, every
time the set is switched on, whether it will actually work! Perfection has
replaced chance in radio reception (Radio Craft magazine, March 1938).
The nomenclature lingers on, however, for we still call the power sup-
ply voltages in modern audio equipment the B+ and B. See also b plus.
BBC Over the years, the British Broadcasting Corporation (BBC) has been
responsible for much research, innovation, and improvement in the qual-
ity of broadcast audio, from the days of Alan blumlein, to digital audio
and the ambisonics system.
Beam Bottle Amateur radio operators slang for beam power tube.
Beam Power Tube A type of tetrode vacuum tube in which a set of plates
is so arranged that the electron stream from the cathode forms two beams
directed at the plate of the tube. The effect is to increase the efciency of
the tube, increasing its power output for a given size of tube. One of the
rst beam power tubes, and also the most commercially successful, was
the 6L6. All the modern power output tubes, such as the EL-34, EL-84, KT-
66, KT-88, 6550, and 5881, are derived from the 6L6 design. Incidentally,
the KT in the British type designations stands for kinkless tetrode,
ostensibly because of a relatively smooth characteristic curve.
Beats When two periodic signals or sounds are less than 30 hertz or so
apart in frequency, and if they are mixed together, the amplitude of the
combined signals will uctuate as they alternately reinforce and cancel each
other. These amplitude uctuations cause loudness uctuations and are
called beats. They can most easily be heard when two sounds of equal
strength are mixed.

40
 Bessel Crossover

Beats

The beat frequency is equal to the difference in frequency between the

two signals. When the difference in frequency rises to 30 Hz or so, the beats
are so fast that they are heard more as a roughness of the sound rather than
as distinct beats. Beats are audible in most musical instruments, and are
very useful in the tuning of instruments, for they will disappear when two
sounds being tuned together have the same frequency. When complex
sounds are mixed, the various partial components may beat at different
rates, giving rise to several simultaneous beat frequencies. It is easy to hear
this in the sound of the piano.
Beats are a physical phenomenon, and occur in air when sounds are
mixed; they also occur when electrical signals are mixed. In addition, there
are subjective beats, which are sometimes heard when a tone of one fre-
quency is heard by one ear and one of a different frequency is heard by the
other.
Bessel Crossover, Bessel Filter A type of crossover utilizing low-pass lter
design characterized by having a linear phase response (or maximally
at phase response), but also a monotonically decreasing passband
amplitude response (which means it starts rolling off at DC and contin-

41
Beta

ues throughout the passband). Linear phase response (e.g., a linear plot
of phase shift vs. frequency produces a straight line) results in constant
time-delay (all frequencies within the passband are delayed the same
amount). Consequently the value of linear phase is it reproduces a near-
perfect step response, i.e., there is no overshoot or ringing resulting from
a sudden transition between signal levels. The drawback is a sluggish roll-
off rate. For example, for the same circuit complexity a Butterworth
response rolls off nearly three times as fast. This circuit is based upon Bessel
polynomials; however, the lters whose network functions use these poly-
nomials are correctly called Thompson lters [W.E. Thomson, Delay Net-
works Having Maximally Flat Frequency Characteristics, Proc. IEEE, part
3, vol. 96 (Nov 1949): 487 490]. The fact that we do not refer to these as
Thompson crossovers demonstrates, once again, that we do not live in a
fair world.3
Beta The second letter of the Greek alphabet, . The current gain of a bi-
polar transistor is called the beta. A beta of 100 means that X amount
of base current will result in 100 times X of collector current in a transistor.
Beta Hi-Fi Beta Hi-Fi is a method of recording high-quality stereo sound
on Beta format video recorders. The technique was introduced by the Sony
Corp., using the same platform as the Sony betamax video recorder sys-
tem. The signal is companded (see compander) and frequency modulated
onto a carrier that is recorded in a gap in the bandwidth between the
chrominance and luminance portions of the video signal. The speci-
cations for Beta Hi-Fi are similar to those of most digital systems, offering
more than 90 decibels signal-to-noise ratio and no wow or flutter,
etc., even though it is entirely analog. The system was developed as an
improvement to the original linear audio tracks on videotape.
A similar but incompatible system was developed for the VHS video
format recorders invented by JVC. It is called VHS Hi-Fi.
Betamax The rst popular method of home video recording, introduced
by Sony, now abandoned. See also beta hi-fi.
Biamplication Some loudspeaker systems with multiple drivers do
not contain a crossover network, and they require a separate ampli-
fier for each frequency range. Such a system is called a biamplied loud-
speaker, and the technique is called biamplication.
Because each amplier in a biamplied system is called on to amplify
only a limited frequency range, it will generate less intermodulation
distortion, and the system can sound cleaner than a conventional loud-
speaker with internal crossover. The biamplied system still requires a
crossover network, but it precedes the ampliers and does not handle the
power amplier output. It thus can be of higher impedance and can be
made variable in its crossover frequency. This allows a choice of crossover
frequency to optimize the performance of the speaker system. Most high-

3. Thanks to Dennis Bohn, Rane Corp.

42
 Bias Trap

power commercial sound systems use biamplication. See also active

crossover.
Bias Bias can be either a small direct voltage applied to a tube grid or
field effect transistor gate along with the signal or a small direct cur-
rent applied to the base of a bipolar transistor along with the signal.
The bias moves the operating point on the characteristic curve of the
device to a more linear portion so as to reduce distortion. The bias must
be carefully adjusted to the proper value for optimum effect.
In an analog magnetic tape recorder, bias is an ultrasonic signal
(usually about 100 kHz) that is mixed with the audio signal and applied to
the recording head. The level of the bias is several times as high as the sig-
nal level, and its purpose is to reduce distortion by achieving anhysteretic
magnetization of the tape. As the tape passes by the gap in the record head,
the high-level bias causes the magnetic domains to be magnetically satu-
rated rst in one direction and then the other at the bias frequency. In so
doing they rapidly traverse their nonlinear magnetic hysteresis curve.
The audio signal is very much lower in frequency and lower in level than
the bias, and it appears to the magnetic particles between the head gap to
be a nearly constant magnetic level. As the tape passes the edge of the gap,
the effect of the bias is gradually weakened, and the magnetic particles are
alternately magnetized as before, but at gradually reduced peak magneti-
zation levels. This gradual reduction in the extremes causes the domains
to experience ever-smaller hysteresis curves and leaves the tape magnet-
ized at the slowly varying level of the audio signal. This is the process of
AC bias, so named because the bias current is alternating.
The level of the bias must be carefully adjusted for optimum effect. Too
little bias increases the distortion, and too much bias reduces the recorded
signal level and reduces the high-frequency response of the tape recorder.
Adjusting the bias is an important part of tape recorder alignment. See also
overbias.
It is possible to use a direct current bias in the tape head along with the
signal, and it does reduce the distorting effect of magnetic hysteresis, but
it is relatively noisy and results in a relatively small dynamic range. This
is called DC bias and has been used in the past in some tape recorders.
Bias Beat A high-frequency whistle is sometimes heard in recordings when
two analog tape recorders are connected in order to simultaneously record
the same signal. The beat is the difference in frequency between the two
bias frequencies of the two machines. To prevent bias beat, the bias oscil-
lators must be synchronized so they have the same frequency. This capa-
bility was provided in most commercial analog tape recorders.
Bias Trap A special filter in the output of a tape recorder to remove any
residual bias from the signal output when the recorder is actually record-
ing is called a bias trap. The bias, being of a very high frequency and at
a high level, can cause distortion and overload problems in subsequent
audio devices if there is no bias trap. When playing an already-recorded
tape, the bias oscillator is not operating, and no bias reaches the output.

43
Bidirectional Microphone

Bidirectional Microphone See velocity microphone.

BiFET A type of integrated circuit opamp that has a differential pair of field
effect transistors as the input stage. They are capable of very low noise
levels and very high input impedances and high slewing rates.
Bilar Winding A method of winding transformers with the primary
and secondary coils wound together in one coil. The wires of the primary
and secondary are simultaneously wound side by side in the coil. This
greatly increases the magnetic coupling between the two circuits and cor-
respondingly reduces the leakage inductance in the transformer. The out-
put transformers of high-quality vacuum tube power ampliers are
sometimes bilar wound.
Biex See whizzer cone.
Bifurcate To divide or split into two parts. Many audio connectors use a
separable pin and cup contact arrangement, where the spring-loaded bifur-
cated cup is designed to separate in half slightly when the pin is inserted
to maintain rm electrical contact for many use cycles.
BIM See bass intermodulation.
Binary A binary number system uses only two digits, as opposed to the
decimal system, which uses ten. Binary numbers consist of a series of ones
and zeros, and are easy to implement in computer systems because the pres-
ence of a voltage can indicate a 1, while the absence of a voltage indicates
a 0.
All digital audio systems use binary numbers to encode the signals.
Binaural Literally, having two ears. Because we have two ears spaced
apart by the width of our head, our hearing mechanism can make use of
several cues to allow us to determine the direction from which a sound is
coming. This ability is called localization. The localization cues depend
on the fact that the two ears hear somewhat different sounds. These dif-
ferences can be categorized as amplitude differences, arrival time (phase)
differences, and spectral (frequency response) differences. If a sound
is heard with greater loudness, at an earlier time, or with a greater high-
frequency content in one ear than the other, the sound will be localized as
coming from that side of the head. The greater the magnitude of the dif-
ferences, the farther to that side the sound will seem to originate. If there
are no differences, the sound will seem to be coming from directly in front
of the head.
All three localization mechanisms operate simultaneously. The distance
between the ears causes the phase differences, and the shading effect of the
head and the shape of the pinnas cause the amplitude and spectral dif-
ferences.
Binaural hearing is a very complex phenomenon and is not completely
understood. It is one of the topics of psychoacoustic research, which is
actively being carried out by several laboratories around the world.
Binaural Dissimilarity The opposite of interaural cross-correlation.
A measure of the degree to which the sounds heard by the two ears are dif-

44
 Binaural Recording

ferent. Listeners tend to prefer environments where the sound has a high
binaural dissimilarity rather than low.
Binaural Recording A system of sound recording where the binaural
localization cues are preserved, and the listener is able to achieve localiza-
tion of sounds as if he were actually at the site where the recording was
made. Advertising hype notwithstanding, stereophonic techniques, espe-
cially two-channel stereo, do not come close to sounding as realistic and
natural as properly made binaural recordings.
A binaural recording system consists of a dummy head, complete with
pinnas, with microphones at the entrances to the ear canals. These micro-
phones are sensitive to the differences in level, arrival times (phase), and
spectral content that a persons ears would sense if at the same location.
The microphones are connected to a two-channel recorder in the same
manner as if a stereo recording were being made. When listening to the
resulting recording, each ear must hear only the sound picked up by the
corresponding microphone in the dummy head, and for this reason head-
phones are used.
The effect of listening to a true binaural recording is completely differ-
ent from listening to a stereophonic or monophonic recording via head-
phones. Stereo through headphones gives a greatly exaggerated sense of
spaciousness and reverberation that is quite unnatural. A monophonic
recording heard via headphones sounds like it originates inside the head
instead of in front of the head. Binaural recording has neither of these defects.
It has been found that a binaural recording and reproducing system must
have the two channels identical in frequency response and particularly
phase characteristics within very narrow limits, especially over the mid-
range of frequencies from about 500 Hz to 3 kHz. It has also been found
that noise reduction systems (see compander) such as Dolby A, Dolby B,
Dolby C, and dbx degrade the localization ability because of slight varia-
tions in gain, noise, and frequency response between the two channels due
to mistracking. Cassette recorders do not provide very good binaural
effects because they rely on noise reduction techniques to achieve a satis-
factory signal-to-noise ratio. A good digital recorder, with a minimum of
signal processing devices between the microphones and the recorder, prob-
ably provides the best available medium for true binaural recording. The
headphones used must also be extremely well matched if the effect is not
to be compromised.
There have been many attempts to convey the binaural effect by using
stereophonically recorded material, with varying degrees of success. The
ster-bin developed by Benjamin Bauer was probably the rst of these. It
was further rened by Martin Thomas and others. These systems reduce
the exaggerations present in stereo recordings when heard via headphones
but do not eliminate the basic incompatibility of stereo and binaural
techniques.
Likewise, there have been many attempts to provide binaural hearing

45
Binaural Synthesis

using loudspeakers as reproducers rather than headphones. These sys-

tems are sometimes called transaural. A transaural system must com-
pensate for the fact that with loudspeaker listening, each ear hears both
speakers rather than just the one it should hear. To do this, a specially equal-
ized and delayed signal is fed to the right speaker in opposite polarity
to cancel the sound the right ear would hear from the left speaker. The same
is done in the other channel. The initial research in this technique was done
by Manfred Schroeder in the early 1960s. In the late 1970s, the JVC Corpo-
ration produced their biphonic system, which used these results. The
Biphonic system sounded quite convincing, but the listener had to be in
the sweet spot, a precisely equal distance from the two loudspeakers, for
the cancellation to work properly. JVC also experimented with a four-chan-
nel system called Q-Biphonic, which used two extra loudspeakers behind
the listener and two dummy heads for the original recording.
Others, including Cooper and Bauck, have demonstrated similar sys-
tems, and have reported exceptional results when listening in an anechoic
environment. The reverberation of the listening room inevitably degrades
the signal cancellation in the two ears and compromises the binaural effect.
The rst consumer-oriented transaural processor was the Lexicon CP-1
developed by David Griesinger. It was a digital system and had several pro-
grams for simulation of different types of acoustical conditions.
Supersonix was a binaural sound system for motion pictures rst
demonstrated in early 1989 in New York. It used a dummy head for orig-
inal recording, and each member of the audience wore a binaural headset
that could be plugged into phone jacks or could use an infrared link that
transmitted the signal to the headsets.
Binaural Synthesis A type of recording-playback system where the bin-
aural microphone dummy head is replaced by two microphones appro-
priately spaced. The microphone signals must be specially processed to
simulate the frequency-dependent time delays that would occur between
the ears of a true dummy head.
Binder Binder is the exible plastic material in which the magnetic parti-
cles are embedded in the coating of magnetic tape. It must be tough
enough to withstand wear and yet remain exible to allow proper head-
to-tape contact in the tape recorder.
Binding Post A binding post is a type of terminal which allows wires,
such as loudspeaker wires, to be connected to the output of an ampli-
fier. Binding posts usually are designed so they can accept alligator
clips and banana plugs as well as bare wires. They are almost always
color coded, the black one being the common, or ground, and the red one
being the hot terminal. They are sometimes called ve-way binding posts,
although one might be hard-pressed to come up with all ve ways.
Biphonic An early experimental transaural stereo system, using a
dummy head for recording, developed in the late 1970s by JVC. Several
years later, JVC extended the system to four channels and called it Q-
Biphonic. Neither system was commercialized.

46
 Bitstream

Bipolar Electrolytic A type of electrolytic capacitor that does not

need a direct polarizing voltage across it to operate properly. It is actually
two regular electrolytic capacitors connected back-to-back. Bipolar elec-
trolytics are commonly used in passive loudspeaker crossover net-
works because they are physically small for their capacitance value and
are relatively inexpensive. They will be marked NP (nonpolarized), or have
no + or - mark denoting a polarization.
Bipolar Transistor See transistor.
Birdies Extraneous, high-pitched whistles sometimes heard in tape-
recorded signals where the high-frequency content of the signal causes beats
with the bias signal. The term is also used to refer to high-pitched inter-
ference in am radio reception.
Biscuit The soft, hot piece of vinyl compound that is placed in an analog
record press and from which the record is formed. A biscuit is also stage-
hand-speak for a Clear-Com brand intercom box, which had a model num-
ber beginning with KB, for King Biscuit.
Bit Depth The size of the digital word used for the audio samples in PCM
audio. The bit depth of the CD system is 16 bits. Some devices can operate
at 20 or 24 bits, and some reduced fidelity formats use 8 bits.
Bite A subjective term for the sharp onset or attack of a musical instrument,
especially brass instruments. Bite is usually desired in recordings of such
instruments; however, excessive bite can result from positioning micro-
phones too close to the instrument or from distortion caused by a
momentary overload.
Bit Rate The time rate at which bits are transmitted in a digital system,
expressed in bits per second (bps). The bit rate of the compact disc is
4.3218 million bps.
Bits Binary digits, abbreviated bits, are the numbers used in the binary
number system. The reason the binary system is used in computers and in
digital audio is that only two numbers, 0 and 1, are needed. This sim-
plies the electronics system, for a 0 can be represented as the absence of
a voltage, while a 1 can be represented by the presence of a voltage. Bits
are commonly divided into groups of eight, which are called binary words,
or bytes.
Bitstream A general term for the ow of data in a digital data channel. Also,
a system for playback of compact discs developed by Philips that per-
forms 256 times oversampling (i.e., the sampling rate is 256 times the
normal 44.1 kHz rate) using 1-bit words. The normal digital word length
in CDs is 16 bits, and to encode the same information with 1-bit words a
technique called pulse density modulation is used, with the bits either +1
or 1 and a bit rate of 11.2 megahertz. Pulse density modulation is some-
what similar to delta modulation. The output of the sampler is a series
of pulses, either positive or negative, of equal height and equal width (as
opposed to the pulse width modulation technique, which varies the width
of the pulses to encode the signal). The average value of the adjacent +1s
and 1s is equal to the value of the signal waveform at that instant. To

47
Biwire

recover the audio signal, all one needs to do is to average the pulses; this
is accomplished with a simple analog low-pass lter of moderate slope,
which minimizes any phase shift problems that occur in conventional anti-
imaging filters.
The Bitstream system is more linear than conventional digital-to-
analog conversion, especially at very low signal levels, and the distor-
tion plus noise is said to be at least 106 dB below the maximum signal level.
Biwire or Bi-wire Biwiring amounts to having a passive loudspeaker
system with the inputs to the high-pass (going to the tweeter) and low-
pass (going to the woofer) sections of the crossover separated; they appear
at separate terminals on the rear panel. Normally, there are shorting bars
across them, in which case the speaker works conventionally (attaching
speaker wires to either section drives them both). BUT, if you want, you
can remove the shorting bars and attach a pair of wires separately to the
two lter inputs. This means that you can select the wiring for each sec-
tion of the speaker according to their sonic needs (ha!). Similarly, three-
and four-way systems can use triwiring or quadwiring. With these sys-
tems, you have further options possible in terms of which portions of the
system you leave strapped together, and/or you wire (and possibly drive)
them separately.
At the amplifier end, you have other options, depending upon the
depth of your pocketbook, or the volume of your bank vault.
1. You can just parallel the wires again and drive them from the same
amplier.
2. You can buy a second (or third or fourth if you are triwiring or quad-
wiring) power amp, parallel its input with your existing amplier, and drive
each section of the speaker from its own amplier. Youll note that both
ampliers are operating full range, just that one is not loaded at the high
frequencies and one is not loaded at the low frequencies. I dont think that
anyone has gured out that you stand to gain more if you lter the inputs
to the power amps rst. Of course, that would probably interfere with the
internal crossovers anyhow but it would still be benecial even if the lters
were somewhat removed from the internal crossover frequency. By elimi-
nating the unneeded signals from the ampliers input you minimize a poten-
tial source of intermodulation distortion.4 See also biamplification.
Black Box An electronic circuit of unknown or secret design is sometimes
called a black box.
Black Vinyl Black vinyl is a pseudonym for the lp stereo record, some-
times used to distinguish it from the compact disc. Vinyl need not be
black, but it hides bubbles and aws that might otherwise bother the con-
sumer.
Bleeder A large resistor connected across the big power supply capaci-
tors in audio amps. The bleeder draws current from the supply and pre-
vents the capacitor charging current from climbing too high. In tube

4. Thanks to Rick Chinn, Uneeda Audio.

48
 Board

ampliers, it prevents the capacitor from suffering over-voltage before the

tubes warm up and start drawing current.
Bleedthrough Synonymous with crosstalk.
Blimp A soundproof solid cover for a motion picture camera to silence it
when shooting on a sound stage. See also barney.
Blocked Impedance The electrical impedance of a dynamic loudspeaker
measured when the cone is held or blocked so it cannot move. If the
cone is allowed to move, the measured impedance will include the mechan-
ical components of the system as well as the electrical impedance.
Blooper Slang term for an early regenerative radio receiver. See also
regeneration.
BL Product The arithmetic product of the length of the wire in a voice coil
(L) and the strength of the magnetic eld (B) in the gap in which the coil is
immersed. The BL product determines the magnetomotive force that
the voice coil experiences for a given coil current and is a measure of the
efciency of the loudspeaker.
Blumlein, Alan The chief engineer at EMI in London in the early 1930s and
British pioneer in the study of stereophonic sound, Alan Blumlein prob-
ably made the worlds rst two-channel phonograph records in 1931. (Bell
Telephone Laboratories were experimenting with stereophonic recording
in the thirties also, but there seems to be no evidence of any communica-
tion between them and Blumlein.) The early Blumlein discs used the two
sides of the groove to record the two channels independently, as is done
today in modern stereo records. They were thought of as a laboratory curios-
ity at the time, and no attempt was made to commercialize them. (This was
also the case with the Bell Labs recordings.)
Blumlein also did fundamental work in the theory and practice of using
coincident microphones for stereo recording. He showed that a stereo-
phonic pickup can be achieved from a single point. An interesting and
detailed account of his ndings can be found in his British patent No. 394325,
issued in 1931.
BNC A type of coaxial connector mostly used on test equipment such as
oscillators, oscilloscopes, etc., and professional video and digital signals.
It is a bayonet type rather than a screw-on type and was invented by Paul
Neill and Carl Concelman of the Bell Telephone Laboratories, hence the
name BNC (bayonet-Neill-Concelman). Its threaded counterpart is the TNC,
but it is less often used although it is a superior connector. 75- and 50-ohm
designs are available.
Larry Blake reminds us that BNC also means Blimped Newsreel Cam-
era, the 35 mm Mitchell Camera model BNC that was the industry stan-
dard for over thirty years.5
Board The control or mixing console of a recording studio or sound rein-
forcement system is often called a board, even though it is generally a very

5. This entry copyright 19992003 by Larry Blake is reprinted with permission.

49
Boom

complex piece of equipment, containing microphone preampliers, vol-

ume controls, equalizers, etc. The lighting control console in theaters is
also often called a board. Sometimes printed circuit boards and
cards are called boards. Not to be confused with breadboard.
Boom A type of microphone stand that has a horizontal element to hold
the microphone at a distance from where the base of the stand touches the
oor. Small units of this type are called baby booms.
The long poles used to support microphones for recording of motion
picture sound are also called booms, or shpoles. Portable booms held
by sound technicians are designed to be light in weight. Studio booms on
sound stages and TV studios can be quite large and complex, allowing
the boom operator to sit on it and to physically zoom and rotate the micro-
phone in order to point it toward the source of the sound being recorded.
Boom is also a derogatory term for the response of low-frequency
loudspeakers that are poorly damped. See also bass reflex.
Boom Box Some poorly designed bass reflex loudspeakers might be
called boom boxes, but there is also a commercial device with this name.
It is actually a subharmonic synthesizer, dividing the lowest octave of the
signal input frequency by two to generate a signal one octave lower.
This signal is then either mixed with the input signal at the output of the
device or is sent to a separate amplifier and then to a subwoofer.
The boom box, of course, has nothing to do with accuracy of sound repro-
duction and is meant to add another octave to the bass response of a sound
system. It has been used to good advantage in some movie theater sound
systems, where it can make sound effects such as explosions more con-
vincing. Also a portable stereo, AKA ghetto blaster.
Boomy The subjective effect of excessive bass response or lack of low-
frequency damping in a loudspeaker system. Some early (and some
present-day) resonant loudspeaker systems were called boom boxes
because of this effect.
Boost Boost refers to an increase in amplitude, usually within a specic
frequency or frequency range. Boost/Cut equalizers, the most com-
mon of which are tone controls, cause an amplication (boost) or a
reduction (cut) of selected frequency ranges.
Bootstrap A clever arrangement where the apparent impedance of a cir-
cuit element is reduced by applying an appropriate feedback voltage to
it. Bootstrapping can improve the linearity of a circuit and thus reduce
distortion. It is especially useful in circuits that are required to carry a
very wide range of power or voltage levels. Bootstrapping is also some-
times used in equalization circuits. Probably the term booting a com-
puter comes from the early computer improvement of being able to start
by itself, by its own bootstraps, so to speak.
Bottle Slang term for a vacuum tube, especially power output tubes used
by amateur radio operators.
Bottoming The mechanical striking of the moving parts of a loudspeaker
or other transducer against the magnet structure as a result of too high

50
 Breakout Cable

a drive level. Bottoming, or bottoming out, can quickly damage a speaker,

so care must be taken to prevent it. The acoustic load on a speaker has a
great deal to do with how easy it is to bottom and is therefore a factor in
the protection of the system. woofers in innite baffles, for instance, are
effectively loaded at very low frequencies, whereas woofers in vented cab-
inets are not and are thus more prone to damage from very low frequency
signals or transients.
Bounce On a multitrack recording device, to re-record mixes of certain
tracks onto other tracks, thus freeing tracks for other use.
Boundary Microphone See pzm.
B Plus, or B+ In the early days of radio, receiving sets were battery-
powered, and the batteries were called by different code letters depending
on their function. The A battery was a low-voltage, high-current bat-
tery that was used to supply current to the tube heaters. The B battery
was a high-voltage, relatively low-current battery used to supply the pos-
itive voltage for the tube plates. The C battery was of low voltage and
low-current capacity, and supplied the bias voltage for the tube grids.
Nowadays, with tubes rarely used and battery-powered tube equipment
seldom used, these terms are obsolete. audio equipment still requires elec-
tric power and this is obtained from the AC line via a power supply, which
is a group of circuits that generate and regulate various voltages for use
by the device. The main voltage, which is used to supply current to the tran-
sistors, is called the B+ as a holdover from the old days. The plus means it
is positive with respect to ground.
Many amplifiers and other devices also need a voltage that is nega-
tive with respect to ground, and this is called B. A power supply that makes
both B+ and B is called a bipolar supply. This is not to be confused with
bipolar transistor. Sometimes the B+ and B are called supply rails.
Breadboard A breadboard is an experimental electronic circuit, put
together temporarily to test a particular conguration, or topology. The
term comes from the early days of radio when experimenters supposedly
used the family breadboard as a base on which to construct such things as
crystal sets.
The term is also used as a verb; to breadboard is to make a breadboard
circuit.
Break Before Make A type of selector switch in which the movable contact,
or wiper, breaks the connection with one circuit before it makes a new
contact with the next selected circuit. Most audio applications call for make
before break switches to reduce clicks and pops when switching.
Breakdown Voltage The reverse voltage that, when applied to a semi-
conductor junction, causes a sudden increase in reverse current. The
junction is then said to be operating in the avalanche condition.
Breakout Cable Any form of cable adapter that converts a multiconduc-
tor connector into many separate cables. Since many products are too small
to physically mount all the connectors needed, a breakout cable converts
the multiple contacts of a small connector into several cables with normal

51
Breakup

connectors. Another example is a microphone cable breakout, a multiple-

wire exible cable containing several microphone lines that is connected
to the microphone inputs of a mixing console with a multi-pin connector
and terminates in several microphone input connectors (usually XLR-3
female) at the far end. These breakout cables, also commonly called
snakes, are often used in recording studios and sound reinforcement ven-
ues, and they are much more convenient to deploy and use than a similar
number of individual microphone cables. Multichannel cables that do not
employ a multi-pin connector at the end are usually called snakes.
Breakup See cone breakup.
Breathing The audible variation in the background noise level of a sound
system that has some type of noise reduction is called breathing, or
noise pumping. See also compander.
Brickwall Filter Some low-pass filters have such a steep cutoff slope
that they resemble a brick wall. Brickwall lters are of several design
types, and are used for anti-aliasing and anti-imaging lters.

Brickwall Filter

Bridge Disc A compact disc that contains information that can be read
by CD-ROM/XA drives and also CD-I players. The specications for the
Bridge Disc were released in the White Book in 1991. The Kodak Photo
CD is an example of a Bridge Disc, and it can be played on a Photo CD
player, a CD-I player, and a CD-ROM/XA player.
Bridging In a sense, bridging is the opposite of matching. When the input
of an audio device is connected to the output of another device, it is a bridg-
ing connection if the second device does not appreciably load the previ-
ous device and essentially no power is transferred. The second device is
sensitive to the output voltage of the rst device, and this is maximized
when the loading is minimized. This is an example of deliberately mis-

52
 Bucket Brigade

matching the impedances. Most audio connections are bridging, although

in systems involving very long lines (e.g., telephone systems), the imped-
ances must be matched to prevent reections of the signal at the ends of
the line from causing an audible echo.
Generally, a circuit is said to be bridged if the load impedance is at
least ve or ten times the output impedance.
Another use of the term bridging is in the connection of two power
ampliers in such a way as to give one double the power output. To do
this, the signal is fed to one amplifier normally, but it is inverted (its
polarity is reversed) before sending it to the other amplier. This puts
the amplier outputs out of phase with each other. The load is then con-
nected between the high, or hot, terminals of the two ampliers. The effec-
tive voltage swing seen by the load is thus doubled and the current carrying
capacity is unchanged, so the power output is doubled. stereo ampliers
are sometimes bridged in this way to make mono ampliers of more
power.
Bridging Transformer An audio transformer with a relatively high
impedance primary that is designed to operate when connected to a low-
impedance circuit so as to prevent any loading of the circuit. Bridging
transformers permit connection of several circuits, such as lines, to a sin-
gle circuit without affecting its impedance. Usually a bridging transformer
is made by simply adding series resistors in the primary circuit. The out-
put, or secondary, of the bridging transformer is not able to supply any
signicant power to its load. The bridging transformer is an example of a
case where impedances are deliberately mismatched.
Broadband Broadband means encompassing or consisting of a wide band
of frequencies, as opposed to narrowband, which contains a relatively
small bandwidth. Almost all music is considered a broadband signal.
Some equalizers, such as tone controls, are broadband in action
because they affect a large bandwidth of the signal.
A high-speed internet connection, such as DSL (digital subscriber line)
or a cable modem is also called a broadband connection.
Brute Force See regulation.
Bucket Brigade The bucket brigade, also called a charge transfer device
(CTD) or a charge-coupled device (CCD), is a type of time delay inte-
grated circuit for sampled signals. It consists of a series of small capac-
itors and a switching circuit whereby the charge from one capacitor is
transferred to the next one and to the next, etc., in response to a series of
timing pulses. The charges move along the line much like the buckets of
water in a reghters bucket brigade. It is important that the size of the
charge does not change as it passes from one storage capacitor to the next,
for otherwise the signal will become distorted as it is passed down the
line.
The bucket brigade is actually a specialized form of shift register, but it
is not a digital device, for the charges are analogs of the magnitude of the

53
Bucking

sampled signal. If the signal has been digitized before entering the delay
line, then the amplitude accuracy of each charge is not important, because
each charge simply represents a binary digit, or bit, rather than an instan-
taneous magnitude of the signal itself. For this reason, shift registers are
much better at delaying digital data than analog data.
Bucket brigades are used in some relatively inexpensive audio delay
devices such as reverberators.
Bucking The cancellation of one signal or frequency component of a
signal by another signal with equal amplitude but opposite polarity is
called bucking. See also hum bucking; phasing; flanging.
Buffer A temporary location in a computers or digital signal processors
memory for temporarily storing digital data that is being processed or
transmitted.
Bulk Eraser See degausser.
Burst A burst is a test signal that lasts only a short time, typically a few
milliseconds. A burst of a sinusoid is called a tone burst, and a burst
of white or pink noise is called a noise burst.
The term burst has a special meaning in television systems. The NTSC
color burst is a short burst of a high-frequency carrier that is included
as a part of the video signal. It is modulated with the color information.
Burst Error In the reading of the individual bits from a compact disc
(CD) or a digitally encoded tape, two types of erroneous readings can occur.
Individual bits can be misread at random; these are called bit errors and
are caused by tiny imperfections on the CD or tape surface. Errors that occur
in groups of adjacent symbols are called burst errors. They are caused by
dirt and scratches in the medium.
Error correction schemes in digital tape and CD systems depend on
knowing the data adjacent to the burst. The maximum correctable burst
error is a very important specication for such an error-correction method.
Bus A bus is a point in an electronic circuit where many connections are
brought together. For instance, a ground bus is a common connection for
all the grounds in a device. The bus will then be connected to the chassis
at one point, usually near the signal input connector. A mixing bus is the
point in an audio mixer where the various signals from different micro-
phone preamps are connected. It is the point at which the mixing is actu-
ally done.
Sometimes the term is used for a bundle of wires that are not connected
but are used for parallel transmission of digital data, sometimes in two
directions. This is called a data bus.
Butterworth Filter A multiple-section filter designed in accordance with
the so-called Butterworth polynomials.
The Butterworth lter is usually an RLC lter, meaning it contains resis-
tors, capacitors, and inductors, although it can be realized as an
active lter using capacitors, resistors, and amplifying stages. Because of
its lack of any peak in its response curve, it is sometimes called a maxi-
mally at lter. It usually contains more than two sections and is made

54
 Cannon Connector

with up to twelve poles, or six sections. It is suitable for use as an anti-

aliasing lter in digital audio devices. It was named after S. Butterworth,
a British engineer who rst described it in his paper On the Theory of Fil-
ter Ampliers, Wireless Engineer, vol. 7 (1930): 53641.
Butt Splice A type of tape splice used in editing where the ends of the
two pieces of tape to be joined are cut at a 90-degree angle and then butted
together and fastened with splicing tape. The butt splice results in an instan-
taneous transfer from one take to the other and frequently adds a tran-
sient, or click, to the sound. For this reason, butt splices are seldom used,
and the tape is usually cut at an angle, such as 30 or 45 degrees, so the tran-
sition is more gentle.
In the case of digital editing, which is usually all electronic, one sig-
nal is rapidly faded out while the other is faded in. This is called a cross-
fade. If the digital splice is instantaneous, it is then still called a butt splice,
even though no actual splicing of the tape is performed.
BWF (Broadcast Wave Format) A digital audio le format introduced by
the EBU in 1996 to facilitate professional editing and le exchange. It is a
special case of Microsofts wav format, but adds extra information deemed
important for professional use such as titles, date, time, origination, edit-
ing data, etc. It is increasingly used on daws and digital recorders.
BWF only carries linear pcm and mpeg audio. Its lename extension is
not .bwf as one might expect, but .wav, as it is merely considered a subset
of .wav. See also wav.
Byte In digital systems, the binary bits are grouped into words. A byte
is a binary word of eight bits. The information-storage capacity of digital
systems is measured in bytes or kilobytes. The astute reader may assume
that a kilobyte consists of 1,000 bytes, but this is not true, for a kilobyte actu-
ally contains 1,024 bytes! This is because 1,024 is a power of 2, whereas 1,000
is not. Learning such irregularities is one of the keys to becoming procient
in the domain of digital systems and computers. See Appendix 10.

C
Calendering In order to reduce the asperities in the surface of magnetic
tape, the tape is squeezed between giant polished metal rollers. This
process is called calendering, and is derived from the paper industry, that
uses a similar process of the same name to make very smooth shiny paper.
Cakebox When ordering bulk quantities of blank CD-R discs or other opti-
cal storage media, they often are packed on a spindle and covered with
shrink-wrap plastic. These units are called cakeboxes, although they usu-
ally are rather tall and thin compared to a standard edible cake unless they
are packed fty discs or fewer at a time.
Cans Slang for headphones.
Cannon Connector, Cannon Plug Literally any connector made by Can-

55
Cantilever

non, but popular usage usually means an XLR-3 connector, a type that was
introduced by Cannon. This is the most common connector used on micro-
phone cables. XLR-3 connectors are made by many manufacturers and are
a worldwide standard. See also xlr connector.
Cantilever The tiny rod, usually of light tubular metal, that holds the sty-
lus in a phonograph cartridge and couples its motion to the sensing ele-
ments that generate the output signal. The cantilever must be very light
and stiff to ensure that it has a resonant frequency above the audible range.
See also cartridge.
Cap Short for capacitor.
Capacitance Capacitance is the quantitative measure of the electrical effect
of a capacitor, and its unit is the farad (in honor of Michael Faraday, the
brilliant English investigator who studied electrostatic effects), abbreviated
F. Capacitors are able to store electric charge; the amount of charge is the
voltage across the capacitor times the capacitance in farads. 1 volt would
impress 1 coulomb of charge on a capacitor of 1 farad.
The farad is a very large unit, and the more convenient microfarad (F,
or one millionth of a farad), and picofarad (pF, or one million millionth of
a farad) are commonly used in practice.
Capacitors react to alternating currents in a somewhat complex way, and
this characteristic is called capacitive reactance.
Capacitive Reactance Capacitive reactance is the part of impedance that
is due to capacitance in a circuit. Capacitors have some resistance,
and the current in this resistance obeys ohms law, but the current in
the capacitive portion of the capacitor does not. The current in a capacitor
is equal to the capacitance times the rate of change of voltage across it.
In other words, at a constant voltage, as frequency goes up, the current
in a capacitor also rises. (This is the opposite of what happens in an
inductor.)
Because the capacitor has an impedance that varies with frequency, it
can be (and is) used in equalizers. The value of capacitive reactance is
calculated as follows:

1
XC=


2f
where Xc = Capacitive Reactance,
and F = frequency in hertz

The capacitor passes very rapid changes in current with ease but has
difculty in passing very slowly changing currents and will not pass DC
at all. The voltage across a capacitor lags behind the current by 90 degrees;
i.e., the capacitor has 90 degrees of phase shift.
Capacitor An electronic component that has capacitance. It used to be
called a condenser, but modern usage favors the more accurate term.

56
 Capsule

Some capacitors, however, still use the old terminology, such as condenser
microphone and tuning condenser.
Capacitors accumulate electrical charge according to the law that says
the charge Q equals the voltage V times the capacitance C. See also capac-
itive reactance.
There has been a great deal of raising of eyebrows (an exercise much
practiced in the audio industry) about the relative merits of different types
of capacitors for use in audio circuits. On the surface, it seems that a capac-
itor is a capacitor, and as long as its capacitance and voltage ratings are cor-
rect for an application, it will do its job perfectly; many people adhere to
this view. After all, it only has to separate AC signals from direct voltages
in most applications, or in some cases its variable reactance with frequency
is utilized. However, if one looks more closely, the capacitor is subject to a
host of potential problems. For instance, capacitors that are made by rolling
up strips of conducting foil with the dielectric in between will also have
some inductance, and this can cause problems, mostly at high frequen-
cies. Many such capacitors will actually resonate by themselves, as if they
were wired in an lc circuit. See also q. Some capacitors also exhibit non-
linearity in the sense that their capacitance and leakage current vary a
little with the signal level.
Another type of nonlinearity possessed by some capacitors is nonsym-
metry, meaning their capacitance is different for positive and negative
swings of the signal. These nonlinearities add distortion to the signal,
especially if the offending capacitor is in a feedback loop of an amplifier.
Capstan In a tape recorder, the rotating part that drives the tape past the
heads is the capstan. Usually, the tape is squeezed between the capstan and
a rubber wheel known as the pinch roller or puck.
One would expect the tape speed to be equal to the surface speed of the
capstan, but this is not exactly true. If the capstan has a small diameter, so
that the tape makes a signicant curve where it makes contact, it will actu-
ally move faster than the surface speed due to the belting action of the
tape. Furthermore, thick tapes will move faster than thin ones. The effect
is quite small, but can be measured. This means that a tape with a splice in
it will move faster as the splice goes over the capstan, causing a slight rise
in pitch. This change in speed can be heard in some recordings, especially
if very thin tape is used.
Capstan Drive Motor A motor drives the capstan in a tape recorder. Most
professional-quality tape recorders will have a separate motor to turn the
capstan, traditionally an ac synchronous motor, which means its speed is
determined by the line frequency rather than the voltage. Thus, it will run
at a constant speed regardless of voltage uctuations. Most modern tape
recorders use dc motors, which are servo-controlled, to drive the capstan,
and this facilitates varying the speed. See also vari-speed.
Capstan Idler Synonymous with pinch roller, or puck.
Capsule The business end of a condenser microphone is sometimes

57
Capture Ratio

called a capsule. The capsule is the transducer, and it contains all the
microphones elements except the preamplifier. In some microphones,
the capsule is removable and interchangeable with others having different
directional characteristics.
Some manufacturers use the term cartridge for the same device, but
cartridge is more commonly used for a phonograph transducer.
Capture Ratio The ability of an fm receiver to lock onto, or capture, a
broadcasting station in the presence of an interfering signal near the same
frequency. Capture ratio is expressed in decibels, and the smaller the
number, the better. A capture ratio of 1.5 dB means the receiver will lock
onto a signal only 1.5 dB stronger than the unwanted one.
Carbon Microphone The carbon microphone was one of the earliest types
of microphone to be devised and is still widely used in older telephones.
It consists of a metal diaphragm, one side of which makes contact with some
small carbon granules contained in a small receptacle. When the diaphragm
moves in response to sound pressure, the granules are alternately com-
pressed and relaxed, and their combined resistance to an electric cur-
rent changes. This causes the current to vary in response to the sound
pressure waveform, and this constitutes the signal from the microphone.
This arrangement is not a perfect transducer, and it suffers from notice-
able distortion and noise compared with most modern microphone
types. However, the carbon microphone is very well suited to telephone
usage because its sensitivity increases with sound pressure level. This makes
it more sensitive to a nearby voice and much less sensitive to background
sounds in the room. This peculiar effect has been called a serendipitous
nonlinearity by our friend John Bareham.
The carbon microphone was developed by Thomas Edison while he was
under contract with Alexander Graham Bell, who patented the design. It
was also used in the early days of radio broadcasting, but it has long been
obsolete in this application.

Carbon Microphone

Card A small printed circuit board is often called a card, from its sim-
ilarity in size and shape to a playing card. Cards are often designed so they

58
 Cardioid

plug into a socket, which may be mounted on another circuit board. This
larger board is called a mother board, and the cards are then called
daughter boards. (This may be an example of the essential lack of sex-
ism in the audio world.)
Cardioid Cardioid literally means heart-shaped, and any microphone
that has a heart-shaped polar pattern is called a cardioid microphone.
The cardioid is most sensitive to sounds arriving from the front, is 6 deci-
bels less sensitive to sounds from 90 degrees to the sides, and, theoreti-
cally, is completely insensitive to sounds coming from the rear. In practice,
the directional qualities of a cardioid are not fully realized due to reected
sounds from room walls and ceiling entering the sensitive area of the micro-
phone.
The most important attribute of the cardioid is its ability to discriminate
between direct sounds (coming from the direction in which the microphone
is pointed) and reverberant sounds, which are coming from all other
directions at random. Compared to an omnidirectional microphone,
the cardioid will give a direct-to-reverberant ratio about 4.5 dB higher. That
is, the direct sound will be about 4.5 dB stronger than the reverberant sound,
and this improves the subjective clarity of the received sound.
Another way to state the same fact is that a cardioid may be placed 1.7
times as far from a sound source in a room as an omni and it will give the
same subjective ratio of direct and reverberant sound. This is an advantage
when the microphone should not be seen, such as in motion picture record-
ing, etc. In most real cases, however, an omni microphone can be placed
about two-thirds as far away as a cardioid, and not only will the reverber-

Cardioid Polar Pattern

59
Carrier

ant balance be the same, but the quality will usually be better because omni
microphones are usually smoother and have a wider sound range than car-
dioids. Also, cardioids suffer from proximity effect when they are near
sound sources, while omni microphones do not.
One of the most important uses of cardioid microphones is in sound re-
inforcement, where the directivity allows the system gain to be higher
without generating acoustic feedback. See also supercardioid; hyper-
cardioid.
Carrier The high-frequency signal transmitted by a radio or television sta-
tion after having been amplitude modulated or frequency modu-
lated by the audio or video signal. Theoretically, the carrier could be any
frequency, and in broadcasting the carrier frequencies are chosen based on
considerations of ease of generating high power output, required antenna
size, etc. In the U.S., the am radio band has carriers between 550 and 1650
kilohertz, and the fm and television bands use frequencies between 54 and
806 megahertz.
The carrier itself does not transmit any information; all the intelligence
is in the modulation sidebands, which are in a band of frequencies on either
side of the carrier frequency. The carrier is used by the receiver to aid in
the demodulation, or detection, of the signal.
Some signals, such as fm stereo, involve more than one carrier to en-
code the information. The lower frequency carrier is called a subcarrier, and
it is mixed with certain parts of the audio signal and used to modulate the
main carrier. In the receiver, the subcarrier is recovered by demodulation
of the main carrier, and it is then demodulated to recover its signal.
Cart Machine Broadcasters slang for tape cartridge machine, which is
a playback machine that uses endless loops of tape in plastic cartridges.
Commercials and announcements are recorded on carts, and some of the
playback machines can handle a dozen or so of them, something like a record
changer. Now obsolete, they were replaced by minidiscs and other digi-
tal storage systems.
Cartridge, Phono Cartridge The cartridge is the transducer that converts
the motion of the phonograph stylus into an electrical voltage. This volt-
age is the output signal of the cartridge.
There are several kinds of phono cartridge, but the moving magnet
and moving coil types are the most common. In the moving magnet car-
tridge, a very small magnet is made to vibrate in response to the modu-
lation of the record groove. This motion induces a current in two nearby
coils of wire, one for each stereo channel. The moving magnet cartridge is
characterized by relatively high output voltage and relatively low cost,
although some good-quality examples are quite costly. Almost all such units
are made to operate correctly when connected to a 47,000-ohm load. Vir-
tually all phono preamps thus have an input impedance of 47,000
ohms.
The moving coil cartridge contains two tiny coils of wire, which are made
to vibrate in accordance with the record groove modulation. They are in

60
 Cathode Bias

the magnetic eld of a small stationary permanent magnet; thus a signal

current is induced in them. Moving coil cartridges are characterized by rel-
atively high prices, very low output voltages, and very low output imped-
ance, in some cases as low as a few ohms. This means they require either
a transformer or a pre-preamp between them and the phono preamp.
As usual in audio, there is some controversy over which is the best type
of cartridge, and it is difcult to present objective data that clearly favor
either one. Some comments may be appropriate, however:
The moving coil types have very low mass, and thus very high reso-
nance frequencies, sometimes as high as 60 kilohertz or so. Even
though this is above the audible range, it can cause subtle distortion of
the sound. This resonance is very difcult to damp, and as a result, the out-
put of the cartridge rises very much at high frequencies. This means the
preamp (and pre-preamp) are subjected to high levels of ultrasonic noise
from the record surface, and this can cause intermodulation distortion
in the audible range. This means that an expensive moving coil cartridge
used with a mediocre-quality preamp usually sounds worse than an inex-
pensive moving magnet with the same preamp. Also, the very low output
voltages can lead to signal-to-noise ratio problems unless the pre-
preamp is very quiet. On the other hand, they tend to be more linear, and
the harmonic distortion of moving coils is lower than that of moving
magnets.
The moving magnet cartridges are somewhat easier to use than the mov-
ing coils because of their higher output and the wide availability of phono
preamps with the correct input characteristics. They also generally have
user-replaceable stylus assemblies, whereas moving coils usually do not.
CAS Cinema Audio Society, an honorary group of people in the motion
picture sound industry.
Cathode The negatively charged terminal of a component such as a diode
or battery is called the cathode. current, which is dened as a ow of
charge, was thought by Benjamin Franklin to be in the direction from
anode to cathode. Franklin, who established many of our conventions
regarding electricity, thought the charged particles which constitute elec-
tric current were positively charged, but today we know that they are neg-
atively charged (electrons). So the actual current is in the direction from
cathode to anode, although many old-timers (the writers included) still cling
to the old convention of positive charge. As long as one is consistent, there
is no problem with this.
Cathode Bias A method of applying bias to a vacuum tube by placing a
resistor between the cathode and ground. The current in the resistor
forces the cathode to be at a higher potential than ground, and the grid,
if referenced to ground, will be at a negative potential with respect to the
cathode. Cathode bias provides a certain amount of negative feedback
around the tube, along with some reduction in gain. If the cathode resis-
tor is shunted by a large capacitor, called a bypass capacitor, the AC gain
is restored and the feedback reduced.

61
Cathode Follower

Cathode Follower If the plate of a vacuum tube is connected to a posi-

tive voltage supply without a series resistor, and if the cathode has
a series resistor to ground, the cathode will vary in its voltage in the
same manner as the grid. This arrangement is called a cathode follower.
It is characterized by a very high input impedance, a very low output
impedance, a voltage gain of less than 1, and a very high current gain.
It also has very low distortion, and is used as an impedance matching
device for driving long lines, such as between preamplifiers and power
amplifiers.
With the general adoption of solid-state equipment, the cathode fol-
lower is essentially obsolete. Similar techniques are used in solid-state cir-
cuits to achieve low impedances.
Cathode Ray Tube, or CRT Traditionally the active display device in an
oscilloscope, the picture tube in a television set, and the display for
a computer, often called simply CRT. Newer technologies such as liquid
crystal and plasma displays are replacing the CRT.
The CRT was invented by Dr. Vladimir K. Zworykin of RCA laborato-
ries in the 1930s. An evacuated glass tube contains an electron gun, which
emits a stream of electrons, and a uorescent screen on which the electrons
impinge. On their way from the gun to the screen, the electrons are accel-
erated by a high-voltage electric eld, and they are deected in their path
by voltages applied to deection plates adjacent to the gun in the case of
the oscilloscope, or by magnetic elds from a deection yoke of wire coils
placed around the neck of the TV picture tube. Electrons have very little
mass and are easy to deect even though they are moving very fast. This
ability to rapidly deect the beam makes the CRT useful for looking at the
instantaneous voltage of a changing waveform, as in the oscilloscope.
When the electrons hit the screen, they are rapidly decelerated, and in so
doing, they emit a certain amount of X-ray radiation. The amount depends

Cathode Ray Tube

62
 CD

on the accelerating voltage used in the CRT, so large CRT computer screens
and TV screens make X-rays that would be harmful to computer users and
TV watchers. To prevent X-ray exposure, the glass in the business end of
the CRT contains a fairly large quantity of lead, which is a good X-ray
absorber. The larger the screen, the more lead required, and this causes large-
screen TVs and large CRT computer screens to be inordinately heavy for
their size.
Cats Whisker See crystal set.
CAV, Constant Angular Velocity When a phonograph or laserdisc main-
tains a constant RPM, the linear velocity at the pickup decreases as the
diameter gets smaller. The problem with a phonograph disc is that since the
actual speed of the disc surface past the stylus is faster at the outer edge of
the disc and much slower at the inside of the disc, the frequency response
declines and distortion increases as the record is played. In the case of
the laserdisc (Laser Video Disc, LVD), this is not a problem, and CAV
refers to an original formatting of the disc that places each video eld on
exactly 12 of the diameter no matter where on the disc. This design sim-
plied slow motion and stop frame effects on early players with electro-
mechanical control. It also limited playing time to 30 minutes per side, so
a constant linear velocity laserdisc was introduced with an hour of play-
ing time per side. The introduction of digital video buffer memory allowed
decent still frame and slow motion effects on CLV discs. See also clv.
C Battery In the early days of radio, receiving sets were all battery pow-
ered if they were not crystal sets. The low-voltage, high-current A bat-
tery supplied power to the tube heaters, and the high-voltage, low-current
B battery supplied the voltage for the tube plates. An addition in the mid-
1920s was the C battery, which supplied a low voltage at very low current
to provide bias to the tube grids. The bias reduced the distortion consid-
erably, but most set owners thought the major benet of the C battery was
to reduce plate current, allowing the B battery to last much longer.
CCD See bucket brigade.
CCIR CCIR was an international organization concerned with the setting
of standards and practices for audio and broadcast equipment. CCIR stood
for Comit Consultatif International Radio and was analogous to the
National Association of Broadcasters (nab) in the United States. The
CCIR joined the itu in 1992 and became ITU-R, the division dealing with
radio communications and standards. CCIR standards are generally now
iec standards.
Many European tape recorders use CCIR-designed pre-emphasis and
de-emphasis curves; tapes made on them are incompatible with ones
recorded in the United States. It is interesting to note that a good case can
be made for the idea that U.S.-made tapes are more suited to CCIR equal-
ization, while European tapes are better optimized by the nab curves.
This is an example of the Law of Probabilistic Adversity.
CD, Compact Disc. The original 120 mm digital optical disc format
invented by Sony and Philips in 1982. It is fully specied in the red book,

63
CD-DA

and all other CD formats are derived from it. It was originally known as
cd-da for Compact Disc-Digital Audio.
CD-DA, Compact Disc Digital Audio The original name of what we
know today as simply the cd, or compact disc.
CD Horn EQ See constant directivity horn.
CD+G A special compact disc format that allows simple graphics and
text to be stored in the subcodes of an audio CD. A special player is needed
to display the data, such as a special karaoke player.
CD-I Acronym for Compact DiscInteractive. A digital disc system for stor-
ing video, audio, and graphics information in such a way that the user can
control the presentation of the various media. The CD-I standard was
released by Philips and Sony in the Green Book in 1987.
CD-MIDI A cd format that stores midi information in the subcode area.
CD-R, Compact Disc Recordable A recordable optical disc with the same
physical dimensions as the normal compact disc. CD recorders are com-
mon in computers, and many stand-alone units are available for home use.
CD-Rs can be used for recording music or recording computer data as in
cd-roms. CD-R audio recorders for home use usually require a special
music CD-R that has a special code on the blank. Extra fees for music
CD-R blanks go to a music industry fund.
CD-ROM Acronym for Compact DiscRead Only Memory. The compact
disc medium modied to store computer information rather than audio.
This is analogous to a magnetic hard drive in a computer, except the stor-
age is by optical means and once created, the CD-ROM can only be read
from and not written to. The CD-ROM must be used in conjunction with
a computer to retrieve the data.
CD-ROM XA A CD specication dened in 1988 by Philips, Sony, and
Microsoft as an extension of the Yellow Book standard. The XA stands for
Extended Architecture. The format shares many traits of the CD-I.
CDS See cinema digital sound.
CD-Single A small compact disc that can record 20 minutes of stereo
music. It is 80 mm in diameter. The record industry will sometimes make
a standard size CD with one or two songs on it and call it a CD single.
CD-V Acronym for Compact Disc-Video. The compact disc medium
modied to record video signals as well as digital stereo audio signals. The
video information is recorded in analog form rather than digital. The sys-
tem was compatible with the laserdisc system and is similarly obsolete.
Not to be confused with Video CD, which is a digital audio/video format
compatible with many DVD players.
CEDAR, Computer Enhanced Digital Audio Restoration CEDAR Audio
began as a research project between the British National Sound Archive and
Cambridge University in the mid-1980s for restoring the sound of old media
such as discs, cylinders, and lm. Now a commercial enterprise, they man-
ufacture CEDAR noise reduction products and offer denoising and
restoration services on existing recordings.
Cell A single unit of an electrical battery is a cell, even though cells are often

64
 Characteristic Curves

called batteries, as in ashlight battery. Various types of cells produce

voltages between 1 and 3 V depending on the chemicals involved; most
are around 1.5 V.
Centering Device The exible mounting that allows the voice coil of a
cone-type loudspeaker to move in and out of the magnetic gap but pre-
vents it from striking the sides of the gap. See also spider.
Center Tap The center tap is an electrical connection to the center of a
transformer winding. The voltage at the center tap will be midway
between the voltages at the two ends of the winding, and this is often the
ground potential in an audio device. Frequently, center taps of audio
transformers are connected to the chassis of the device. This has the effect
of reducing the 60-hertz, common-mode voltages in the circuit and
thus reducing the hum and noise.
In the case of the phantom power of microphones, the center tap
of the input transformer primary winding is usually used as the point where
the phantom voltage is applied.
Cents Besides 1/100th of a dollar or a Euro, musical cents are frequency
increments of 1/100th of a semitone. Some audio devices with variable
speed can be adjusted in cents. One cent is a frequency ratio of the 1200th
root of 2, or about 0.06 percent.
CF, Compact Flash A format of semiconductor memory package. Often
used as storage for digital cameras, higher capacity has led to use in digi-
tal audio recorders.
Channel An independently processed or recorded signal is called a chan-
nel. A tape recorder able to record two simultaneous signals is a two-channel
machine, and so on.
Channel Bit Rate The actual bits being read from a digital tape or com-
pact disc (CD) are much greater in number than would theoretically be
required to encode the audio. This is because error correction and
synchronization bits, etc., are added to the audio code.
In the CD, the audio bit rate is 1.41 million bits per second. The chan-
nel bit rate is actually three times as high, or 4.32 million bits per second.
Channel Separation Channel separation refers to the crosstalk between
the channels of a multichannel system. It is the inverse of inter-channel
crosstalk, measured in decibels. A small amount of crosstalk is equiva-
lent to a large channel separation.
Channel Strip On an inline mixing console, each input and its controls, such
as sends, EQs, and fader, is arranged vertically, and is known as a channel
strip.
Characteristic Curves It is convenient to present the electrical characteris-
tics of a tube or transistor in terms of graphs or curves that relate the
various voltages and currents at the inputs and outputs of the devices.
The most important curves for a tube are probably the plate characteris-
tics, which are a family of curves made by plotting plate current versus plate
voltage at various values of grid voltage. Similarly, the collector charac-
teristics of a bipolar transistor are a family of curves that plot collector cur-

65
Charge

rent versus collector voltage at various values of base current. By referring

to these curves, an engineer can predict the performance of different cir-
cuit designs.
Charge Charge is a measure of the quantity of electricity, and its unit is the
coulomb, named after the famous French physicist Charles Augustin de
Coulomb. In an electrical circuit, charge consists of negative charges, or
electrons. A positive charge can be thought of as simply an absence, or
deciency, of electrons. The electron is the smallest possible quantity or unit
of charge. Charge is what is moving in an electric current; the electric cur-
rent exists, but does not itself ow. (By analogy, the current in a river is a
ow of water and the current exists but does not ow.) Use of the phrase
ow of current should be discouraged.
Positive charges exist in the form of protons, which are one of the fun-
damental particles from which all matter is made, but they do not take an
active part in electrical circuits.
The symbol for charge is Q.
Charge-Coupled Device, or CCD Often referred to as charge transfer
device. See bucket brigade.
Chase-Lock A type of controller for a video or audio recorder that will lis-
ten to the smpte time code signal from another device and will adjust
its own speed to nd the correct time and then will lock into synchroniza-
tion with the external time code.
Chassis (Pronounced chassey.) The metal mechanical structure that
houses electronic circuits and shields them from external magnetic and
electric elds is called the chassis. It is always connected to ground and
to the shields of audio cables. Many times one of the two wires carrying
the signal is also connected to the chassis. See also Appendix 6.
Chebyshev Filter A multiple-pole, low-pass filter designed according
to Chebyshev polynomials is a Chebyshev lter. It has a steeper cutoff
than does the butterworth filter, but it has irregularities in its fre-
quency response, called ripple.
The Chebyshev is called an overcompensated lter in that it has peaks
in its passband. These peaks are collectively called ripple, and a ripple of
1 decibel means the response remains within 1 dB of at in the passband.
The lter trades ripple depth for cutoff steepness; the steeper the cutoff,
the deeper the ripple. The Butterworth lter can be considered a ripple-
free Chebyshev lter. Chebyshev lters can be used as anti-aliasing fil-
ters in digital audio devices. Pafnuty Lvovich Chebyshev, also spelled
Tschebyscheff and Tchebysheff, (18211894) was a Russian mathematician,
best remembered for his work on the theory of prime numbers.6
Chip During the mastering of an acetate, the cutting stylus removes
a hairlike lament of the acetate material as it cuts the groove. This is called
the chip and, in most mastering setups, is removed by a vacuum in a small

6. Thanks to Dennis Bohn, Rane Corp.

66
 Cinema Digital Sound

metal pipe beside the cutterhead. The cutting stylus is heated with a small
coil of resistance wire, and because the chip is ammable care must be taken
that it does not catch re on contact with the heater. In Britain, the chip is
called the swarf.
The term chip is also used to mean an integrated circuit, probably
because a small chip of silicon is the basis of the device.
Choke, Choke Coil Obsolete term for inductor. The term probably
stems from the idea that an inductor chokes alternating currents
in proportion to their frequency. The most common usage of the word
has been and is in connection with power supply inductors, which are used
to reduce ripple. Otherwise, inductors are generally so called.
Chorus An electronic music effect that modies the sound of a single instru-
ment to simulate a large group of the same instruments, for example, a vocal
chorus or a string section. The subjective effect of a real chorus is caused
by the fact that the many sound sources being mixed together all have
slightly different frequencies and also do not have precisely steady fre-
quencies. The mixture becomes extremely complex as the relative phases
of the signals cause partial cancellation and reinforcement over a broad fre-
quency spectrum.
The synthetic chorus effect was rst attained by subjecting the input
sound to a series of very short time delays and mixing the delayed sounds.
The delays were then randomly varied, or modulated, to increase the
uncertainty of the combined pitch. This could be called the time domain
chorus synthesis and can be quite expensive if enough delay times are used
to ensure a satisfactory result. Another type of chorus device operates in
the frequency domain and is somewhat simpler and at the same time more
convincing. The signal is split into many frequency bands by a series of band-
pass lters, and each band is randomly varied in phase and amplitude,
after which they are recombined.
Chromium Dioxide Tape Chromium dioxide can be used as a magnetic
medium in recording tapes, in addition to the usual oxides of iron. Chrome
tape, as it is sometimes called, has better high-frequency performance and
less noise than iron oxide tape. It is especially useful with very short recorded
wavelengths, i.e., when the tape speed is low, as in cassettes.
It requires more bias current than conventional tapes, and a different
pre-emphasis curve; therefore, cassette decks designed for chrome tape
will have a switch to change these parameters, or the machine may do the
switching automatically via standardized slots in the cassette shell.
Some specially formulated iron oxide tapes are similar in their charac-
teristics to chrome tape and are called chrome compatible. Although
DuPont owns the patent for chromium dioxide and the tradename, crolyn,
the iec has a standard for chrome and chrome-compatible tapes and calls
them Type II cassette tapes. They contain cobalt-doped gamma ferric
oxide, which has similar magnetic characteristics, but avoid using any
chromium dioxide.
Cinema Digital Sound A system of digitally recording motion picture

67
CinemaScope

sound onto the lm via a laser beam, that reportedly combines the dynamic
and frequency ranges and low distortion of the compact disc on six
discrete channels. It was a joint project between the Eastman Kodak com-
pany and Optical Radiation Corporation of Azusa, California. Five chan-
nels encompass the full audio bandwidth, and the sixth is designated a
subwoofer channel and contains only the lowest frequencies. The CDS-
encoded lm was capable of being shown with conventional stereo opti-
cal sound but required a special sound system to reproduce the six
channels digitally. The 1990 feature Dick Tracy in 70-mm format was the
rst production to use CDS. The system succumbed to other digital cin-
ema sound systems.
CinemaScope The trademark of the rst true stereophonic motion pic-
ture sound system that had the sound tracks on the same lm with the pic-
ture. The system was developed and introduced by 20th Century Fox, and
the rst lm to use the process was The Robe, in 1953. The process used a
wide screen with a 2.35 to 1 aspect ratio. The image was compressed side-
ways by a factor of 2 to 1 onto the lm by a special cylindrical lens and was
then expanded to full width by a similar anamorphic lens on the pro-
jector. (This system could be thought of as an optical compander!)
The rst CinemaScope movies were produced with full stereophonic
three-channel sound, with a fourth channel for ambience effects reproduced
from loudspeakers in the rear and sides of the theater. This proved to be
very expensive in practice, and later lms had stereo music but dialogue
and sound effects panpotted to the various channels from a monaural
source. The early CinemaScope movies were very painstakingly made and
had a three-dimensional audio realism that was quite remarkable in 1953,
and is seldom exceeded today!
There were four independent magnetic sound tracks on the lm, two
on each side of the rows of sprocket holes. To accomplish this, Fox reduced
the width of the holes to allow enough space for the mag tracks. This meant
that special sprockets had to be installed in the projectors (sometimes called
Fox-hole sprockets), but these new sprockets were still usable with stan-
dard non-CinemaScope lms, and virtually all projectors have them today.
Cinerama A motion picture format introduced in 1952 that used three sep-
arate synchronized 35-mm projectors simultaneously to produce three
images that fused on the screen into one very wide panorama. The sound
system consisted of ve channels of amplifiers and loudspeakers (later
seven channels) and ve or seven magnetic sound tracks on a separate 35-
mm magnetic lm. The synchronization of the projectors and sound repro-
ducer was quite complex and entailed precise threading of the machines.
The sound system was capable of very high quality results for the time,
but the problems in producing ve- or seven-channel stereophonic sound
tracks were formidable and turned out to be prohibitively expensive, as
was the production of three simultaneous 35-mm picture lms.
Circuit An arrangement of electronic components for performing a par-
ticular task is called a circuit. Literally, a circuit is a path through which the

68
 Clipping

signal current exists. The circuit path is always in the form of a loop,
although it may be very convoluted. Most audio circuits are formed on
etched circuit boards, where the wires consist of a copper plating on the
berglass board.
Circuit Board See printed circuit board.
Circumaural A circumaural headset has a large cushion that surrounds the
ear and makes contact with the head. Such headsets exclude external noise,
unlike supraaural designs.
Clapboard See slate.
Class A, Class AB, Class B, Class C, Class D, Class G, Class H See power
amplifier.
Click Track A click track records a series of clicks, like a metronome
sound, on one channel of a multitrack tape recorder. The click track is used
to synchronize the recording of subsequent tracks by playing it back via
headphones to the musicians while they are playing for the added tracks.
A midi sequencer is sometimes used to generate clicks. This has the advan-
tage of easy on the spot editing and control over such things as varying
tempo.
Clipping If a signal waveform is passed through an amplifier or other
device that cannot accommodate its maximum voltage or current

Clipping

69
Clock

requirements, the waveform is sometimes said to be clipped, because it looks

like it has had its peaks clipped by a pair of scissors. A clipped waveform
contains a great deal of harmonic distortion and sounds very rough
and harsh. Clipping is what typically happens when an audio amplier
output is overloaded or its input stage is overdriven. The clipping point
of an amplier is dened as the maximum sine-wave signal level which,
when viewed on an oscilloscope, shows no signs of at-topping to the
trained observer.
It is interesting that clipping does not usually reduce the intelligibility
of speech signals; in fact, it has been shown that clipped speech is easier to
understand than normal speech in noisy environments. Probably the rea-
son for this is that clipped speech has much greater high-frequency con-
tent due to the added distortion, and this is easier to hear under noisy
conditions. See also hard clipping; soft clipping.
Clock, Clock Frequency A high-frequency signal, usually generated by
a crystal-controlled oscillator, which is used as a master timer to con-
trol the events in a digital system. For instance, the sampling frequency,
even though it is lower in frequency, is determined by the master clock
frequency.
It is important that the clock is accurately controlled in frequency. If a
signal is sampled at one frequency during recording and at a slightly dif-
ferent frequency at playback, the resulting musical pitch of the signal will
be different. Also, the clock must have a very high degree of stability in fre-
quency to reduce sampling errors due to jitter.
Closed Circuit Television A television system where the signal is routed
through coaxial cables from the camera or cameras to the monitor(s) rather
than being broadcast.
Closed Loop An amplifier or amplier stage is said to be operating in
the closed-loop mode if it has feedback around it. See also open loop.
Closed-Loop Tape Drive A drive system for the tape in a magnetic tape
recorder in which the tape is pinched between a capstan and a puck on
each side of the tape heads. This double-capstan arrangement is capable
of excellent control of the tape motion and can result in very low values of
flutter and wow. Sometimes only a single capstan is used, and the tape
is looped around and is forced against the opposite side of the capstan by
a second puck.
Close Miking The placement of a recording or broadcast microphone
extremely close to a vocalist or a musical instrument for the purpose of elim-
inating as much extraneous sound as possible is known as close miking.
Close miking rarely provides a natural-sounding pickup of the instrument
or voice, but accuracy may not be desired in many instances. See also jolly
green giant effect.
CLV Acronym for Constant Linear Velocity. The compact disc is an exam-
ple of a medium whose rotational speed (rpm) is varied depending on what
part of the recording is being played. It runs fastest at the inner part, and
gradually slows down towards the outer part as the disc is played so as to

70
 Code

achieve a constant translational velocity to the laser pickup beam. The pur-
pose of CLV is to conserve disc space. The laser videodisc can also utilize
CLV as well as cav.
CMRR Acronym for Common-Mode Rejection Ratio. See common mode;
Appendix 6.
Coaster A popular term for a damaged recordable audio CD, named after
the round object you rest your drinking glass on so that it doesnt mark the
table.
Coaxial Loudspeaker A coaxial loudspeaker is a two-way system (woofer
and tweeter) combined into one unit such that their centers are in line.
Coaxial literally means coincident axes. The tweeter in a coaxial speaker
is usually a horn-loaded unit with the throat of the horn formed by bor-
ing out the center of the woofer magnet. The advantage of the coaxial
arrangement is that at the crossover frequency, the phase irregularities are
minimized.
One manufacturer some time ago introduced a three-way loudspeaker
in one unit and called it triaxial, literally, with three axes, but this did
not describe the loudspeaker in questionan example of the misuse of
language.
Cockpit Trouble Often problems with recording or reproducing equipment
are not actually in the equipment but are caused by an inept user. This is
referred to by some professionals as cockpit trouble, sometimes also called
a short between the earphones.
Cocktail Party Effect A psychoacoustic effect that allows us to localize
the sources of sounds around us. It arises because we have binaural hear-
ing. When a person is in a sound eld where sounds are arriving from many
directions at once, it is possible to direct ones attention in a particular direc-
tion and ignore sounds from all other directions. For instance, at a crowded
cocktail party, one can concentrate on a talker at a certain location in the
room to the exclusion of all the others; the other voices merge into amor-
phous background noise. If, on the other hand, a single-channel (monau-
ral) recording is made at the same party, it is not possible to isolate
individual voices in the playback. A binaural recording and playback
restores the ability.
The effect contributes greatly to the sonic clarity one experiences in lis-
tening to live music in an auditorium and explains why monaural record-
ings must be recorded to exclude much of the reverberation of the room to
help increase the clarity of the playback. Stereophonic recording restores
part of the localization ability, but only a true binaural recording is com-
pletely effective.
The cocktail party effect was investigated and so named by Irwin Pol-
lack and J. M. Pickett in 1957. They reported it in an article in the Journal of
the Acoustical Society of America in February 1958.
Code In digital audio devices, the signal exists in the form of binary
words, and this digital information is called code. The analog audio
signal is sometimes said to be encoded. The set of instructions that make

71
Codec

up a computer program is usually in cryptic form, and this is also often

called code.
Codec Abbreviation for Code-Decode. In the audio world, a codec is a soft-
ware application that encodes a digitized audio signal le into a different
format, often for the purpose of reducing its size. This process is also called
data compression, or simply compression. The encoded le must be
decoded to recover the original audio le, a process one might expect to be
called decompression, but we have not heard this term used in practice.
An example of a familiar codec is PKZip.exe that encodes any computer
le into a Zip le that is smaller than the source le. The le can then be
stored in its compressed form and then unzipped to recover the le in
its original form. Such a codec is called lossless because no information
is lost in the coding-decoding process. Another example of a lossless codec
is used by analog-to-digital converters (ADCs) and digital-to-
analog converters (DACs), used to encode and decode the audio sig-
nal to be recorded onto a compact disc. Since there are usually errors in
writing and reading coded audio, some people think it should not be called
lossless. Elaborate error correction and error concealment schemes
are used in CD players for this reason. Some codecs do not recover the orig-
inal signal exactly, and they are called lossy. An example of a lossy codec
is Dolby AC-3, which uses psychoacoustic principles to remove parts of
the signal that cannot be heard by the human ear due to masking by louder
sounds, a technique called perceptual coding. This allows the coded les
to be more compressed, resulting in much smaller coded les than are pos-
sible with lossless coding. Other examples of lossy codecs are the ATRAC
system on MiniDiscs, audio MP3, and video MPEG2.
Codecs have been around for a long timeSamuel F. B. Morses so-
called Morse Code for telegraphy is a codec, as are all secret codes for
communication.
Coercivity Coercivity is a measure of a magnetic materials resistance to
being demagnetized. Coercivity is an important specication of mag-
netic tape. Tapes with low coercivity are easily demagnetized, meaning
they can become erased if exposed to relatively small magnetic elds. Units
of coercivity are oersteds.
Cogging Uneven torque as a function of rotation produced by an electric
motor because of the nite number of poles in the motor. Cogging in direct-
drive turntable motors or tape recorder spooling motors can cause flut-
ter unless the motors are carefully designed to reduce the effect.
Coherer A type of detector invented in 1892 by the Frenchman Edouard
Branly after a discovery by D. E. Hughes of England and used in very early
radio receivers that received signals from spark gap transmitters. The
coherer consisted of a small insulating cylinder containing zinc and silver
lings with a contact wire on each end. The rf signal received by the
antenna, or aerial, caused the particles to cohere, or stick together, and
to detect the signal. The Russian experimenter Popoff added a small ham-

72
 Coloration

mer, actuated like the hammer on an electric doorbell, which rattled the
particles around, breaking them apart and readying them for the next spark
reception. It was called, obviously enough, the decoherer.
Coil Literally, a coil of wire. Coils abound in audio devices; most loud-
speakers have voice coils as active elements; transformer primaries
and secondaries are coils; a phono cartridge contains coils in which
the signals are induced.
A coil has electrical inductance, the amount of which is determined,
among other things, by the number of turns of wire in the coil. The induc-
tance of a coil is not linearly related to the number of turns, however, but
is proportional to the square of the number of turns. Thus, two coils whose
turns differ by a factor of 2 will have inductances differing by a factor of 4.
Coil Former The cylindrical structure on which the voice coil of a
dynamic loudspeaker is wound is called a coil former or simply a for-
mer. The design of the coil former has a great deal to do with the perfor-
mance of the loudspeaker. In olden days, it was made of paper impregnated
with varnish or shellac, but such a voice coil is not very tolerant of heating
at high signal levels. A former made of metal, such as aluminum, is much
better from a temperature tolerance standpoint and also conducts the heat
away from the coil, but it is also an electrical conductor, so it behaves as a
shorted turn, or the secondary of a transformer with the voice coil as
primary. Its motion in the magnetic eld would also be greatly impeded
by eddy currents induced in it. To reduce these effects, a metal former
must be slotted so it does not make a complete turn.
In some loudspeakers, especially tweeters, the voice coil is simply
glued together without a former. This saves space in the magnetic gap, elim-
inates eddy currents in the former, and increases the sensitivity or efciency
of the speaker.
Coincident Microphones In the recording of intensity stereo, where
all the directional cues are caused by differences in loudness of the sig-
nals in the two channels, the two microphones must be at the same place
to avoid response differences between the two signals. This is impossi-
ble in practice, but the microphones are placed as close together as possi-
ble and they are called coincident microphones. They must be directional
microphones if each is to pick up its side of the stereo image, and they
are usually cardioids or bidirectional microphones.
In a special case of intensity stereo one of the coincident microphones
is a bidirectional (figure 8), and the other one is an omnidirectional,
or cardioid. By combining the outputs of the two microphones in vari-
ous ratios, the apparent width of the stereo image can be changed. This is
called ms stereo standing for mid-side stereo: the bidirectional microphone
is the side and the cardioid the mid.
Coloration Coloration is a term for subtle distortion that results in a
change in the timbre of a sound, without that sound being noticeably
distorted.

73
Colortek

If the same signal is routed to various loudspeaker systems, the

resulting sounds will be easily discerned as different, although it may be
difcult to determine which is the most accurate reproduction of the orig-
inal. Each speaker is said to possess its own coloration.
Colortek A four-channel discrete optical motion picture sound system
invented by John Mosely and introduced in 1977 with the movie A Bridge
Too Far. Its full name is the Colortek Optical Stereophonic Sound Film Sys-
tem, and was the successor to Quintaphonic Sound, which was an
obscure two-track optical lm sound system. The innovative design of the
system employs, as a sound reader, a charge-coupled-device scanner (CCD
scanner) and applies video technology to the readout of soundtracks. It scans
the four soundtracks transversally at 10 MHz, resulting in a CCD frame
rate of 40 kHz, as the tracks pass longitudinally at the normal speed of 24
lm frames/s (457.2 mm/s) through the sound reader. In 1978, Colortek
was modied and reborn as Kintek Stereophonic Sound, only to be rein-
carnated in 1980 with further modications as Cinesonics. One feature
that seems to have been retained throughout this metamorphosis is the use
of the dbx compander system. See also Comtrak.
Column Loudspeaker A column loudspeaker, also called a line array, is
a loudspeaker system with the drivers arranged in a vertical column,
usually from three to six feet long, and six to eight inches wide. This
conguration causes the system to have a narrow vertical coverage angle;
the longer the column, the narrower the angle. The horizontal coverage is
the same as that of the individual drivers. The value of this arrangement
is that in an auditorium with a sound reinforcement system it is important
that the loudspeakers cover only the area where the audience is, for any
other coverage simply adds to the reverberant sound, decreasing clar-
ity and speech intelligibility.
The column effect is wavelength dependent, and this means a given
column will become more and more directional as the frequency goes
up. For this reason, some manufacturers make two-way systems where the
highest frequencies are fed to a shorter column and the mid and low fre-
quencies to a longer column, with a crossover network dividing the
signal between them.
The column speaker was introduced in the 1950s and since that time has
almost revolutionized sound reinforcement, especially in the case of rela-
tively small systems. It is important to note that the column speaker must
be mounted vertically to achieve its benets. A natural position for one is
above the proscenium in an auditorium.
Comb Filter A comb lter is a filter that has a series of very deep
notches, or dips, in its frequency response. The spacing of the notches
along the frequency axis is at multiples of the lowest frequency notch,
so they look evenly spaced along a graph plotted on a linear frequency scale.
On the more common logarithmic frequency scale, the notches become
closer together on the paper as frequency increases.
A comb lter is produced when a signal is time-delayed and added

74
 Compact Cassette

to itself. Frequencies where the time delay is one-half the period and mul-
tiples of these frequencies are canceled when the signals are combined
because they have opposite polarity. If the signals are of equal strength,
the cancellation is perfect and the notches are innitely deep on a decibel
scale. See also flanging.

Comb Filter

Combination 14-inch-XLR Jack A clever dual female connector that incor-

porates two jacks in one spacea 14-inch stereo phone jack and a 3-pin xlr
jack. The 14-inch jack hole is nestled between the 3 holes for the XLR pins.
Also known as a TRS/XLR jack.
Common Mode Common mode refers to equal voltages induced in the
two wires of a signal-carrying pair.
In a balanced audio circuit, the signal voltages are of opposite
polarity in the two signal wires. Any voltage that appears with the same
polarity on each wire is called a common-mode voltage. Usually noise,
such as 60-hertz hum, is induced in audio cables equally and in the same
direction, and so is a common-mode voltage. If the signal is connected to
a differential amplifier input, the common-mode voltages will can-
cel, while the signal voltages, being of opposite polarity on each input ter-
minal, will add together. This is the reason why balanced lines are less prone
to induced noise from external inuences.
Common-Mode Rejection Ratio See common mode; Appendix 6.
Compact Cassette The ofcial name for the ubiquitous audio cassette. The
name and the detailed specications for the compact cassette recording and
playback system were established and patented by Philips many years ago,
and they will grant licenses to make cassette equipment only if the
specications are precisely followed, ensuring compatibility among equip-
ment and tapes from various manufacturers.

75
Compact Disc

Compact Disc The compact disc system, also sometimes called the DAD,
for digital audio disc, or CD-DA, for compact disc-digital audio, was the
rst truly digital audio system to be widely available for home use. It
is an outgrowth of the laserdisc technology, although the laserdisc is not
a digital system (It would later gain digital audio capabilities).
The system consists of a fast-rotating, variable-speed, aluminized disc
with a very ne spiral pattern of extremely small pits in its surface, which
are detected by a laser beam. The pit depth is one-quarter wavelength
of the laser light. The spiral is oriented as it is played from the inside out.
The disc is scanned by a low-power laser focused onto its surface and a
photodetector, which receives the reected light. The existence of a pit will
cause the reected laser light beam to be delayed by one-half wavelength
(one-quarter wavelength each way), causing cancellation. (For this system
to operate, the distance between the laser and the disc surface must be very
closely controlled. This is done by a special servo system that continuously
adjusts the position of the laser lens.) The pattern of the pits in the disc cor-
responds to the digital bits of the encoded audio signal, which consists of
the audio signal and a series of extra bits that are used in the decoding
process for error correction. In playback, the coded series of bits is
decoded with a complex program that checks the recovered digital signal
from the disc for errors caused by mistracking, scratches, or foreign mate-
rial on the disc surface and corrects the errors if they are not too severe. The
error-corrected digital signal is then passed through a digital-to-ana-
log converter to recover the original audio signal waveform.
All optical storage systems for digital information use a laser beam to
read the data. The beam is generally a small gallium arsenide semicon-
ductor laser. To read the information, the laser beam is focused on the
spiral track consisting of microscopic pits. The laser light is reected in
different ways by the pits and the surface land between the pits. On strik-
ing a pit, the light is diffracted while light striking the land is reected
back to the laser and is detected by a photo detector. To do its job, the laser
must be focused down to a tiny circular spot. The pits are 0.6 microns wide
and 0.12 microns deep. (A micron is one millionth of a meter!) The length
of the pits and lands can vary between 0.833 microns and 3.56 microns,
and the track density on the disc is about 16,000 tracks per inch. The length
of the track on a compact disk is about 4 miles. To illustrate what these
gures mean, if the diameter of a CD were enlarged to 120 meters (about
the size of a baseball diamond), the tracks would still be less than 0.5 milli-
meters wide!
Compactron A multiple-purpose vacuum tube introduced by the General
Electric company in 1960. Some types had as many as three separate tubes
in one small glass envelope. The advent of the integrated circuit and
the increasing popularity of the transistor doomed the compactron.
Compander A shortened version of compressor-expander. A compander
is a device for reducing noise in audio devices such as tape recorders and
is actually a type of lossless codec. The compander will reduce the

76
 Compander

dynamic range of the signal, or compress it, before sending it to be

recorded. The compression makes the softer passages louder so the dynamic
range recorded on the tape is less than it would be if it were not compressed.
Then, on playback of the tape, the signal is expanded; that is, the softer
passages, which are too loud on the tape, are reduced in volume to match
the original signal to restore its dynamics.
In the expansion, which is similar to a fast-acting automatic volume con-
trol, the noise introduced by the tape-recording process is effectively
reduced. This is because the music, when loud, masks the noise, and dur-
ing the soft passages, the volume is turned down, making the noise that
much less loud.
The effectiveness of a compander system is dependent on the ability
of the music to mask the noise during loud passages. If it is not able to do
so, the action of the compander will cause the audible noise level to uctu-
ate up and down with the signal level. This is annoying and is called noise
pumping or breathing. Signals that have relatively strong high-frequency
content suffer less under a compander than do simpler signals.
There have been many designs for companders over the years, with the
Bell Telephone Laboratories taking credit for the earliest ones. The rst com-
mercially successful compander system in recent times was the Dolby A
system, which was widely used commercially in analog recording studios
and in Dolby Stereo optical movie sound tracks. This system divides the

Compander

77
Compatibility

audio bandwidth into four frequency bands and independently com-

presses and expands each one. This reduces the possibility of low-frequency
signals causing noise pumping in the high-frequency range and makes the
system much less audible than it otherwise would be. The Dolby A system
has been simplied for use with home cassette recorders, where it is called
Dolby B and Dolby C. These systems operate in the high-frequency range
only, where tape cassettes are noisiest. Dolby C offers about 20 dB of noise
reduction above 1 kilohertz (kHz), and Dolby B will provide about 10 dB
above 4 kHz.
No doubt, the highest development of the analog compander is the Dolby
SR (for spectral recording), system. The Dolby SR system has also been
simplied for use with cassette recorders, where it is called the Dolby S sys-
tem. It provides 24 dB of noise reduction at high frequencies and 10 dB at
low frequencies.
The dbx noise reduction system is a linear compander in that it com-
presses and expands the signal equally at all frequencies. It employs high-
frequency pre-emphasis and de-emphasis to reduce the effect of noise
pumping.
Compatibility Usually the term compatibility is used to mean stereo-
monaural compatibility, which has several manifestations. stereo record-
ings are said to be mono compatible if they can be played over a monaural
sound system without distortion or loss of tonal balance. This is one of
the concerns of fm radio broadcasters, who have no control over the type
of receivers the listeners may be using.
One of the design criteria of the stereo tape cassette medium was that
it be mono compatible. This means the stereo tracks are placed next to each
other in such a way that if played on a monaural cassette machine, the sig-
nal from both channels will be heard equally.
quadraphonic records, which were released in the 1970s, were
designed to be compatible with two-channel stereo playback.
In former times, after the introduction of stereo records, some were
designed to be compatible with the old monaural playback cartridges. This
meant reducing the stereo separation between the channels, especially at
low frequencies. Virtually all playback cartridges made today are
designed for stereo, so this is no longer a requirement.
Audio devices are also said to be compatible with one another if they
operate correctly when connected together. Often this involves the signal
levels and impedances of the inputs and outputs of the devices. If
recorded media play properly on other machines, the machines are also said
to be compatible. See also impedance; matching.
Complementary Symmetry Complementary symmetry is a circuit
conguration, or topology, which uses a pnp and an npn transistor
connected in a push-pull arrangement. It is often used in the output stage
of audio power ampliers.

78
 Compliance

Complementary Symmetry

Complex Impedance The total impedance, taking into consideration the

magnitude and the phase is called the complex impedance. Sometimes
the nyquist plot of the impedance, which is the real part plotted versus
the imaginary part, is called the complex impedance plot. For more detail,
see Appendix 11.
Complex Wave A sound or an electrical signal that consists of more than
one sine wave is a complex wave. It may be thought of as a combination
of a number of sine waves, in accordance with the fourier theorem.
This should not be confused with complex numbers, which involve the
imaginary unit, or square root of 1. See Appendix 11.
Compliance Compliance is the reciprocal of stiffness. It is a measure of the
ease with which a mechanical structure can be deected. The greater the com-
pliance, the less the stiffness.
Compliance is an important specication of the stylus cantilever in a
phonograph cartridge, and is measured both vertically and laterally. Too
much compliance results in too low a frequency for the tonearm res-
onance, and will cause problems in tracking.
Sometimes one hears about the compliance of the cone in a loud-
speaker. This relates to the low-frequency resonance of the speaker. Cone
compliance is one of the thiele-small parameters.

79
Component

Component An element of an electronic circuit, such as a resistor,

capacitor, transistor, or diode, is called a component.
Component Video A method of handling video signals where the picture
information is separated into three components. One method separates the
picture into Red (R), Green (G), and Blue (B), and is known as RGB. The
other method usually associated with the term separates the picture into
the mathematically equivalent luminance (Y), and two color-difference sig-
nals, Red minus Y (R-Y) and Blue minus Y (B-Y). Whether processing, edit-
ing, or conveying the picture, the three components are processed separately
and do not interfere with one another, hence higher picture quality is
retained. s-video is similar, but only separates chrominance and luminance.
Component video has many variations and notations, often used incor-
rectly. One may see references to Y/R-Y/B-Y, Y/U/V, Y/PB/PR, Y/CB/CR,
and others.
hdtv uses a wide-bandwidth version of component video.
Component video transmits the highest quality picture available to the
consumer. See also composite video.
Composite Video An analog video interconnection scheme that combines
the chrominance (color) with the luminance (brightness signal, Y) and the
synch into one signal. Composite video is available on almost all consumer
video devices. On consumer devices, it usually uses a single rca jack with
a yellow insert, and is patched with a yellow ended cable. The yellow cables
should be 75-ohm video cables. Industrial devices may also use a BNC con-
nector, or on some old devices, a UHF connector. The quality of compos-
ite video is better then that of video over rf, which has to be demodulated
to recover the video signal and also contains the sound information. Com-
posite video cables and connectors are usually good for digital audio con-
nections, such as S/PDIF. Early digital tape recorder systems would encode
the bitstream into a composite video signal and use ordinary video
recorders. See also dat.
Compound The mixture of polyvinyl chloride and additives used to make
analog phonograph records.
Compression A dynamic range problem that occurs in loudspeakers
caused by nonlinearity under conditions of high input power levels. It is
sometimes called power compression. As the power input to a loudspeaker
is increased, the sound power output should increase by the same amount.
This will be true at low and moderate power levels, but at very high lev-
els, the acoustic output increases more slowly or ceases to increase alto-
gether as the input power is increased. Not only does the acoustic output
fail to increase as it should, but distortion is produced and the fre-
quency response curve is usually also different at very high levels. In a
multidriver system, each driver will begin to compress at different power
levels, and the overall frequency response of the system therefore becomes
dependent on the level.
Compression is a characteristic of all loudspeakers, but the better
designs begin to compress at power levels well above those that are gen-

80
 Compression Driver

erally encountered (or should be encountered) in music listening. It is rare

to nd a discussion of compression in the specications provided by loud-
speaker makers.
Compression also refers to the use of a compressor.
Compression Driver A special type of dynamic loudspeaker designed to
t onto the small end of a horn. The horn acts like an acoustical trans-
former, with the driver providing a relatively high sound pressure at the
throat of the horn, and the mouth of the horn providing a large area of low
pressure to radiate the sound efciently into the air.
The compression driver was developed by the Bell Telephone Labora-
tories people in the early 1930s, and their basic designs are still used with
very little change. The voice coil is directly connected to a small metallic
dome, or diaphragm, whose inside surface radiates the sound into the
horn throat. One of the design problems for such a unit is how to ensure
that sounds from all parts of the diaphragm arrive at the throat at the same
time in order to prevent phase cancellation at certain frequencies. The

Compression Driver

81
Compressor

diaphragm must be large enough to handle high power levels but must
be small enough to avoid the phase problems. The solution is the phas-
ing plug, which consists of a series of annular or tangential slits through
which the sound passes.
In high-power compression drivers, it is possible to attain such high
sound pressure levels in the throat of the horn that nonlinearities in the com-
pression of the air are introduced. This causes very signicant harmonic
distortion and is partly responsible for the objectionable sound of some
high-power horn systems.
Compressor An audio device that reduces the dynamic range of a sig-
nal. The compressor is the rst part of a compander.
The effect of the compressor is to make the loud parts of a signal softer
and to make the very soft parts louder. Compressors are frequently used
in recording popular music and in radio broadcasting, where very soft pas-
sages may be lost in the background noise of the listening environment.
For instance, when music is playing on the radio in a car, the cars noise
level will easily mask the quieter musical passages.
The limiter acts something like a compressor but operates only at the
top end of the dynamic range. The subjective audibility of a compressor
depends strongly on its time constants, and they are selected with care
to minimize obvious pumping of the volume. To restore the original
dynamics to a compressed signal, a volume expander can be used, but
great care must be taken that the time constants match those of the
compressor.
COMTRAK According to its developer, the late John Mosely, COMTRAK
stood for Combined Academy Monophonic and Compatible Four-Track
Stereophonic Photographic Sound Track. Among the myriad of audio-
related acronyms and abbreviations this is surely the most obscure.
The COMTRAK system was a form of optical motion picture sound track
that provided four independent stereo channels (left, center, right, and sur-
round) plus a conventional variable area monaural track. The stereo
channels were recorded alongside the monaural track and were of constant
area and so were not picked up by the standard photocell in a conventional
projector. A special reproducer that looks at only the edge of each track was
used to recover the stereo information. In addition to the four stereo tracks,
there was a control track that could be frequency modulated with very
low frequencies for special effects. To attain good dynamic range, the four
channels were companded.
The COMTRAK system was attractive from several standpoints, not the
least of which was that it was compatible with the then-widespread con-
ventional monaural theater sound. It also provided four independent
channels of audio, whereas other sva systems used matrixing to encode
four channels onto two. COMTRAK was developed in the late 1970s and
was never used commercially. Mr. Mosely lamented that acceptance and
commercialization of the system was a classic chicken and egg situation.
See also colortek.

82
 Condenser Microphone

Condenser Condenser is the old-fashioned name for the capacitor. It is

interesting that some uses of the term have lingered on, such as condenser
microphone.
Condenser Microphone One of the earliest types of microphone to be
invented after Dr. Lee DeForest invented the audion amplier in 1906 was
the condenser microphone. Thomas Edison is sometimes credited with its
invention, but this seems to be in doubt. At any rate, Wente, of Bell Tele-
phone Labs, designed a condenser microphone in 1917 and introduced it
commercially in 1931.
The condenser microphone is a very simple mechanical system, with
almost no moving parts compared to other microphone types. It is sim-
ply a thin stretched diaphragm held very close to a metal disc called a back
plate. This arrangement is an electrical capacitor, and it is given an elec-
tric charge by an external voltage source or a permanently charged
electret material. When sound acts on the diaphragm, the pressure vari-
ations cause it to move slightly in response to the sound waveform. This
causes the capacitance to vary in like manner, and because the charge
is xed, the voltage on the back plate will vary according to the laws gov-
erning the capacitor. This voltage variation is the signal output of the
microphone. The condenser microphone has extremely high output
impedance, and must be placed very near a preamplier to avoid loss
of the signal.
In its simplest form, the condenser microphone is a pressure micro-
phone, sensitive to sound pressure variations. It behaves like a very high
speed barometer, sensing instantaneous variations in atmospheric pressure.
The pressure exerted by a sound is independent of direction, and so a pres-
sure microphone is nondirectional, or omnidirectional. It is generally
agreed that the highest quality microphones are condensers.
It is possible by special treatment of the back plate and by combining
several microphone elements to attain various directional patterns, includ-
ing bi-directional, cardioid, and super-cardioid.

Condenser Microphone

83
Conductance

Conductance The reciprocal of resistance, or electric current divided

by voltage. The traditional unit of conductance was the mho, which is
ohm spelled backward. The si metric unit is the siemens, named after the
German inventor who made the rst moving-coil dynamic loudspeaker.
See also transconductance.
Cone The surface, usually conical in shape, which actually radiates the
sound in a direct radiator loudspeaker is called the cone.
The traditional material for speaker cones was paper, and most cones
today are made of a paperlike material that has added llers for damping
of resonances. Recently, the plastic polypropylene has been used for cones,
and it has excellent characteristics for this use. It has very high internal
damping and is strong.
Cone Breakup The motion of a loudspeaker cone in different modes of
vibration such that it no longer moves as a solid piston.
It is desirable to operate a loudspeaker only at frequencies below that
of the rst mode of vibration, but this is difcult to achieve in practice. Oper-
ation at frequencies of the higher modes causes nonuniform radiation with
rough frequency response and polar response of the speaker. damp-
ing can be applied to the cone to reduce the effects of breakup, and this
adds mass and reduces the rst resonant frequency, which is not necessarily
undesirable.
Condence Head Some tape recorders have extra hardware that allows
playing the actual recorded signal moments after the fact, during the
recording process. If the signal thus played sounds OK, one has some
condence that the recording is proceeding correctly. The term is primar-
ily used with DAT recorders. Open reel and cassette recorders with such
systems are referred to as three head.
Console In studio parlance, a mixing console, or in Britain, a desk.
Consonant Literally sounding together. Musical tones that are consonant
sound harmonious or in tune when sounded together rather than dis-
cordant or harsh. Musical intervals composed of tones that have relatively
simple frequency ratios are more consonant than ones with more complex
ratios. The most consonant interval is considered to be the octave, which
has a frequency ratio of 2 to 1. See also Appendix 8.
Constant Directivity Horn, CD Horn Conventional horn-loaded, high-
frequency loudspeakers have a directivity pattern, or q, which varies with
frequency. Typically, the horn becomes more sharply directional as the
frequency rises. In recent years, designs have been introduced which com-
bine the operating principle of the column loudspeaker and the horn.
To do this, the horn has two shapes, although the rate of area expansion
(are rate) is constant over its length. The rst section of the horn will be
narrow and at, terminating in a vertical slot. This slot is connected to
another horn that will are both vertically and horizontally and has a square
mouth. Such a horn will distribute the high-frequency energy more uni-
formly in the horizontal direction, but at the expense of ragged frequency

84
 Control and Display Symbols

response, which is compensated for by special equalization called CD Horn

EQ.
Constant Displacement See constant velocity.
Constant Q See equalizer and q.
Constant Velocity A magnetic phonograph cartridge is sensitive to the
velocity of tip motion, and will produce the same output voltage at any
frequency if this velocity is held constant. Records are cut so that the veloc-
ity of the stylus is constant over a certain frequency range, and at a given
rms velocity, as frequency rises, the stylus displacement decreases, and con-
versely, as frequency goes down, the displacement increases. If this con-
stant velocity were maintained down to low frequencies, the displacement
of the grooves would be so much that they would have to be too widely
spaced on the record, and the maximum recorded time would be too short.
(See also variable pitch.) Therefore, below a certain frequency, called the
turnover frequency, the recording is made at constant displacement. In
order to achieve at frequency response, the low frequencies must be
boosted to compensate for the reduced velocity. This is accomplished by
the riaa equalization curve built into the preamplifier.
The riaa pre-emphasis and de-emphasis also involves a high-
frequency boost in cutting the record and a high-frequency cut in repro-
ducing it in order to reduce the effect of high-frequency surface noise.
Contact Microphone Strictly speaking, a contact microphone is not a
microphone at all, but is a vibration transducer that generates an elec-
tric signal proportional to the vibration of the surface to which it is applied.
Such a device can be used as a microphone, although the signal produced
will not generally resemble the signal an air microphone would produce
from the same instrument. Nevertheless, contact microphones are often used
for sound reinforcement and recording of musical instruments. The primary
advantage of contact microphones is that they are not sensitive to the
acoustics of the room, and they do not pick up the sounds from other
adjacent instruments.
Continuous Wave One of the important innovations in early radio trans-
mission was the use of a continuous carrier signal modulated by the audio
signal. The technique was rst perfected around 1902 by Reginald Fessen-
den, the great American pioneer of early radio. Before Fessenden, radio
transmission was by damped-wave carriers, which were the result of using
a spark gap to generate the radiated signals. Spark gap transmitters trans-
mitted telegraph messages. Fessenden wanted to be able to transmit speech
rather than telegraph messages by radio, and he knew the damped-wave
technique would not allow this. The rst continuous wave (CW) trans-
mitters used a carbon arc as part of the resonant transmitter circuit and
were quite inefcient and noisy. Fessenden later used a high-frequency alter-
nator, or AC generator, to create the carrier signal, and this became the stan-
dard of the day.
Control and Display (C&D) Symbols See subcodes.

85
Control Track

Control Track One track of a multitrack magnetic tape recorder used for
recording special signals that provide control information to the recording
console during automated mixdown. Also, VCRs and some VCR-based
digital audio recorders record a linear track of timed pulses to synchro-
nize the tape speed and helical scan head tracks, which is called a control
track.
Control Voltage A direct voltage, usually varying, used in electronic
music synthesizers to control various parameters of the signal being pro-
duced. Control voltages are used for envelope control, frequency con-
trol, and lter bandpass and cutoff frequency control, etc. Suitable control
voltages can be generated in various ways, one of the most straightforward
of which is by a standard keyboard.
Convolution In any linear system or device, the output signal is a func-
tion of the input signal and the characteristics of the device. The interac-
tion between the input and the device is described by a mathematical innite
integral called convolution. The output is the input convolved with the
impulse response of the device. The mathematical operation of convo-
lution is complex, and it is not generally easy to ascertain the characteris-
tics of a device by looking at the output signal directly. However, a fortunate
circumstance is that the Fourier transform converts the convolution into a
simple multiplication. This is why spectrum analysis is used so much in
audio measurements. The spectrum of the output of a device is simply the
spectrum of the input multiplied by the frequency response of the device.
The Fourier transform is easy to implement with modern fft analyzers.
See also deconvolution.
Copper Loss A loss of power in a circuit because of heating due to the
resistance of the wire. The term is most commonly used in conjunction with
power transformers and power supplies.
Copy Code A technique, proposed by CBS in 1982, that puts a narrow deep
notch at 3.8 kHz in the frequency response of compact discs and a
circuit in r-dat recorders to sense the absence of signal at this fre-
quency. If any signal exists at this frequency, the dat machine will not
enter the record mode. The idea was to prevent the unauthorized copying
of CDs. CBS claimed that the notch was inaudible, which of course was
nonsense, and no one ever adopted the system. See also scms.
Core The magnetic material in a transformer around which the coils are
wound. The cores of most audio transformers are made of thin laminations
of a high-permeability material such as iron or mumetal. The central
metal part of a magnetic tape head is also called the core.
Corner Frequency The frequency at which the response of an equalizer
or other audio device is reduced by 3 dB. This is also sometimes called the
half-power point and can refer to both low-pass and high-pass response
curves. The term probably comes from the appearance of the frequency
response graph, which bends or turns a corner near this point.
Cosine Pattern The polar response curve of certain loudspeakers and
microphones can be expressed mathematically as a polar plot of the

86
 Coulomb

trigonometric cosine function. This curve consists of two circles if plotted in

linear units. The responses of audio devices are usually plotted in decibel
units, and the cosine curve plotted this way looks like two distorted circles.

Cosine Pattern

Cough Button A push-button usually placed within reach of a radio

announcer that disconnects his microphone in case he has to cough, hic-
cup, fart, etc.
Coulomb The coulomb is the unit of electric charge, abbreviated C, and
is the quantity of electricity transferred in 1 second by a current of 1
ampere. It is named after C. A. de Coulomb, a French physicist who worked
in electrostatics.

87
Coupling

Coupling In an audio device with more than one stage of amplication,

the stages are connected, or coupled together. Frequently a capacitor
is used to transfer the AC signal and to block any DC component. This is
called capacitive coupling, and the capacitor is called a coupling capacitor.
In a transformer, the coils are said to be magnetically coupled,
although there is no direct electrical connection between them. The amount
of coupling can vary depending on the physical arrangement of the coils
on the core. Coils that are close together or even interleaved are said to
be tightly coupled, and this increases the efciency of the transformer action.
Magnetic coupling can become a source of unwanted noise, especially
60-hertz hum. Stray magnetic elds from power transformers or from
wires carrying large amounts of 60-Hz current can induce 60-Hz noise
in signal wiring. See also balanced.
CRAP Completely Ridiculous Audio Performance. According to Dennis
Bohn of the Rane Corp., a favorite acronym describing poor quality sound
equipment.
Crescendo A gradual increase in loudness of a musical sound. In con-
ventional music, a crescendo is usually attained by gradually increasing
the number of instruments sounding. In electronic music, a crescendo may
be achieved by simply increasing the volume of the generated music signal.
Crest Factor Crest factor is the ratio of the peak level of an audio signal
to its rms value, usually measured in decibels. For instance, a sine wave
has a peak value of 1.4 times its rms value. This is a crest factor of 1.4, which
in decibel notation amounts to 20 log 1.4, or 3 dB.
Musical waveforms are never smooth and uniform, but are rough and
peaky. transients, such as the beginning attacks of percussive sounds,
have high instantaneous levels, but fall rapidly to a much lower continu-
ous level. Speech sounds have short-duration transients as well, especially
if the microphone is close to the source of the sound. The conventional
way of measuring the level of an audio signal is the venerable vu meter,
which responds to the average value of the signal, but this does not tell very
much about the peak level, and it is the peak level that causes overload
of audio devices and consequent distortion. The crest factor is a mea-
sure of how peaky the signal is, and the greater its value, the greater the
headroom needed in the audio chain. See also modulometer.
Critical Band In hearing, the critical band for loudness is that fre-
quency band within which the loudness of a band of noise of constant
sound pressure level is independent of its bandwidth. If the noise
bandwidth exceeds the critical band, the loudness will increase.
The critical band is somewhat less than one-third octave wide over
much of the audible frequency range. It acts something like a one-third
octave filter whose center frequency can be anyplace along the audible
frequency axis. The loudness of several tones, if they lie within one critical
band, can be predicted by summing the power of all the components. If,
however, the several tones are in different critical bands, then each one will
have a particular loudness, and these loudnesses will sum to arrive at the

88
 Cross-Field Head

combined loudness. Thus, components more widely spaced will sound

louder than if they are close together in frequency.
The measurement of the width of the critical band is difcult, and dif-
ferent methods give different answers. One of the earliest investigators into
the critical band was Harvey Fletcher, who was working in the 1920s and
1930s.
Critical Distance In a reverberant space, the sound perceived by a listener
is a combination of direct and reverberant sound. The ratio of direct to rever-
berant sound is dependent on the distance between the source and the lis-
tener and upon the reverberation time. At a certain distance, the two
will be equal (the direct-to-reverberant ratio will be 1). This is called the
critical distance, and the concept is useful to designers of sound reinforce-
ment systems, where the sound heard by the listener must not be too rever-
berant if clarity and intelligibility are to be maintained.
The critical distance depends on the sound source: a source that is direc-
tional in its output will have a longer critical distance than a nondirectional
source because less of the rooms reverberation will be excited. For this rea-
son, directional loudspeakers are used in sound reinforcement systems.
Critical Frequency The highest radio frequency that will be reected
from the earths ionosphere to allow reception at great distances from the
transmitter. The ionosphere is a complex and constantly changing area in
the upper atmosphere. It reects frequencies lower than the critical fre-
quency and transmits higher ones to outer space. For any particular atmos-
pheric condition and depending on the time of day, the season of the year,
and the sunspot activity, there will be a maximum usable frequency, or MUF.
Crolyn A trademark and patent of DuPont for their formulation of
chromium dioxide recording tape.
Cross-Correlation Cross-correlation is a mathematical measure of how
much a signal resembles another signal. Two signals that are identical will
have a cross-correlation of 1, and two signals that are completely different
from each other will have a cross-correlation of 0. The cross-correlation
between two signals is also a function of the time delay between them.
Cross-correlation can be used to measure the accuracy of an audio com-
ponent, or series of components. If, for example, the input voltage to a
loudspeaker is correlated with the output of a precision microphone
in the room, the cross-correlation should be 1 for the time delay equal to
the transit time of the signal through the air, neglecting reverberation.
Any inaccuracy in the loudspeaker (and any reverberation or noise added
by the room) will cause the cross-correlation to be less than 1. See also inter-
aural cross-correlation (iacc).
Cross-Fade In editing of recorded material, it is possible to more or less
rapidly fade one signal out while the other is faded in. This is called a cross-
fade. See also butt splice.
Cross-Feed See ster-bin.
Cross-Field Head The cross-eld head is a type of magnetic tape head
for applying high-frequency bias to the tape during recording of a signal.

89
Crossover Distortion

It is on the back side of the tape opposite the audio recording head, and
applies the bias eld through the plastic backing.
Normally, the bias is mixed with the signal and applied to the record
head. This conguration causes self-erasure of high frequencies due
to the strong bias eld in the record head gap. The cross-eld method
reduces this tendency, and allows the recording of higher levels of high
frequencies. One disadvantage is that the position of the head is some-
what critical, and different tape thicknesses result in different effective bias
levels.
Crossover Distortion Crossover distortion is a type of distortion present
in some amplifiers that increases for low-level signals. In many
ampliers, the output devices (usually transistors) are so connected that
one of them is active during the positive half of the waveform and the
other one is active for the negative half. There is thus a region near zero
current where the signal is transferred from one to the other. If this is not
done smoothly, there will result a small discontinuity in the output wave-
form. This discontinuity causes high-order harmonic distortion, and,
being constant in value, it is more noticeable with low-level signals than
with stronger ones.

Crossover Distortion

90
 Crossover Network

The phenomenon is reduced by application of negative feedback

around the output stages of the amplier, and by proper bias current in
the transistors. Modern ampliers are nearly free of the effect, but early
solid-state ampliers were plagued by it.
Crossover Frequency See crossover network.
Crossover Network In multiple-driver loudspeakers where a different
speaker is used to cover each of several frequency ranges, some means
is needed to allocate the frequency bands to the appropriate drivers. This
is the crossover network, which is actually a type of filter. It varies in
sophistication depending on the number of drivers used and on their type.
In a two-way system, the signal for the low-frequency driver will pass
through a low-pass filter, which attenuates signals above a certain fre-
quency, and the high-frequency unit will be fed from another lter, which
attenuates frequencies below that frequency. This frequency is called the
crossover frequency, and at that frequency, both drivers will receive the same
amount of energy. The two lters will typically have 3 decibels of atten-
uation at the crossover frequency so the sound output of the system is
at over the crossover region.

Crossover Network

The high- and low-pass lters of a crossover network may have various
rates of attenuation outside their passband, and values of 6, 12, and 18 dB
per octave are common. The larger rates of attenuation are generally
required for speaker systems using horn-loaded, high-frequency drivers
rather than direct radiators because horn-loaded units will not han-
dle very much power below their cutoff frequency.
Crossover network design is somewhat complicated because lters
cannot be designed that have the needed frequency response charac-
teristics without also having signicant phase shift near the crossover

91
Crosstalk

frequency. This prevents the sounds from the individual drivers from adding
together to form a coherent sound. This phase nonlinearity can be heard
in many speaker systems, especially those with crossover frequencies near
the middle of the ears most sensitive range of about 1,000 to 4,000 hertz.
The crossover network is usually built into the speaker system, where
it must handle the entire power output of the amplifier. It is then called
a passive crossover. It also may be placed before the amplier, in which
case a separate amplier is needed for each frequency range and driver.
This is called an active crossover. Such a system is often called bi-
amplied, or triamplied in the case of a three-way system. Such a crossover
need handle only very small signal levels, and many units of this type use
active electronic components (transistors or integrated circuits),
which is why they are called active crossovers. Many of them are able to
provide variable crossover frequencies to tailor the response to suit vari-
ous different loudspeaker congurations.
Crosstalk In multichannel audio transmission systems, such as tape
recorders, record players, or telephone lines, a signal leaking from one
channel to one or more of the others is called crosstalk. The term comes
from early telephone company usage.
Crosstalk is almost always measured as a power ratio and is expressed
in decibels, dB. channel separation in stereo systems is the inverse
of crosstalk and is also expressed in decibels. Channel separation of 30 dB
is the same as saying the crosstalk between the channels is minus 30 dB.
Phono cartridge performance is ordinarily stated this way. Crosstalk usu-
ally varies with frequency, so a simple number does not tell the whole
story. Crosstalk can be reduced by using balanced lines, proper shield-
ing, and correct equipment design.
In a stereo music system, there is not a great need for extremely low
crosstalk, for crosstalk of less than 20 dB or so cannot be heard because of
the similarity of the two channels and because of masking of each chan-
nel by the other. But if the adjacent channels carry different programs, such
as is the case with quarter-track tape recordings that are recorded in both
directions, the crosstalk must be much lower to be inaudible, on the order
of 60 dB or better.
CRT See cathode ray tube.
Crystal The frequency-determining device in radio or television transmit-
ters and digital clocks is a small vibrating piezo-electric crystal, usu-
ally of quartz. Small pieces of such crystals will naturally vibrate at a certain
frequency when an alternating voltage is applied across them. The frequency
of vibration can range from kHz to GHz, and it depends on the size and
dimensions of the crystal. All digital audio systems use a crystal to gener-
ate the high-frequency clock signal used to time the events involved in
analog-to-digital converters and digital-to-analog convert-
ers. See also crystal set; crystal synch.
Crystal Set An early type of radio receiver, known as a crystal set, used a
small crystal of galena, or lead sulde, as a detector. The galena crystal

92
 Curve

acted as a solid-state rectier, or diode, to recover the audio signal from

the radio wave. To achieve suitable sensitivity, the listener had to carefully
place a small wire, the cats whisker, at various places on the crystal while
listening to the program over headphones. It is interesting to note that the
rst radio receivers were of solid-state design, although most people think
of solid state as a recent development.
Crystal Synch Crystal synch is a method of running the tape recorder and
camera in a motion picture production setup at precisely constant speed
to ensure that they will remain synchronized when combined into the edited
sound lm. No interconnection between camera and recorder is needed.
A quartz crystal is used as a frequency standard and the camera drive
motor is controlled by this stable signal. The tape recorder used for the
audio recording also contains a crystal, which records an accurate frequency
on the tape along with the audio. When the tape is reproduced, this fre-
quency is sensed and the speed of the machine is controlled to hold it con-
stant, or to keep it in synchronization with the movie projector.
The advent of the smpte time code has made crystal synch obsolete.
CSA Acronym for Canadian Standards Association, which is loosely anal-
ogous to the National Institute of Standards and Technology (nist) in the
United States.
CTD Acronym for charge transfer device. See bucket brigade.
Cue To position a recording at the start of a recorded selection or to posi-
tion a playback stylus at the beginning of a record cut is to cue or cue up
the selection. Cuing is one of the most important activities that a disk jockey
performs, although the skill is different today with digital media.
The signal sent to the monitor loudspeakers for studio musicians to
listen to is sometimes called a cue mix. The monitor speaker is also some-
times called the cue box.
Current In electrical and electronic circuits, the signal consists of elec-
tric currents, and these are the ow of electrical charge. Audio signals are
always alternating currents, meaning the current reverses direction
each time the signal waveform passes zero. The current from a battery,
in contrast, is a direct current, meaning it always has the same direction.
Electric current does not ow in a circuit; it is the electric charge that
ows, and the ow of charge constitutes the current.
Curve A curve is a graphic representation of the relationship between two
quantities, called the independent and the dependent variables. The depen-
dent variable is said to be a function of the independent variable.
In a frequency response curve, the response (dependent variable) is
plotted vertically as a function of the frequency (independent variable),
which is plotted horizontally. In this case, the best curve is not curved but
straight, as in flat response.
Frequently in audio, the variables are not plotted directly, but their log-
arithms are plotted. In frequency responses, it is almost always the case
that the horizontal axis will be log frequency, and the vertical axis will be
decibels, which is a log amplitude scale.

93
Cut

Cut A reduction in amplitude of a signal level, usually referred to a

specic frequency or frequency range, is a cut. equalizers, such as the
common tone controls, cause cuts and boosts of selected frequency
ranges.
The making of an acetate master record is often called a cut. Some-
times in recording studios, the recording of a tape is called a cut, proba-
bly because of earlier usage in record cutting. One individual song on a
record may be called a cut.
Cutoff Frequency In a high-pass or low-pass filter, the cutoff frequency
is where the amplitude response is 3 decibels below the response in the
passband. This is also sometimes called the half-power point. The response
of the filter falls off more or less rapidly above or below the cutoff fre-
quency depending on its design.
Cutterhead In the making of an acetate master record, a tiny, pointed,
chisel-shaped stylus is used to cut the groove, and this stylus is moved
in response to the audio signal by the cutterhead. To cut a stereo record,
the stylus must be moved in two orthogonal directions at once. Each
motion is directed to cut one side of the groove, and the motions are 90
degrees from one another. The stylus is driven by two small voice coils
attached to the stylus lever arm. The two coils operate like miniature loud-
speakers without cones and they must be designed so that each one drives
the stylus in one direction only and does not interfere with the other motion
imparted by the other coil.
The stylus assembly in a cutterhead is very stiff, and it takes consider-
able power to drive the coils, typically several hundred watts. Most of
this power is dissipated as heat, and the cutterhead must be cooled, usu-
ally by blowing helium around the coils.
To ensure a smooth and noise-free cut, the stylus itself is heated by a
small coil of resistance wire wrapped around it.
CW See continuous wave.
CX A type of compander noise reduction system for stereo records, intro-
duced with much ballyhoo by CBS in the early 1980s. The system works
by compressing the dynamic range on the record and expanding it on play-
back to restore the musical dynamics to the original levels. The expansion
on playback effectively reduces the background noise added by the record
medium. CX works in a manner similar to that of the dbx system except
CX operates on only the upper 40 dB of the input signal dynamic range,
providing 20 dB of noise reduction. Signal levels below -40 dB are not
altered. To properly decode the recorded signal, the decoder has to be
matched in level to the recorded signal level. This level-matching require-
ment is similar to that of the Dolby noise reduction systems. To achieve this
level matching, decoder manufacturers provide a small record containing
a reference tone to be played into the decoder in order to set the gain to the
proper value. About fteen manufacturers were initially licensed to pro-
duce decoders.
At the time of its introduction, CBS made the claim that even though

94
 Damp

the CX-encoded records are not compatible with conventional playback,

the degradation of sound when playing them undecoded was not signicant
for most records or most listeners, even though the upper 40 dB of musi-
cal dynamics were compressed into only a 20-dB range. This, of course,
was preposterous, and the system died an early death in the marketplace.
The CX system found success in the audio tracks on laser video discs,
and most video disc players included the necessary decoders.
Cycles per Second, cps Until 1948, the accepted unit of frequency. In 1948
hertz, abbreviated Hz, was adopted as the standard unit of frequency.
Cylinder The rst commercial phonograph, built by Thomas Edison,
recorded on a hollow cylinder of wax; the records are called cylinders. Cylin-
der records had the advantage of a constant speed of groove with respect
to stylus, in contrast to the disc record, whose relative speed decreases as
the groove nears the center of the record. This fact convinced Edison that
the disc record would never overtake the cylinder. The deciding factor was,
of course, that the disc requires much less storage space for a given play-
ing time. The reduced stylus velocity is, however, a real disadvantage, along
with the curvature of the groove, which causes tracking error.
C-Zero An empty tape cassette shell that is to be loaded with tape as the
tape is being recorded in a high-speed tape duplication machine.

D
DA See distribution amplifier.
DA-88 The rst in a family of modular digital multitrack tape recorders
introduced by Tascam in 1993. It uses Hi-8 video tape cassettes to record 8
tracks of PCM digital audio, and units can be added to create up to 128
tracks. DA-88 and its siblings, the DA-38 and DA-98, are also called
DTRS for Digital Tape Recording System. The format was also offered
by Sony, and is considered a competitor to the similar but incompatible Ale-
sis adat system.
DAC Abbreviation for digital-to-analog converter.
DAD Short for Digital Audio Disc, also known as CD-DA, for Compact
Disc-Digital Audio. See compact disc.
Damp, Damping Damping is the addition of friction to a resonance to
remove energy from it and thus reduce its magnitude. For example, a clock
pendulum is a resonant system with very little friction. If it is placed in a
jar of molasses, it will no longer vibrate because of the frictional damping
of the molasses.
Almost all mechanical systems have resonances, and they tend to vibrate,
or ring, at the resonant frequencies. This is desirable in a bell, but not in
a loudspeaker or phono cartridge, which should not contribute any
sound of its own to the music. Therefore, speakers and cartridges have fric-
tional damping added to them, sometimes without complete success.

95
Dampen

The electrical analog of friction is resistance, and it is used to damp

resonating electrical circuits, such as crossover networks and filters.
See also q.
Dampen To add water to something, to make damp. This has nothing
to do with damping, and it is surprising how often the term dampening is
used when damping is intended.
Damping Factor The damping factor refers to the ratio of the rated load
impedance of an amplifier to the actual output impedance of the
amplier. It is the ability of the amplier to control the motion of a loud-
speaker cone after a signal disappears. An amplier with a high damp-
ing factor looks more like a short circuit to the speaker, reducing its
vibration when the signal stops. This is because the speaker acts like an
electrical generator when the cone is moving, and the low impedance of
the amplier absorbs the energy from the generated signal, damping the
cone.
Damping factors above 15 to 20 are not signicantly better than those
slightly lower because the resistance of the loudspeaker wires must be added
to the output impedance of the amplier, reducing the effectiveness of damp-
ing. There is a common misconception that the impedance of the voice coil
of the speaker also must be added to the amplier output impedance when
calculating the effect of the damping factor, but this is not true. Voice coils
of higher impedance generate greater voltages, which are correspond-
ingly reduced more by the low impedance of the amplier. For the same
reason, loudspeakers connected in series will experience the same damp-
ing when connected to an amplier as they do when connected in parallel.
The damping factor of an amplier will vary with frequency, and
sometimes a manufacturer will publish a curve of damping factor versus
frequency. The effect of high damping factors is most audible at low fre-
quencies, where the primary resonance of the woofer cone is reduced in
level.
DAO Disc At Once, a method of writing an audio CD-R where all infor-
mation on the disc is written in one pass, without the writing laser ever
being interrupted. See also tao.
DAP See digital acoustics processor.
DASH Short for Digital Audio Stationary Head. DASH is one standard for-
mat for digital multitrack studio recorders that use stationary, rather than
rotating, heads. The open reel format covers a wide range of versions, such
as from 2 to 48 tracks and tape speeds from 12 to 76 cm per second, and
was agreed on by such manufacturers as Sony and Studer. A competing,
and similar but incompatible system called ProDigi was promoted by Otari,
AEG, and Mitsubishi.
DAT Generally, any Digital Audio Tape recorder, but more commonly, a
standardized 4mm cassette format for recording digital audio on mag-
netic tape with a rotating head. Originally designed for consumers, the
standardized DAT cassette format became widely used by professionals

96
 dB FS

instead. The same cassette format is used for data recording, but should be
called DDS (Digital Data Storage) in that application and not DAT. DDS
tapes are functionally the same as audio DATs, however, they are labeled
in tape length in meters rather than time, and come with some data for-
matting already recorded. The 60 meter DDS tape is equivalent to the 120
minute audio DAT, and many people nd the lower price of the DDS and
equal performance attractive. Longer DDS tapes are not recommended for
audio use as the tape and coating is thinner, and the audio transport is not
designed to accommodate that. However, it is one way to get 3 or more
hours on an audio DAT if one is careful.
The rst consumer digital audio tape recorders were actually adapters
to allow audio recording on videocassette recorders. Examples are the Sony
PCM-F1 processor and the dbx 700 system. In these devices, the signal
must be formatted so it resembles a video signal (in other words, broken
up into frames so the recorder will accept it). This also puts certain con-
straints on the sampling rate.
There are two general DAT technologies: the R-DAT, which uses a rotat-
ing head assembly similar to a VCR; and the S-DAT, which uses a station-
ary head. The R-DAT records diagonally across the tape while the S-DAT
records several linear parallel tracks of digital signals. The R-DAT is more
complex mechanically and the S-DAT requires more electronics. The eiaj
rst issued a standard for R-DAT and later issued one for S-DAT. The R-
DAT standard includes a four-channel format that would permit record-
ing of ambisonics, although it is seldom seen. The standard 4mm DAT
format utilizes the rotating head system, and is always called simply DAT
instead of R-DAT. S-DAT systems are usually referred to by their trademark
format such as DASH or ProDigi.
DAW Acronym for Digital Audio Workstation. A DAW consists of a com-
puter with a large amount of disc storage space and specialized software
for editing and modifying digitized audio signals. It usually also includes
high-quality analog-to-digital and digital-to-analog convert-
ers. DAWs vary in their complexity and capabilities, and the best (and most
expensive) ones can perform many modications to audio signals such as
editing, noise reduction, frequency equalization, adding reverberation
and/or echo, dynamic range compression and expansion, etc. on many
tracks. DAWs can be assembled from personal computers, or can be bought
as a manufactured product.
dB See decibel.
dBA Sound pressure level measured using an a-weighting lter. See also
a weighting.
dBf Literally, decibels referred to 1 femtowatt, or 10 to the minus fteenth
power watts. The dBf is used to specify signal levels at the RF inputs to
fm tuners.
dB FS The absolute digital signal level referenced to full scale on a digi-
tal level detector.

97
dBk

dBk The signal level in decibels referred to 1,000 watts. dBk is most often
used in radio and television transmitting stations.
dBm Literally, decibels referred to 1 milliwatt. The term dBm is now obso-
lete, and was replaced by dB (mW). Strictly speaking, the 1 milliwatt must
be dissipated in a load of 600 ohms. The dB (mW) is used in stating power
levels of signals in broadcast, recording consoles, and in tape recorders. 1
dB (mW) into 600 ohms will result in .775 volts rms. This causes some
confusion. The dB (mW) is a power level and only results in .775 V when
the load impedance is 600 ohms. See also decibel.
DBS, Digital Broadcast Satellite A digital satellite system that uses geo-
synchronous earth satellites to broadcast more than 200 channels of high
quality audio and television programming to home subscribers using small
dish antennas. Both DISH Network and DIRECTV use the DBS system.
dBu Formerly the signal level in decibels referenced to one microwatt into
an impedance of 600 ohms. The term has been redened in recent years to
the signal voltage level referenced to 0.775 Vrms with no load impedance
specied. The new dBu (perhaps as in dB unterminated) is equivalent to
dBv with a lower case v. The use of dBv is discouraged due to likely con-
fusion with dBV.
dB The signal level in decibels referred to 1 microwatt.
dBv See dBu.
dBV Literally, decibels referred to 1 volt rms. This is an unfortunate
usage because decibels cannot properly compare simple voltages unless
their impedances are the same. See also decibel.
dBW Literally, decibels referred to 1 watt. Power amplier output lev-
els are sometimes specied in dBW, a 100-W amplifier is then 20 dBW.
Power rating in dBW is numerically equal to ten times the common log-
arithm of the power output in watts. See also decibel.
DC Short for direct current (see also alternating current). Direct cur-
rent always has the same direction, namely, from the positive terminal
(anode) to the negative terminal (cathode). This convention of describing
electricity owing from positive to negative was established by Benjamin
Franklin, who hypothesized that charge consisted of tiny units of positive
electricity, and the convention has been retained to the present day. How-
ever, it has been known for a very long time that an electric current usu-
ally consists of a ow of electrons, which are negatively charged and move
from negative to positive. Sometimes in a semiconductor the current can
consist of positive charges, called minority carriers or holes, and in these
cases, Franklins convention was right. In any case, it would be inconven-
ient to reverse the convention, for all the shortcuts engineers like to use,
such as right-hand rules would have to be changed.
DC Amplier An amplifier whose frequency response extends all the way
to zero frequency, or to DC. Such an amplier is also called direct cou-
pled, meaning there are no capacitors in the signal chain. An example
of a DC amplier is an opamp in its raw state with no input or output capac-
itors. Some audio power ampliers are direct coupled, and it is true that a

98
 Decibel

DC amplier will not have any phase shift caused by the low-frequency
rolloff that capacitive coupling provides. However, if the bandwidth
of an AC amplier extends to sufciently low frequencies, this phase shift
is negligible.
DC Bias See bias.
DCC See digital compact cassette.
Dead In acoustics of rooms, the effect of very little reverberation is
called a dead room. A dead recording is one with very little or no recorded
reverberation or ambience of the space where the recording was made.
Debouncing The removal of noise and multiple making and breaking of
switch contacts used for voltage-controlled switching. Switches, especially
push-button switches, do not usually make a clean contact when they are
closed but tend to bounce, causing ambiguous or erroneous signals to
be sent to the controlling elements. Debouncing can be as simple as shunt-
ing various combinations of resistors and capacitors across the switch
or can become quite complex with the use of pulse-forming and shaping
integrated circuits.
Decade A frequency ratio or interval of 10 to 1, as opposed to an octave,
which is a 2 to 1 ratio. Sometimes the rolloff of a lter or equalizer is
expressed in decibels per decade rather than in decibels per octave. A
rolloff of 20 dB per decade is equal to a rolloff of 6 dB per octave. The decade
interval has no musical signicance.
Decay Time Decay time is a synonym for reverberation time, or the
time it takes a sound to decay to one millionth of its former strength. This
is a 60-decibel reduction in level.
Decca Tree A three-microphone array for two-channel stereo recording
originally devised by Decca Records in England. The microphones used
are omnidirectional, making the Decca tree a special case of spaced omnis,
or a-b stereo. The right and left microphones are spaced about 2 meters
apart, and the third, or center, microphone is on the center line between the
other two but is located about 1.5 meter in front of them. The signal from
the center microphone is mixed into both of the stereo channels. The for-
ward location of the center microphone distinguishes the Decca tree from
more conventional AB placement. The location results in the center micro-
phone picking up sounds from the center of the stereo eld slightly earlier
than the side microphones, and this tends to emphasize centrally located
sources due to the haas effect, essentially eliminating the hole in the
middle sometimes exhibited by AB recordings.
Decibel, or dB Literally, one tenth of a bel. The bel is named after Alexan-
der Graham Bell, and the number of bels is dened as the common loga-
rithm of the ratio of two powers. Thus, two powers, one of which is ten
times the other, will differ by 1 bel; 10 watts are 1 bel higher in level than
1 watt. A 360-horsepower car is 1 bel more powerful than a 36-horsepower
motorcycle. Any power ratio may be expressed in bels, and it is important
to note that only power ratios are allowed. A bel is a pure number with no
dimensions.

99
Decibel

The bel had its origin in the Bell Telephone Labs, where workers needed
a convenient way to express power losses in telephone lines as power ratios.
Because the bel is a power ratio of 10, and this is a rather large ratio, it is
convenient to divide it into tenths of bels, or decibels, abbr. dB. Ten dB is
1 bel; thus the decibel is ten times the common log of the ratio of two pow-
ers. The decibel was originally called the transmission unit, or TU, by
the Bell Labs people.
Because of the properties of logarithms, it is easy and convenient to form
some rules of thumb about decibels. The common log of 2 is .301, so a
power ratio of 2 is 3.01 dB, normally taken as 3 dB. Therefore, any two pow-
ers differing by a factor of 2 will be 3 dB apart, and this applies to any type
of power whatever, so long as it is power. Two light bulbs of 100 W and 50
W differ by 3 dB, just as electric motors of 1 horsepower and 2 horsepower
differ by 3 dB. So, any time a power is doubled, it is increased by 3 dB. An
increase of 6 dB represents two doublings of power, or a power ratio of 4.
If an orchestra increases its sound pressure level by 3 dB, its acoustic
power output will be doubled.
Another rule of thumb that is useful to remember is that 10 dB is a power
ratio of 10. Any time a power is increased tenfold, it is increased by 10 dB;
thus, a 200-W power amplier will put out 10 dB more electrical power than
a 20-W amplier, and its sound power output will also increase by 10 dB.
Another way to think of decibels is to think in terms of percentages. We
all know what 10 percent means, and nobody thinks of percentages as being
quantities of anything. A decibel is nothing more than a power change of
27 percent, 3 dB is a power change of 100 percent, and 10 dB is a power
change of 1,000 percent.
Decibels have caused untold confusion among audio people, and most
of this is due to the failure to realize that decibels are not quantities of any-
thing and can represent only power ratios. The trouble starts when we mea-
sure audio signals in volts rather than watts. If we note that power is
proportional to the square of voltage, then a power ratio would be the
same as a ratio of two squared voltages. Then, because the log of a squared
ratio is twice the log of the simple ratio, we can see that the number of deci-
bels is twenty times the log of the voltage ratio between two signals. There-
fore, we can still measure in volts and express power ratios in decibels simply
by multiplying the log of the voltage ratio by 20 instead of 10. Thus,

P V
dB = 10 log10
1 = 20 log10
1
P2 V2

Because we usually measure voltages in audio circuits instead of power,

we need some more rules of thumb for voltage ratios. Here they are: a dou-
bling of voltage is a 6-dB increase, or a power factor of 4. A tenfold increase
in voltage is a 20-dB increase, or a power factor of 100.

100
 Decilog

In the decibel equation, we could replace the power in the denominator

with a xed amount of power. Then, dB would represent the power in rela-
tion to this reference. For instance, if P2 in the equation above is always 1
milliwatt, then the decibel level will represent a specic amount of power.
In other words, 3 dB would be 2 mW, 10 dB would be 10 mW, etc. This is
commonly done, and if the reference power is 1 mW, the measured quan-
tity is in dB(mW); dB(mW) = 10 log P1/.001 W. The reference power could
be any value, and sometimes different values are used, for instance dBW uses
a reference of 1 W. Thus X dBW is X dB above 1 W, and 3 dBW is 2 W, etc.
Decibels are also used in measuring sound pressure. Because sound
pressure squared is proportional to sound power, sound pressure in deci-
bels is equal to 20 times the pressure ratio. If a reference pressure is cho-
sen, then we can measure actual sound levels in decibels, and this is done.
The reference pressure is 20 micropascals (Pa), 1 Pa being a pressure of 1
newton per square meter; 20 Pa is about the threshold of hearing at 1,000
Hz for most people, so it is a convenient reference. In any case, it has been
agreed upon by all international standards. A sound pressure level in deci-
bels is always referred to this standard sound pressure.
The gain of an amplifier commonly is expressed in decibels, and this
can lead to further confusion: an amplier with 3 dB of gain literally means
it doubles the power in the signal, and an amplier with 60 dB gain would
be expected to increase the power one million times. This would be true if
the impedance of the input and output of the amplier were the same, but
this is seldom the case in practice. For instance, a power amplier typically
has a high input impedance and a very low output impedance for
driving a loudspeaker. The true power gain of such an amplier is very
high because the input signal has essentially no load and thus supplies
almost no current or power. This power gain could be correctly expressed
in decibels if the actual input and output powers were taken into consider-
ation. This, however, is rarely done; it works out correctly only if the two
impedances are the same, as is sometimes the case in ampliers used in
broadcasting and some sound systems. The impedance of such units is usu-
ally 600 ohms.
When amplier gains are expressed in decibels, what is usually meant
is voltage gain, and as we have seen, a simple voltage ratio cannot be
expressed in decibels by the denition of the decibel. The formula used to
calculate amplier gain is dB (gain) = 20 log [V(output)/V(input)], and
again, it would be correct only if the impedances at the input and output
were the same. It is small wonder that decibels cause so much confusion!
Decilog, Del, dL A logarithmic ratio of two numbers, dened as 10 times
the common logarithm of the quotient of the numbers, i.e., dL = 10 log A/B.
The decilog differs from the decibel in that the quantities A and B can be
any numbers. Whereas the decibel is dened only as the logarithmic ratio
of two powers, the decilog can express the ratio of any two quantities, such
as the relative sizes of two bank accounts, populations, etc.

101
Decoherer

Decoherer See coherer.

Deconvolution A mathematical process for separating two signals that
have been convolved (combined by the process of convolution). Con-
volution is a mathematical operation involving an innite integral, and it
describes how certain acoustic, electrical, and vibration signals combine in
linear systems. For instance, the human voice is a complex signal consist-
ing of an acoustic input from the vocal cords that is modied by the com-
plex, variable resonances of the vocal track. These resonances accentuate
certain frequency ranges and attenuate others to produce the sounds we
know as vowels. The vocal track modies the input signal by impressing
its own characteristics on that signal. Mathematically, the input signal is
said to be convolved with the impulse response of the vocal track. Once
this convolution is done, it is not an easy task to separate the various con-
tributors to the nal sound, but it is theoretically possible by mathemati-
cal deconvolution. It is also sometimes possible in practice and can be
accompanied by certain postprocessing steps in a dual-channel fft ana-
lyzer. Also, if the input signal is accurately known, the deconvolution can
be done mathematically with a computer program.
It is frequently desirable to measure the characteristics of an audio sys-
tem, such as a sound system in a building, but it is difcult to separate the
effects of the input signal, the system itself, the reverberation of the room,
and the characteristics of the measurement system (microphone). When
the impulse response of the sound systemroom combination can be mea-
sured directly, it is easy to calculate the frequency response, for it is sim-
ply the Fourier transform of the impulse response. The fft analyzer does
this handily. One way to determine the impulse response of the sound
system (or any other audio device) is to input pseudorandom noise, or
a maximum-length sequence signal, and then record the resulting out-
put signal. A deconvolution can then be done on this signal, which gives
the impulse response.
Decoupling Decoupling is the presence of filtering at the power supply
rails of ampliers. The ltering is between gain stages of the amplifier
to prevent some stages from modulating the supply voltage and affect-
ing the amplication of other stages.
Improper decoupling can cause an amplier to be an oscillator, or
at least to be unstable and begin to oscillate under certain signal condi-
tions. Decoupling must be effective at both high and low frequencies to effec-
tively isolate the amplication stages from each other. Decoupling is
usually done with series resistors between the supply rails and the gain
stages and shunt capacitors from the stage to ground, but using voltage
regulator circuits for each stage is another way to do the same thing. Dig-
ital circuit ICs will also commonly have small decoupling capacitors on their
power leads.
Decrescendo A musical term indicating a gradual reduction in loudness.
The opposite of crescendo.
DED Dark Emitting Diode, the inverse of the led. DEDs are said to be used

102
 Deglitcher

by the CIA and other clandestine operatives for power-off indicators in their
electronic equipment.7
De-emphasis The complementary equalization that follows pre-
emphasis is called de-emphasis. The riaa phono equalization curve is
a de-emphasis. See also pre-emphasis.
De-esser A de-esser is a special type of compressor that operates only at
high frequencies, usually above 3 or 4 kilohertz. It is used, especially in
the broadcast industry, to reduce the effect of vocal sibilant sounds, which
are normally too strong when singers and announcers use very close-up
microphones. When the high-frequency energy exceeds a preset thresh-
old, the compressor starts to operate to reduce the high-frequency response.
Low-level high-frequency sounds are not reduced.
De-essers are not used for nonvocal music.
Degausser Degaussers are used to erase tape recordings and to demagnetize
metallic objects such as tape recorder capstans and tape guides and heads;
syn., demagnetizer. Magnetized components such as these will add noise
to tape recordings, although modern tapes have sufcient coercivity that
they are not so sensitive to the effect as were older formulations.
Degaussers consist of a coil of wire wound on an iron core. The coil is
designed to accept the 120 volt line, and this produces a strong magnetic
eld pulsating at 60 times per second, which corresponds to the 60 hertz
line frequency. This pulsating eld alternately magnetizes the part being
degaussed in one direction and then the other. As the degausser is pulled
away from the part, the eld strength gradually diminishes and the part is
left with no net magnetization. There is a battery-operated degausser that
contains an oscillator to supply the pulsating eld. It is used for
degaussing cassette recorder heads.
A tape degausser (or bulk eraser) usually takes the form of a box with
a very large and powerful coil/core assembly inside. Tape reels or cassettes
are passed over the eld and are erased while leaving the eld. These
degaussers are very strong and can erase wide tapes (up to 2 inches), and
come with warnings to avoid use if wearing a pacemaker. They also affect
wristwatches, possibly magnetizing mechanical watches and affecting
their accuracy, or affecting the stepping motor on electric watches. Purely
digital watches are not affected. A tape recorder head degausser takes the
form of a wand where the core extends out to form a tip and the hand grasps
the encapsulated coil.
It is possible to erase tape by passing it over a strong permanent mag-
net. This saturates the magnetization of the tape in one direction, wiping
out any remnant of the recording. This is actually a gausser rather than a
degausser, and it results in a relatively high barkhausen noise level on
the tape.
Deglitcher When a dac is called upon to move from one voltage state

7. Thanks to Dennis Bohn, Rane Corp.

103
Delay Line

to another, it sometimes introduces an undesired excursion due to nonsi-

multaneous operation of the analog switches involved in the conversion.
This is a gross nonlinearity and is called a glitch. An external circuit which
reduces the effect is called a deglitcher.
A deglitcher can be made by following the DAC with a sample and
hold circuit, which disconnects the DAC from its output and holds it at
the previous level while the transition occurs. The sample and hold must
be very fast and have low control feedthrough; in other words, its transi-
tions must not introduce signicant disturbances in the output signal.
Delay Line An audio signal processing device, called a delay line, is used
to simulate the effect of an acoustic echo or to simulate reverberation.
Delay lines are usually digital processing devices, but there exist ana-
log delay lines as well. The rst delay lines were made by using tape
recorders to record a signal while playing it back on the same machine. The
distance between the record and reproduce heads causes a time delay. This
technique is called tape delay.
Delta-Sigma Modulation Delta-sigma modulation is an analog-to-digital
conversion scheme rooted in a design originally proposed in 1946, but not
made practical until 1974 by James C. Candy. Inose and Yasuda coined the
name delta-sigma modulation at the University of Tokyo in 1962, but due
to a misunderstanding, the words were interchanged and taken to be sigma-
delta. Both names are still used for describing this modulator. Character-
ized by oversampling and digital ltering to achieve high performance at
low cost, a delta-sigma A/D thus consists of an analog modulator and a
digital lter. The fundamental principle behind the modulator is that of a
single-bit A/D converter embedded in an analog negative feedback loop
with high open loop gain.8
In the simplest form of delta modulation, only the direction in which
the signal must go to arrive at the next sampled level is encoded. In other
words, if the next sample is higher than the previous one, the code is a 1,
meaning increase in level. If the next sample is lower, the code is a 0, mean-
ing decrease in level. The rate at which the level moves is xed, and the
resulting encoded waveform looks like a series of sawteeth, zigzagging
up and down in an approximation of the signal shape.
It would seem that this is a very inaccurate way to encode a signal, and
this is true unless the sampling frequency is extremely high. In one such
early system, made by the dbx company, the sampling frequency is 700 kilo-
hertz, compared to conventional PCM, which usually uses 44 to 48 kHz.
But the digital words needed for each step contain only 1 bit, rather than
the 16 or more usually required. Thus delta-sigma modulation trades word
size, with its inherent complication, for sampling rate. The bit rate in each
system is about the same, meaning the bandwidth required to record the
encoded signal is nearly the same.
An advantage of delta-sigma encoding is that it provides 1-bit infor-

8. Thanks to Dennis Bohn, Rane Corp.

104
 Derating

mation at a very high rate and in a format that a digital lter can process
to extract higher resolution (such as 20-bits) at a lower sampling rate.
Dematrix To separate signals that have been combined by matrixing. Gen-
erally, dematrixing is actually another form of matrixing.
Demodulation See detection.
Demodulator A demodulator is a device that recovers the audio signal from
a modulated waveform. It is also called a detector. See also amplitude
modulation; frequency modulation.
Depth In stereophonic reproduction of music, depth refers to the appar-
ent distance between the listener and the various instruments in the sonic
image. The perceived depth also has to do with the high-frequency con-
tent of the instrument in question as well as relative time delays between
it and other, nearer instruments. Also important is the perceived size of the
reproduced ensemble, a small group seldom exhibiting a large amount of
depth and vice versa.
Derating The reduction in a theoretical rating, such as power-handling
capacity, due to some environmental factor such as high ambient temper-
ature. For instance, a 1-watt resistor may be derated to .5 watt if it is used
at an elevated temperature.

Detection

105
Desk

Desk British terminology for a recording console, or board; also called a

mixing desk.
Detection, Detector The process of recovering the audio waveform from
a modulated carrier is called detection or demodulation. In the case
of amplitude modulation, simple rectication followed by some
smoothing is all that is needed.
Detection of a frequency modulated waveform is more complex.
Some fm detector types are the ratio detector, slope detector, and frequency
discriminator.
Deutlichkeit Deutlichkeit (pronounced Doyt-lish-kite) is the German word
for clarity, and is a measure of the relative energy in early reverbera-
tion of sound in a room compared to later-occurring reverberation; liter-
ally, the ratio of the sound pressure squared and integrated over the rst
50 milliseconds to the pressure squared averaged over all time. See also
envelopment.
DI See directivity index. See also direct box.
Dial Cord A length of strong string that connects the tuning knob on a radio
or tuner to the station indicating needle and the actual tuning condenser
that selects the station. Dial cord has been used since the introduction of
commercial radio broadcasting, but is becoming obsolete on higher qual-
ity products due to the proliferation of digital tuning controlled by push
buttons and indicated with digital displays.
Dialnorm Dialog Normalization. In digital television broadcasting and
receiving using Dolby audio coding, dialog normalization is a method to
automatically adjust the sound volume of received TV programs so that
the level of speech is relatively uniform from program to program. In the
U.S., the Federal Communications Commission (fcc) mandates that digi-
tal television broadcast stations encode their signals with sound volume
information so that the home television sets can read it and adjust the sound
volume accordingly.
The reason this is needed is illustrated by this example: The dialog of a
typical news broadcast does not require very much headroom because
sound effects that could be much louder than the dialog are generally not
used in such broadcasts. The listener adjusts the volume to a comfortable
level in her living room. Then, if the next program is a motion picture, the
sound track will have music and sound effects that might be momentarily
much louder than the dialog, and the broadcast signal will have to be
reduced in level to avoid excessive loudness. The result is that the dialog
will be too soft, and the listener will have to raise the volume to compen-
sate. Dialnorm adds the headroom information to the sound track so the
Dolby decoder can vary the volume automatically. However, it is suspected
that the Dialnorm system is usually turned off during commercials! See also
metadata.
Diatonic A musical scale of eight notes spanning one octave, consisting
of an ascending pattern of two whole steps, a half-step, three whole steps,
and another half-step. The familiar do re me fa sol la ti do is a diatonic

106
 Differential Amplier

scale. Most Western music is built on diatonic scales, but there are many
other possibilities, such as the pentatonic scale with ve tones in an octave
used in some Chinese music.
DI Box See direct box.
Dichotic Dichotic refers to headphone listening where each ear hears a
different signal, as opposed to diotic, where the same signal is presented
to both ears. binaural listening is an example of dichotic presentation.
These terms are used mostly by researchers into psychoacoustics.
Dielectric An electrical insulating material, as opposed to an electric con-
ductor. The term is most commonly used to mean the material between
the plates of a capacitor, where it serves as a spacer/insulator and
increases the capacitance of the capacitor. The relative amount the capaci-
tance is increased by using a dielectric between the plates rather than by
placing the plates in a vacuum is called the dielectric constant of the mate-
rial, and this can vary from about 2 to over 50 for various materials.
The dielectric strength of a material is the highest electric eld the mate-
rial can stand without breaking down, and this determines the highest volt-
age that a given capacitor will tolerate before shorting.
Difference-Tone IMD See twin-tone.
Differential Amplier Normally, one of the signal input terminals of an
amplifier is connected to the chassis of the amplier; in common parl-
ance, it is grounded. This is sometimes called a single-ended connec-
tion. The amplier is then sensitive to the voltage existing between one
input terminal and ground.
It is possible to build an amplier that has neither input terminal
grounded and is sensitive to the difference in voltage between the two input
terminals. This is called a differential input, and the amplier is called a
differential amplier. Before the advent of integrated circuits, differ-
ential ampliers were made by adding a transformer to the input and
not connecting either side of the primary to ground; this is still done in the
case of some microphone preamplifiers.
Integrated circuit differential ampliers without transformers are com-
monly available. They allow balanced congurations to be easily built
at lower cost than using a good transformer.

Differential Amplier

107
Diffraction

Diffraction The bending of a sound wave around an obstacle and the

reection of a sound wave from a discontinuity in its path are called dif-
fraction. It is wavelength dependent. Where the wavelength is short (rel-
atively high frequencies) compared to the obstacle, reection will occur
as well as bending of the wave front. When the wavelength is long (lower
frequencies) with respect to the obstacle, little reection will occur and the
bending will be more pronounced.
The diffraction of light is analogous to that of sound, although it is sel-
dom seen because of the extremely short wavelengths of light.
Digital See digital audio; Appendix 7.
Digital Acoustics Processor, DAP A consumer audio device that attempts
to simulate the acoustics of an auditorium or other room by adding suit-
able digital time delays and synthetic reverberation to recorded signals. It
is the latest in a long line of analog devices that began with the stereo-
phoner designed by the Viennese orchestra conductor Hermann Scherchen
in the 1950s and the xophonic made by Radio Craftsmen in the U.S. in
the 1960s. The Xophonic had multiple outputs for feeding surround loud-
speakers and came with preprogrammed delay settings to simulate certain
existing concert halls. DAPs are often used to process the stereo audio sig-
nals from prerecorded video tapes and video discs to enhance the illusion
of ambience. Illusion is the proper word, since DAPs do not produce very
accurate simulations of actual acoustical spaces.
Digital Audio The application of digital computer-based technology to the
recording and reproduction of music is somewhat loosely called digital
audio. For an overview of the techniques, see Appendix 7; and for a dis-
cussion of test measurements on digital devices, see Appendix 5.
Digital Audio Broadcasting, DAB As an alternative to am and fm broad-
casting, several types of digital audio broadcasting have been proposed.
One promising system is known as Eureka 147. Its audio quality is said to
be comparable to that of the compact disc, and it does not suffer from
fringe area fading or multipath distortion. It is broadcast by earth-
bound transmitters and requires much less radiated power from the sta-
tion than conventional broadcasting does (only about 1,000 watts, compared
with up to 50,000 watts for AM and up to 100,000 watts for FM), and 16
different programs can be broadcast from a single 1,000-watt station,
greatly reducing hardware costs to the broadcasters. The Eureka 147 sys-
tem was developed by the Eureka 147 Project, an international consortium
of broadcasters, network operators, consumer electronic industries, and
research institutes. The promotion and marketing of the system worldwide
was taken over in the year 2000 by an independent group called WorldDAB.
While the Eureka 147 system has emerged as clearly superior in labo-
ratory and eld tests carried out by CEMA (Consumer Electronics Manu-
facturers Association), the National Association of Broadcasters (nab)
made public on January 23, 2003, its opposition to the adoption of Eureka
147 in the U.S. This opposition is based on lack of new spectrum, dislike of

108
 Digital Signal Processing

sharing transmitters in the multiplex, and concerns that DAB would intro-
duce new competition.
The iBiquity Digital Corporation has now developed a more limited in-
band solution (originally named IBOC, for In Band On Channel, but now
trademarked HD Radio), utilizing existing FM and AM transmitters. See
also iboc.
DAB signals can also be broadcast by satellites, and two such compet-
ing systems named SIRIUS and XM are now in operation.
These systems are not compatible with one another, and each requires
a special receiver to pick up its signals.
Digital Compact Cassette, DCC A consumer recording format announced
by Philips in 1990 that either failed in the marketplace or was never ade-
quately marketed, depending on how you look at it. DCC was an interesting
system designed to compete with the r-dat format and allow compati-
bility with analog cassettes. The DCC was a tape cassette the same size
as a standard analog compact cassette, and its tape speed was the same. It
utilized several linear (as opposed to helical scan in the R-DAT) digital tracks
in each direction to encode the audio signal at a sampling rate of 44.1 kHz,
the same as used in the compact disc. The DCC was quite similar to the
s-dat format.
This system had some interesting characteristics. Since it used the same
sampling rate as the CD, it could be used to make direct digital copies from
CDs without going through digital-to-analog conversion. It contained a ver-
sion of scms to prevent unauthorized proliferation of digital copies. The
system also allowed for the recording and playback of analog cassettes as
well as DCCs on the same machine. A major advantage of the DCC format
was that it allowed high-speed duplication of prerecorded cassettes, which
would have reduced production costs. The system used Precision Adaptive
Sub-band Coding (PASC) to reduce the recorded bit rate compared with the
CD format. The PASC is a European standard for digital radio broadcast-
ing. The dynamic range was said to be 110 dB although the audibility of the
bit rate reduction could be questioned. The DCC has been discontinued.
Digital Dubber A motion picture sound post-production term for a
specialized multitrack audio recorder with removable media used to mix
lm sound tracks. Older analog dubbers usually used 35mm sprocketed
fullcoat.
Digital Equalizer An equalizer that operates in the digital domain; that
is, it operates on signals that have undergone analog-to-digital conversion
and uses digital signal processing techniques. The rst commercially
available unit of this type was the Yamaha DEQ7, introduced in 1987.
Digital Signal Processing, or DSP The manipulation and modication of
signals in the digital domain (after having undergone analog-to-digital con-
version). A great many electronic music instruments use DSP, as do certain
test equipment types such as the fft analyzer. DSP devices have a micro-
processor inside them to do most of the work.

109
Digital Time Delay

Digital Time Delay A digital device that provides an adjustable time

delay. Time delays are used in articial reverberation systems, for spe-
cial echo effects in music recording, and to provide delayed sound to cer-
tain loudspeakers in some sound reinforcement systems. Before the
advent of digital delays, the only way the effect could be achieved was by
tape echo, or by placing a loudspeaker at one end of a long tube and a
microphone at the other. This gives a delay of about 1 millisecond per
foot of length, and it becomes bulky when long delays are needed.
Digital-to-Analog Converter, DAC, D/A The component within a digital
audio device that converts binary digital words into an analog signal
that can then be amplied and sent to a loudspeaker, etc. The DAC is
the last link in the digital chain, just before the anti-imaging filter.
When the compact disc system was rst commercialized, 16-bit
DACs with sufcient linearity for use in high-quality audio were pushing
the state of the art in solid-state electronics. Like the adc, the DAC is
subject to certain inaccuracies, as follows:
1. Absolute accuracy: The difference between the actual analog output
and the correct output expected from a given digital code word.
2. Differential linearity error: The deviation from the ideal output change
of one lsb when the input goes from one value to the adjacent value.
3. Feedthrough: Undesirable signal coupling around switches or other
components that should be turned off.
4. Gain drift: A change in the DACs full-scale voltage output caused
by a change in temperature of the device.
5. Linearity: A measure of the deviation of the actual transfer function
curve from the ideal curve.
6. Offset drift: The actual change in DAC output voltage with variation
in temperature with all bits in the input digital word equal to zero.
7. Settling time: The time required for the DACs output voltage to sta-
bilize to its nal value, measured from the time the digital input code
changes.
Because DACs are costly, many less expensive compact disc machines
use only one DAC which is time-shared between the two stereo chan-
nels. This may cause degradation of sound quality because of the increased
number of glitches in the output signal due to the switching involved.
(The fact that one channel is time-offset by one-half the sampling period
in this scheme is of no audible consequence. It is simply equivalent to mov-
ing one of the loudspeakers by a fraction of an inch.)
Digitization The process of analog-to-digital conversion. See analog-to-
digital converter.
Diminuendo Synonym for decrescendo.
DIN Acronym for Deutsches Institut fr Normung, which is the German
standards organization roughly equivalent to the American nist, for-
merly the National Bureau of Standards. The DIN determines standards
for almost all types of products and procedures for making measurements.
In many cases, the corresponding DIN standard is different from the Amer-

110
 Direct Box

ican standard, and care must be used in comparing specications of audio

equipment specications made in the two countries.
DIN Hub Also known as a CCIR/DIN hub, which is a recording tape reel
style used in Europe. American machines use the nab style hub on their
large tape reels.
Diode A diode is a circuit element that will pass current in one direc-
tion only, from the anode to the cathode.
The rst diode was an Edison light bulb with a second electrode enclosed
in the bulb. Electrons boiled out of the lament and migrated to this plate
when it was connected to the positive pole of a battery. The cats whisker
of the old crystal set radio was a solid-state diode, which served as
a detector of the received radio waves.
Today, most diodes are made of silicon and are used as rectiers to change
the alternating household current to direct current for powering all man-
ner of audio devices. Some diodes, called signal diodes, are very fast-acting,
and are used as detectors for radio signals.
A special type of diode called a zener, or avalanche, diode is designed
to operate in its avalanche mode, in which it conducts current in the oppo-
site direction from normal. An avalanche diode will have a constant volt-
age across it regardless of current through it, and it is used as a voltage
regulator circuit.
Diotic Literally, with two ears. Diotic generally refers to headphone lis-
tening whereby the two ears hear the same signal, as opposed to monotic,
where only one ear hears the signal. See also dichotic.
Dipole With reference to radio or television antennas, a dipole is simply
two metal rods extending out from a common point where they are con-
nected to the transmission line.
In loudspeaker designs, a dipole radiator is a system that radiates for-
ward and rearward with equal energy but with opposite polarity. Exam-
ples of dipole radiators are electrostatic loudspeakers and planar
loudspeakers. Some cone-type speakers have been arranged so they form
dipole radiators. For a dipole radiator to have adequate low-frequency
response for full-range music reproduction, it must be quite large to pre-
vent the rear wave from canceling the front wave. Also, the dipole radia-
tor must not be placed close to and parallel to a wall; it will work best when
not near reecting surfaces.
Direct Box The direct box is an adapter to allow connecting an instrument
pickup, instrument preamplier, or power amplier directly to the mic or
line input of a recording or sound reinforcement console. This avoids hav-
ing to use a microphone for acoustic pickup, or offers a different sound
quality. Electric guitars, basses, and keyboards are typical candidates for
this treatment. Such boxes can offer several attenuation options to accom-
modate this wide range of signal inputs. They can also be passive, usually
with a small transformer, or active, typically with a FET input to min-
imize any loading of a pickup that might affect the sound. Sometimes called
simply a DI (pronounced dee-eye) as in direct input or direct inject.

111
Direct Coupling

Direct Coupling A connection between two devices that allows both direct
current (DC) and alternating current (AC) between them. Direct coupling
is by a simple wire rather than through a series capacitor. See also dc
amplifier.
Direct Current, abbr. DC An electric current that is always in the same
direction, such as the current supplied by a battery, is called direct cur-
rent. In classic electricity theory, the direction of the current was thought
to be from positive to negative. It has since been learned, however, that cur-
rent almost always constitutes a ow of negative charges, or electrons, and
they ow from negative to positive.
Thomas Edison was the great champion of power distribution to homes
via direct current, and he built up such a power distribution network, some
of which is still in use in some eastern cities. It seems he never was con-
vinced that this was a mistakeone of his few lapses of judgment. DC
power has many disadvantages compared to alternating current, one
of the most important of which is that DC voltage cannot be changed by
the use of transformers.
Nicola Tesla, the famous Yugoslavian-born physicist, was convinced very
early that AC was the only practical way to distribute power, and of course
history has proven him correct.
Directivity Factor A measure of the directionality of the sound output from
a loudspeaker. See also q.
Directivity Index, DI A measure of how directional a particular loud-
speaker is, as compared with a completely omnidirectional one. To deter-
mine DI, an omnidirectional loudspeaker connected to an amplifier and
variable-frequency signal source is placed in a free field. The sound
pressure level is measured at a xed distance as a function of frequency.
Then the dut is put in the same place and its input power adjusted so its
sound power output is the same as that of the omnidirectional speaker.
The DUT is aimed directly at the measurement microphone. Because it is
directional, the resulting measured SPL will be higher by a certain number
of decibels, and this number is the directivity index for that loudspeaker
as a function of frequency. Accurate measurement of DI is quite involved
and time-consuming. Certain highly directional speakers, such as high-fre-
quency horns, may have DIs of 15 dB or so. For another denition of loud-
speaker directivity, see also q.
Direct Radiator A loudspeaker that does not have a horn between the
moving element and the air is called a direct radiator. Most loudspeakers
for home use are direct radiators, and most commercial loudspeakers for
sound reinforcement are horn-loaded types. Direct radiators generally pro-
vide smoother, more uniform response, while horns are much more efcient,
providing greater output level for a given power input. Also, horns are more
directional, which is desirable in sound reinforcement systems.
Direct-to-Disc Direct-to-disc is a type of vinyl LP disc mastering in which
a master tape is not used. The signal directly from the control console is
used to cut the original acetate disc. This means a direct-to-disc record-

112
 Dispersion

ing cannot be edited in the sense of making a better performance by splic-

ing several performances together. The recording is made live.
There are audio people who think the tape is a weak link in the audio
chain, and that signicantly better quality can be attained by eliminating
it. However, there are many problems with direct-to-disc mastering.
Increased stress on the musicians is only one. The mastering engineer has
no way of knowing in advance just how loud a performance will be and
must set the volume controls by educated guesswork. A single note that
causes an overload means the disc must be rejected and the whole process
started over. Mastering from a tape is thought to be an art by many, and
direct-to-disc mastering must be more so.
Disc Any disc that stores music or data in optical or mechanical form is
spelled with a c as in compact disc or phonograph disc. See also disk.
Discrete Refers to a 1:1 relationship of recorded tracks on a movie and the
resulting number of speaker channels. For example, a 5.1 digital mix will
be reproduced through six channelsleft-center-right-LFE-left surround-
right surroundin the theater. Obviously each of the surround channels,
and possibly the LFE channel, will have more than one speaker. Discrete
playback is often contrasted with 4:2:4 matrix encoding/decoding.9
Discrete Circuit An electronic circuit that uses individual transistors
as active elements rather than integrated circuits (IC). It is thought by
many that ICs, especially IC operational amplifiers, do not lend them-
selves to high-quality audio applications. They are used extensively
because of their relatively low cost and ease of use. In almost all cases, a
discrete circuit will be superior if well designed.
Discrete Operational Amplier The somewhat poor performance of early
integrated circuit opamps led to the design of an operational ampli-
fier using discrete components. The best known of these is the JE990,
designed by the late Deane Jensen. It has been optimized for audio use and
is generally credited with outperforming all IC opamps and most other dis-
crete amplier designs.
Discriminator One type of detector in an fm radio receiver is the fre-
quency discriminator, a circuit invented by Major Edwin Armstrong,
the originator of FM broadcasting. One disadvantage of the discriminator
is that it is sensitive to amplitude modulation as well as frequency
modulation. It is therefore sensitive to radio interference unless it is pre-
ceded in the circuit by limiters that remove the am from the signal.
The discriminator is actually a specic circuit, but any FM detector is
often called a discriminator.
Disk The spelling disk is used for disks that store data or music mag-
netically as in a computer oppy disk rather than optically. See also disc.
Dispersion Usually refers to the ability of a loudspeaker to radiate sounds
over a wide anglesynonymous with angular coverage. This usage is

9. This entry is copyright 19912003 by Larry Blake and is reprinted with permission.

113
Dissipation

unfortunate because dispersion actually means the variation in speed of

sounds of different frequencies. In air, all frequencies travel at the same speed
(luckily), but some media are dispersive. See also group velocity.
Dissipation The absorption of energy and its conversion to heat, analogous
to damping. loudspeakers are limited in the amount of energy they can
dissipate because of temperature limitations of the voice coil assembly.
Sometimes the descriptive specications for a loudspeaker will include its
maximum power dissipation in watts. The less efcient a loudspeaker is,
the more dissipation it will have. The surrounds of loudspeaker cones,
however, are designed to have relatively high dissipation to help prevent
cone breakup and spurious resonances. As Gilbert Briggs, of the Wharf-
dale Wireless Works, once said, dissipation appears to be better for audio
products than for their producers.
Dissipation Factor Is the reciprocal of the q-value of a capacitor.
Dissonant The opposite of consonant.
Distortion Theoretically, any addition or modication to a signal caused
by any type of equipment could be called distortion, but the term has
come to be somewhat more restricted in its use.
Distortion may be conveniently grouped into six types.
1. nonlinear distortion, manifested as harmonic distortion and
intermodulation distortion. For many years, engineers have strug-
gled to formulate an objective measure of the subjective effects of non-
linear distortion, but this has proved to be a nearly impossible task.
Low-order harmonic distortion, for instance, does not sound unpleasant
to most listeners and in fact adds a certain richness or brightness to
music. Similar percentages of intermodulation, however, are much more
objectionable.
2. frequency distortion, the unequal amplication of different fre-
quencies. This relates to the frequency response of audio devices. It has been
shown that differences in frequency response curves amounting to less
than 1 decibel can be distinguished by the human hearing mechanism. It
is also very important that the frequency response of a device is the same
for all signal levels. Audio transformers are sometimes the cause of
response variation with output level.
3. phase distortion, an effect caused when phase shift in an audio
device is not a linear function of frequency. In other words, different fre-
quencies experience different time delays. This changes the waveform of
the signal, and is especially injurious to transients.
4. transient distortion, including transient intermodulation dis-
tortion. Many experiments have shown that the beginning and ending
transients on musical sounds are largely what determine their timbre, rather
than their harmonic, or overtone, content. An audio device that passes
steady-state signals perfectly well may distort the transients, causing audi-
ble coloration of music. Low transient distortion means a device must
have wide frequency response, no phase distortion, and no hangover.
5. Scale distortion, or volume distortion. Because the human ear has a

114
 Dither

sensitivity that varies with frequency and with loudness level (see
fletcher-munson effect), a musical ensemble must be reproduced at
the same loudness as the listener would experience at the actual event if
frequency distortion is not to occur. For example, if a symphony orchestra
is reproduced at 70 phons loudness level, whereas the concertgoer in a seat
in the auditorium would experience 90 phons, the apparent bass response
will be decreased by about 10 dB. On the other hand, if the reproduced level
is above that of the original, the bass (and to a lesser extent, the extreme
treble) will be too loud.
6. frequency modulation distortion. Examples of this are flutter
and wow, and doppler distortion caused by the motion of loudspeaker
cones.
There are other factors that cause music reproduction to be untrue to
the original, but are not considered distortion. An example is background
noise. Another is lack of directional realism and proper ambience due
to the use of too few channels of reproduction.
There have been many subjective terms used for describing the sound
of reproduced music, some of which are summarized below:

Frequency range
Extreme lows: below 40 hertz
Lows: 40 to 300 Hz
Mid-range: 300 to 4,000 Hz
Highs: 4,000 Hz to 10,000 Hz
Extreme highs: 10,000 Hz to 20,000 Hz

Distortion
Nonlinear: dirty, strident, rough, metallic, harsh
Intermodulation: thick, bassy, fuzzy, wiry, brittle
Noise: frying, sizzle, popping
Transient distortion: hangover, loose, slow

General terms
Source size: live, broad, dead, at, compressed
Detail: veiled, transparent, clear, opaque, focused
Realism: presence, canned, natural, thin, muddy, dead

Distribution Amplier, DA A distribution amplier is a device designed

to feed one signal to many different outputs, such as to a series of headsets
or to several remote audio lines. The DA outputs are usually individual
ampliers all fed from the single input, although gain is usually not a pri-
mary purpose. Instead, the benet is that many outputs can be had with-
out overloading the source, and if an output is short-circuited accidentally,
the source and other outputs are not affected.
Dither In order to reduce the effect of quantization noise and distor-
tion in digital audio systems when the signal level is very low, a

115
Diversity Receiver

broadband random noise with an amplitude of about 12 quantizing step

is added to the signal. This causes the signal to vary in level and to make
the quantizing process more linear, reducing distortion. The dither actu-
ally increases the noise content of the signal, trading wideband noise for
quantization noise and distortion, but the improved accuracy in the quan-
tization results in a subjective improvement in noise performance of the
system. Dither noise is usually white noise, at low level; quantization noise
contains much greater quantities of lower-frequency components and is
much more audible.
To fully attain the advantage of using dither, the signal must be averaged
in time after being converted back to analog. In the case of a low-level
sine wave, averaging over several seconds allows the recovery of a virtu-
ally undistorted signal. Music, however, does not wait for us to do this
averaging, especially with short transient sounds, and the ear in any case
is not able to do a long time-average, so the improvement, while signicant,
is not ideal.
Aside from being a funny sounding word, dither is a wonderfully accu-
rate choice for what is being done. The word dither comes from a twelfth-
century English term meaning to tremble. Today it means to be in a state
of indecisive agitation, or to be nervously undecided in acting or doing,
which, if you think about it, is not a bad description of noise. Dither is one
of lifes many trade-offs. Here the trade-off is between noise and resolu-
tion. Believe it or not, we can introduce dither (a form of noise) and
increase our ability to resolve very small values; values, in fact, smaller than
our smallest bit.10
Diversity Receiver In some wireless microphone systems, a special
receiver is used which has two separate antenna systems. The receiver senses
which antenna can supply the best quality signal, and switches to that
one. If the signal starts to fade in this channel, the receiver will switch very
quickly to the other antenna. As the transmitter is moved around within
the reception area, the receiver constantly evaluates both receiving antenna
signals and switches back and forth as fast as necessary to always receive
the least noisy signal.
Diversity receivers have greatly increased the reliability and quality of
wireless microphone systems, which used to be notorious for picking up
interference from various other transmissions and for picking up noise due
to poor signal levels at the receiver.
Dividing Network See crossover network.
dL See decilog.
DMM A trademark standing for Direct Metal Mastering, which is a process
whereby the phonograph record master is cut into a disc of copper instead
of acetate. The advantage of the process is said to be better high-frequency
response and less noise in the nal records. The copper master is used as

10. Thanks to Dennis Bohn, Rane Corp.

116
 Dolby, Ray

a mother to make stampers, simplifying the normal three-step

processing.
The copper is much harder than the acetate, and the groove cut in it is
more precise. It does not develop horns at the top of the groove, and the
stylus is made with sharper edges. This reduces the distortion that
builds up in the inner grooves of records. Up till now, most metal master-
ing has been done in Germany.
DMM also stands for digital multimeter, which is a multimeter with
a digital readout.
Dolby See compander.
Dolby A See compander.
Dolby Digital Dolbys name for its format for the digital soundtrack sys-
tem for motion picture playback. It utilizes their AC-3 system of lossy dig-
ital compression. When used in 35mm motion pictures, the digital signal
is optically printed between the sprocket holes. Now being found in home
theater installations on dvds, Dolby Digital may use any number of pri-
mary audio delivery and reproduction channels, from 1 to 5, and may
include a separate bass-only effects channel. The designation 5.1 describes
the complete channel format. Surround decoder systems with Dolby Dig-
ital automatically contain Dolby Pro Logic processing to ensure full com-
patibility with the many existing program soundtracks made with Dolby
Surround encoding.
Dolby Digital EX, Dolby Digital Surround EX provides a third
surround channel on Dolby Digital movie soundtracks. The third surround
channel can be decoded at the cinemas or home viewers option for play-
back over surround speakers located behind the seating area, while the left
and right surround channels are reproduced by surround speakers to the
sides. To maintain compatibility, the back surround channel is matrix-
encoded onto the left and right surround channels of an otherwise con-
ventional 5.1 mix, so no information is lost when the lm is played in
conventional 5.1.
Dolby NR Any of Dolby Labs analog noise reduction systems. See
compander.
Dolby, Ray Ray Dolby, founder and Chairman of Dolby Laboratories, Inc,
was born in Portland, Oregon, in 1933. Between 1949 and 1952 he worked
on various audio and instrumentation projects at Ampex Corporation
where, from 1952 to 1957, he was mainly responsible for the development
of the electronic aspects of the Ampex video tape recording system. In 1957
he received a B.S. degree from Stanford University, and upon being awarded
a Marshall Scholarship and a National Science Foundation graduate fel-
lowship, he left Ampex for further study at Cambridge University in
England. He received a Ph.D. degree in physics from Cambridge in 1961.
Dolby is a fellow and past president of the Audio Engineering Society
and a recipient of its Silver and Gold Medal Awards. He is also a fellow of
the British Kinematograph, Sound, and Television Society and an Honorary
Member of the Society of Motion Picture and Television Engineers, which

117
Dolby Stereo

in the past has also awarded him its Samuel L. Warner Memorial Award,
Alexander M. Poniatoff Gold Medal, and Progress Medal. The Academy
of Motion Picture Arts and Sciences voted him a Scientic and Engineer-
ing Award in 1979 and an Oscar in 1989, when he was also presented an
Emmy by the National Academy of Television Arts and Sciences. In 1986
Dolby was made an honorary Ofcer of the Most Excellent Order of the
British Empire (OBE).
In 1997 Dolby received the U.S. National Medal of Technology, the IEEEs
Masaru Ibuka Consumer Electronics Award, and the American Electronic
Associations Medal of Achievement. That year he also received an hon-
orary Doctor of Science degree from Cambridge University, and in 1999 he
was awarded an honorary Doctor of the University degree by the Univer-
sity of York. He holds more than 50 U.S. patents and has written papers on
video tape recording, long wavelength X-ray analysis, and noise reduction.11
Dolby Stereo Dolby Stereo is an analog matrix 4 channel sound system

Dome Tweeter

11. From Dolby Labs. literature.

118
 Double System Sound

for movie theaters that produces three channels of sound in the front (left
and right for music and effects and center for dialog) and a surround chan-
nel for effects. It was introduced in the late 1970s and effectively revolu-
tionized motion picture sound. The lm sound track consists of two optical
channels that have the original 4 channels matrixed into them, somewhat
similar to the QS and SQ matrixing systems used in the quadraphonic
systems sold in the early 1970s.
Dome Tweeter A high-frequency loudspeaker that uses a small hemi-
spherical dome as a radiating surface is called a dome tweeter. It was intro-
duced by Acoustic Research Inc. and has been widely used by others since
the 1960s.
It is characterized by smooth frequency response and relatively wide-
angle dispersion. It suffers from quite low efciency and so is not suited
for large auditorium sound systems. The dome may be made of different
materials such as aluminum, titanium, cardboard, or coated textile and so
may be called hard or soft dome. Each has unique characteristics.
Dome tweeters are often covered with a protective metal grill, which
degrades their performance by reections.
Dongle A small device that must be plugged into a computer communi-
cation port to allow certain software to run. It contains a small amount of
Read Only Memory (ROM) that contains code that authenticates the
legality of the software. It is very effective in preventing software piracy.
Sometimes any small device or adapter with a short cable may be called
a dongle.
Doppler Distortion The Doppler effect, named after a German physicist,
is the apparent change in the pitch of a sound when the source of the sound
is moving with respect to the listener. A car horn sounds higher in pitch as
the car approaches and lower in pitch after the car passes us. If a loud-
speaker cone is reproducing both low and high frequencies at the same
time, the low frequencies will cause the cone to move alternately toward
and away from the listener. While the cone is moving toward the listener,
the high frequency will rise in pitch, and when the cone is moving the
other way, the high frequency will fall in pitch. This is actually a frequency
modulation of the high frequency by the low frequency, and it is called
Doppler distortion.
In cases where small cones are producing the full range of music,
Doppler distortion is audible as a general muddying of the sound, but in
multi-speaker systems, where low, mid, and high frequencies are produced
by different drivers, the effect is very small. Some loudspeaker design-
ers, notably the late Paul Klipsch, are very much concerned about Doppler
distortion, while others seem to pay little attention to it.
Doppler Effect See doppler distortion.
Double System Sound Any of several methods of producing sound
motion pictures where the sound track is recorded on a magnetic tape
recorder that is synchronized with the movement of the lm in the cam-
era. The synchronization is done by recording a special tachometer signal

119
Doublet

on one track of the tape. This synch signal is generated by the camera
and is 60 Hz when the camera runs at 24 frames per second. When the
tape of the sound track is played back, the synch signal, or pilot signal,
is sensed by a device called a resolver, which adjusts the speed of the
tape player to exactly match the speed of the movie projector. In this way
the playback of the track will always be at the correct time to match the
action in the picture. There have been many different types of systems to
do the synchronization, but the one most commonly used until the advent
of crystal synch and the smpte time code system was the Neopi-
lot system invented by Stefan Kudelski, maker of the Nagra tape recorders.
SMPTE time-code synchronization of audio recorders and cameras is
almost universally used today.
Doublet A loudspeaker system with both sides of the transducers open
to the air, such as an electrostatic or planar speaker. Also sometimes called
a dipole radiator. See also loudspeaker.
Double Tracking Originally, double tracking meant the recording of a vocal
track on one tape recorder track, then listening to this while recording
another similar track. The two tracks are combined and re-recorded into a
single track, which will sound more diffuse due to slight differences in the
two original tracks. Double tracking gives a slight chorus effect to voices,
and is frequently used in recording popular music.
In the 1980s double tracking could be done with digital signal proces-
sors, which introduce small randomly varying time delays to one signal
and then combine it with the original signal.
Doublet Response Literally, the imaginary, or quadrature, part of the
impulse response. The impulse response is a time-dependent function
and can be either positive or negative in value. Its phase increases uniformly
with time. Another function can be created by simply adding a 90-degree
phase shift to the impulse response. This process is called the Hilbert trans-
form, after the mathematician David Hilbert, and the new function is called
the doublet response. If the doublet response is squared and added to the
square of the impulse response, and the square root of the sum is extracted,
this quantity will always be positive, and thus it can be expressed on a log-
arithmic, or decibel, scale. This is called the magnitude of the impulse
response and is also known as the energy-time curve. The log conver-
sion allows a much greater dynamic range to be visible on a graph.
Doubling If a loudspeaker is driven too hard in its low-frequency range,
it will produce second harmonic distortion, sometimes with greater
amplitude than the fundamental. This is called frequency doubling,
or simply doubling.
The doubled frequency sounds one octave higher than the fundamental
and is therefore not musically annoying to the casual listener, greatly eas-
ing the task of the loudspeaker designer.
It may also refer to a studio technique of playing a track twice in a mul-
titrack recording and using both in the nal mixdown. One musician can
make a sound like two bass players or two singers, etc.

120
 Drop-Frame Time Code

DPDT Acronym for Double Pole Double Throw. This refers to a switch that
simultaneously routes each of two independent inputs to one of two outputs.
DPST Acronym for Double Pole Single Throw. This refers to a switch that
turns two independent current paths off or on.
Drain One of the terminals of the field effect transistor, the other
two being the source and the gate.
In some audio cables, the shield is in the form of an aluminum-coated
mylar wrap which surrounds the signal conductors. This plastic mate-
rial cannot be effectively connected to other circuit elements, so a bare
wire is enclosed inside the shield for the connection. This is called a drain
wire.
DRAW Acronym for Digital Read After Write (we have also heard Direct
Read After Write), which applies to an erasable compact disc that can
be re-recorded. The CD-RW (Compact DiscRead Write) is the commercial
implementation of the idea. The term DRAW is seldom used except for some
digital tape systems that offer the ability to monitor the actual recording
quality during recording.
Driven Shield A technique by which the shield of an audio cable is not
grounded but rather is connected to an audio voltage that is essentially
the same as the signal being carried by the cable. By driving the shield
at the signal voltage, the effective shunt capacitance between the sig-
nal conductor and the shield is much reduced, allowing much longer lines
to be used without high-frequency attenuation. Driven shields are some-
times used for low-level high-impedance circuits, such as the output of
magnetic tape heads when the amplifier is some distance from the head
assembly.
Driver Individual loudspeakers are often called drivers, especially if they
are in a system where several are used. Also, in the case of a horn-type loud-
speaker, everything except the horn is called the driver.
In an audio amplifier, the stage before the power output stage is called
the driver.
DRM Digital Rights Management. The sale of music is poised for change
as distribution changes from physical formats such as cds to online les
distributed over the internet. Integral to the change is some way to pro-
tect the rights of the music performer or owner, and the needs of the con-
sumer. Digital rights management schemes encoded in the les provide
various ways of reducing unauthorized copying while allowing fair use
by the consumer.
Drop-Frame Time Code This code is a version of the smpte time code
and is used for color video recording. The original time code labeled each
frame of a motion picture or videotape with a precise time. The frame rate
of movies is 24 per second and that of black and white television is 30 per
second in the U.S. With these formats, the time code follows real time; that
is, it runs in synch with the clock on the wall. The introduction of ntsc
color television required the frame rate to be changed to 29.97 frames per
second to prevent the color carrier from interfering with the video and

121
Dropout

sound carriers, which were xed in frequency. This means there are 108
extra frames in a 1-hour color program, and the time code frame count
would add up to 3.6 extra seconds of time in the hour. To avoid this, the
drop-frame time code was introduced to drop two frames at the begin-
ning of each minute, except at the beginning of every tenth minute. This
gets rid of the extra 108 frames, and the frame times run in synch with
the wall clock. The term may be abbreviated DF or DFTC. If a sound
recordist on the set does not properly set his time code for DF or non-
DF, his synch will be audibly off. The editors might be able to x it in post-
production, however.
Dropout In analog magnetic tape recording, the quality of the
recorded signal depends on the uniformity of the magnetic coating of
the tape. If its sensitivity varies from place to place on the tape, the sig-
nal level will be reduced periodically, and these reductions in level are
called dropouts.
A dropout seldom is complete, causing only a few decibels of signal
reduction, but the combined effect of small dropouts is an increased noise
level in the reproduced signal. Newer recording tapes are much more free
from dropouts than older tapes were.
In digital recording, dropouts cause momentary loss of data and would
result in very strong noises in the audio if it were not for error correc-
tion schemes, which either correct the error or make an approximation or
prediction of what the data would look like if they were there. A serious
digital dropout may simply mute the audio.
Dry Lacking in reverberation; dead.
DSD Direct Stream Digital, a proprietary audio codec developed by
Sony and Philips for use in the Super Audio Compact Disc (sacd). DSD
bandwidth is normally 2.8224 Mb per channel (64 times 44.1 kHz), with
optional sampling rates of 32 or 128 times 44.1 kHz. DSD uses a delta-sigma
adc to generate a 2.8224 MHz, 1-bit signal, a rate chosen as a simple mul-
tiple of the lowest common high-fidelity pcm sampling rate, 44.1 kHz.
The 1-bit data stream is recorded directly to disc, inherently improving the
resultant audio quality by doing away with anti-imaging brick wall fil-
ters, and simplifying error correction. Sony claims that the sampling
rate is so high that it more nearly approximates the original analog sig-
nal, allowing equalizers and other effects processors to better simulate
analog effects. A number of dsp algorithms are available that allow the opti-
mization of either bandwidth or dynamic range.
The delta-sigma digital-to-analog converter uses a negative feed-
back loop to accumulate the audio waveform. If the input waveform,
accumulated over one sampling period, rises above the value accumulated
in the negative feedback loop during previous samples, the converter out-
puts a digital 1. If the waveform falls relative to the accumulated value,
a digital 0 is output. As a result, full positive waveforms will be all 1s.
Full negative waveforms will be all 0s. The zero point will be represented
by alternating 1s and 0s. Because the instantaneous amplitude of the ana-

122
 Ducted Port

log waveform is represented by the density of pulses, the method is some-

times called Pulse Density Modulation (PDM).
DSD is probably an updating and renement of the Philips bitstream
technology of the 1980s.
DSP See digital signal processing.
DTMF Abbreviation for Dual-Tone Multifrequency. DTMF techniques are
used for sending coded information over transmission channels. In design-
ing such a system, care must be taken that distortion in the transmission
channel will not interfere with the signals. The frequencies are chosen to
be noncommensurate with each other so harmonics of one tone cannot be
confused with the other tones. The most common example of DTMF sig-
nals are the touch tones generated by modern telephones. The term
Touch Tone was led as a trademark owned by AT&T in 1960, but the
trademark was canceled in 1984. The term is apparently now a trademark
of a series of percussion mallets!
DTS A company, Digital Theater Systems, and their products. DTS was orig-
inally developed for the lm industry and is a 5.1-format theater surround-
sound system that uses 6 discrete channels and a perceptual encoding
scheme for digital surround-sound on a cd-rom interlocked with a time
code to a movie projector. They also have similar systems for laserdisc,
dvd, and cd. A special DTS decoder is always required for playback. A DTS
CD carries 6 channels of digital audio in 5.1 format with 20-bit words at a
44.1 kHz sample rate and a compression ratio of about 3:1. One major dif-
ference of DTS on lm over dolby digital is that the lm itself does not
carry the digital sound track, only a time code. DTS was rst used in 1993
in the lm Jurassic Park.
DTV, Digital Television The standard for digital broadcast television in
the U.S. including hdtv and sdtv, formulated by atsc (Advanced Tele-
vision Systems Committee).
Dual Capstan Some tape recorders have two capstans, one on either side
of the head assembly. This is sometimes called a closed-loop drive, and the
motion of the tape is very well controlled with such a system. Perturba-
tions in speed due to supply reel or take-up reel disturbances are minimized.
See also closed loop tape drive.
Dub A copy of a recorded tape is called a dub, and the process of making
the copy is called dubbing.
Ducker A special device used in paging/PA systems and some radio sta-
tions to reduce the volume of the broadcast signal in the presence of
another signal, usually an announcers voice. The ducker enables the voice
to be heard clearly over the background regardless of the former strength
of the background.
The ducker develops a control voltage from the voice signal and uses
this to control the gain of a voltage-controlled amplifier through which
the background signal is passed. The voice signal is then mixed with the
reduced-level background signal, and the mixture is broadcast.
Ducted Port A modication of the bass reflex loudspeaker system where

123
Dummy Head

a tube, or duct, is placed internally over the hole in the cabinet. The pur-
pose is to reduce the frequency of the helmholtz resonance, allow-
ing a smaller cabinet than a standard bass reex would require. However,
the ducted port has a higher q, i.e., it has less damping, and can therefore
sound more boomy than a conventional bass reex. In some ducted port
systems, the air velocity is quite high, and it can produce audible sound,
called port noise.

Ducted Port Loudspeaker

Dummy Head See binaural recording.

Dummy Load For convenience in power amplifier testing, the loud-
speaker is usually replaced with a noninductive resistor capable of dissi-
pating the full power output of the amplier. This resistor is called a dummy
load. The resistive dummy load is a poor simulation of an actual speaker
load, however, due to the complex impedance of a real speaker.
Dust Cap A circular, usually dome-shaped, piece of lightweight cardboard,
aluminum, or similar material that is glued to the center of loudspeaker
cones to protect the voice coil from dust and iron lings. Note that the
diameter of the dust cap has nothing to do with the diameter of the voice
coil that it covers. Besides covering the area, the cap is also decorative, and
perhaps it contributes to the piston action of the cone. Small dents in the
cap are unlikely to be detrimental and can often be simply pulled out with
sticky tape.
DUT Electronic technicians shorthand for Device Under Test.
DVD Originally an abbreviation for Digital Video Disc, then gradually
changed to Digital Versatile Disc, and now stands for . . . nothing, just DVD.
The DVD has the same dimensions as a regular music cd, but it can hold
about six or more times as much information as a CD. A DVD cannot be
played on a standard audio CD player; however, many DVD players can
play music CDs and other disc formats. Several standards apply to the DVD.

124
 Dynamic Headroom

DVD audio discs can handle digital quantization of 16, 20, or 24 bits at
sampling rates of 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz, 167.4 kHz, and 192
kHz. The DVD is also sometimes called the Multi Media Compact Disc
(MMCD) or the Multi-Media Video File (MMVF). The digital audio on a
DVD movie is dolby digital in the countries conforming to the ntsc
television standard. Variants of the DVD format for data have been intro-
duced into computers.
DX Slang for distance as applied to radio reception. In the early days of
radio, hearing a station more than 20 miles away or so was called DX
reception.
Dynagroove A proprietary system for adding a certain amount of har-
monic distortion to signals recorded on analog stereo records in such
a way that the tracing distortion caused by the playback stylus is effec-
tively cancelled out. It was introduced by RCA, but was not very effective
because of the lack of standardization in playback stylus dimensions.
Dynamic, Dynamics When applied to microphones, headphones, and
loudspeakers, the term dynamic means that the operating principle is
a coil of wire moving in a magnetic eld. It seems strange that the term has
not been applied to moving coil phono cartridges, which are called
moving coil cartridges.
The magnetic eld is nowadays always provided by a permanent mag-
net, but in earlier times, electromagnets were used, especially in loud-
speakers. Such loudspeakers were called electrodynamic.
The term dynamics refers to the changes in loudness in a piece of
music.
Dynamic Filter A dynamic lter is a type of single-pass noise reduction sys-
tem that uses one or two filters whose cutoff frequencies are controlled
by the level of the signal. As the signal level falls during soft passages, the
high-frequency response is reduced (like turning down the treble tone
control), and when the signal level is high, the full bandwidth is restored.
The effective operation of such a system depends on the fact that the
noise will be masked by the signal during loud passages, and this is true
in many, but by no means all, cases.
A key element in the design of dynamic lters is the time constants
during which the bandwidth is changing. If they are too fast, distortion
will result, and if too slow, the noise will be heard to swish in and out as
the signal level changes.
There are several manufacturers making such noise lters. The rst one
to be placed on the commercial market was H. H. Scotts dynaural system.
Dynamic Headroom The dynamic headroom of a power amplifier is a
measure of its ability to handle short bursts of power without overload.
Such strong short signals are common in music.
According to the EIA, it is to be measured as the ratio of the level of a 1-
kilohertz tone lasting 20 milliseconds, which is at the clipping point, to
the average continuous sine wave power output. In other words, it

125
Dynamic Microphone

expresses how much more instantaneous power an amplier will produce

compared to its continuous power, and it is expressed in decibels.
Ampliers vary widely in this ability, and ones with high dynamic headroom
sound louder than similar-sized ones with lower dynamic headroom.
Dynamic Microphone A dynamic microphone consists of a diaphragm with
a coil of wire attached to it such that sound pressure moving the
diaphragm causes the coil to move in a magnetic eld supplied by a per-
manent magnet. Motion of the coil causes an electric current to be
induced in it, and this is the signal output of the microphone. It is similar
to a dynamic loudspeaker operating in reverse.

Dynamic Microphone

A dynamic microphone in its simplest form is a pressure microphone,

and so is omnidirectional in its polar response curve, but more com-
plex types are made with cardioid and super-cardioid patterns.
Dynamic microphones were rst made by the Western Electric company
in the early 1930s, and they had a somewhat irregular response to various
frequencies. They were rapidly improved to take their place among other
high-fidelity types. But a high-quality dynamic microphone is not easy
to make, and some of them are more expensive than quality condenser
microphones.
Because the moving parts in a dynamic microphone are relatively mas-

126
 Dynaural

sive and quite highly damped, this type of microphone does not respond
as readily to transient sounds as do condenser microphones, and the
result is a certain lack of smoothness. Dynamic microphones are quite
rugged and will stand up to a lot of punishment. They are easy to use
because no power supply is needed, as contrasted with most condenser
microphones.
Dynamic Range The dynamic range of a sound is the ratio of the strongest,
or loudest, part to the weakest, or softest, part; it is measured in decibels.
A full orchestra may have a dynamic range of 90 dB, meaning the softest
passages are 90 dB less powerful than the loudest ones. Dynamic range is
a power ratio, and has nothing to do with the absolute level of the sound.
An audio signal also has a dynamic range, which is sometimes con-
fused with signal-to-noise ratio. Rarely is the dynamic range of an
audio system as large as the dynamic range of an orchestra because of sev-
eral factors. The inherent noise of the recording medium determines the
softest possible recorded sound, and the maximum signal capacity of the
system (clipping level) limits the loudest possible sound. Many times an
extremely wide dynamic range is not desirable (e.g., in radio broadcasting
for listening in cars) and broadcasters frequently use compressors and
limiters to reduce the dynamic range of the signals before they are trans-
mitted. This type of signal processing distorts the music in a more or
less noticeable way, symphonic music being the most sensitive to it.
The dynamic range of recorded signals can be increased by volume
expanders, such as the ones made by the dbx company.
Dynaural Dynaural was the trade name for a noise-reduction system
introduced by H. H. Scott in the 1940s. It was a dynamic filter that
reduced the high-frequency response as the signal level became lower.
Then, in loud passages, full response was restored. The theory of opera-
tion is that low-level music does not need as full a range of reproduction
because of the fletcher-munson effect, i.e., our ears become less sen-
sitive to high frequencies as the level is reduced. The loud passages of the
music, which require full range, effectively mask the high-frequency noise.
This type of noise-reduction system is called a dynamic lter, and several
others have been marketed in recent years.
The psychoacoustic phenomenon of louder sounds masking softer
ones is the basis upon which all noise-reduction systems operate. If the
music does not mask the noise during loud passages, then no noise-reduc-
tion scheme will work.
Dynaural was what is today called a single-pass system in that the sig-
nal was processed only after being recorded. Other systems, such as com-
panders, process the signal before and after being recorded in order to
reestablish the proper dynamics to the music. One problem with the
Dynaural was that its high-frequency reduction was all too audible during
softer passages, and the lter could be heard swishing up and down with
the dynamics of the music.

127
Earbuds

E
Earbuds Small, lightweight earphones that t into the concha at the
entrance to the ear canal of the outer ear. They are typically used with
portable cd and mp3 players and radios.
Earphones See headphones.
Earth See ground.
EBU European Broadcasting Union. An international professional society
that, among other things, helps establish audio standards. Active mem-
bership of the EBU is open to broadcasting organizations or groups of such
organizations from a member country of the International Telecommuni-
cation Union (itu) situated in the European broadcasting area.
Echo Commonly used incorrectly to mean reverberation, echo, techni-
cally, is a discrete sound reection arriving at least 50 milliseconds after the
direct sound. It also must be signicantly above the level of the reverber-
ation at that time.
Echo chambers are reverberation rooms, which are carefully designed
to be without echoes. If an actual echo is desired in a recording, a digital
time delay device is used, where the time delay is variable. An analog
tape recorder has sometimes been used to add a time delay, the delay rep-
resenting the time it takes the tape to move between the record and repro-
duce heads. This was called tape echo, and is appropriate usage.
Eddy Currents Eddy currents are induced in electrical conductors by uc-
tuating magnetic elds in the conductors. These currents are localized, and
they circulate in rather small areas of the material. Iron, being a reasonably
good conductor, as well as having a large magnetic permeability, has eddy
currents induced in it when it is used as a core of a transformer. The
eddy currents cause the iron to heat due to its electrical resistance, and this
represents a power lossthe higher the frequency and the more rapid
the changes in the magnetic eld, the greater the loss. To reduce the effect,
the transformer core is made up of many layers or sheets of iron, which are
insulated from one another. This keeps the eddy currents down to small
areas, reducing the power loss from heating and raising the efciency of
the transformer.
Some magnetic materials, such as ferrite, have high resistance, so eddy
currents are small in strength. Ferrite is a very efcient material for trans-
former cores as a result.
There is a type of electric motor, called the eddy current motor, which
uses copper rotors in which eddy currents are induced. The magnetic eld
set up by the eddy currents opposes the magnetic eld that generated them,
causing the rotor to turn. The eddy current motor was used in some early
78-rpm phonographs, where a mechanical governor controlled its speed.
Edge A subjective impression of a certain roughness in the reproduced
sound of a musical instrument. It is usually caused by nonuniform high-
frequency response in the loudspeaker or other audio device.

128
 Editing

Edge Slotting Edge slotting is the cutting of small grooves in the surface
of an analog tape recorder head where tape edges would otherwise con-
tact a ledge caused by wear. This allows tape to always maintain contact
with the head surface in spite of wear or wider than normal tape. As an
analog tape recorder head wears, it creates a trough in the face of the head
surface that is nominally the width of the tape used. The width of tape varies
slightly within a tolerance, and it is possible to get a portion of wider than
normal tape. When wider tape tries to go through the wear trough, the
edges of the tape may ex or the tape may move away from the surface of
the head. Edge slotting alleviates this problem. Inexpensive consumer heads
would never be slotted, but use a felt pressure pad to keep the tape against
the head. Pressure pads cause many other problems, though. Also known
as head slotting.
Edison Cell A type of storage cell developed by Thomas Edison. It uses
nickel and iron electrodes and an alkaline electrolyte consisting of
a solution of potassium and lithium hydroxides. Edison cells are somewhat
complex to build but are very rugged and difcult to damage, and they
have an energy storage capacity much greater than the same weight of lead-
acid cells. However, they have a higher internal resistance, making them
unsuitable when high current is required, such as for starting cars. Edi-
son cells found much use as current sources for railroad switching appli-
cations and also were sometimes used in early radio b batteries.
Edison Effect The ability of a vacuum diode to carry a unidirectional elec-
tric current was originally called the Edison effect, after Thomas Edison,
who rst investigated the phenomenon. In working with his newly invented
light bulb, Edison conducted many experiments to determine why it dark-
ened with use and why the negative end of the lament was always where
it burned out. He noted that if another electrode were introduced into
the bulb, and if it were connected to the positive terminal of a battery, a
unidirectional current was established. Edison could not explain it, nor did
he envision a practical use for it, but, being Edison, he immediately
patented it, calling it an electrical indicator. This patent, issued in 1884,
is considered the worlds rst electronics patent. William Preece, an English
investigator, visited Edison and obtained some of his mysterious bulbs. It
was Preece who coined the term Edison effect. See also fleming valve.
Edit Decision List, EDL When editing a sound or video recording or
motion picture, the various takes are auditioned and a list of the desired
ones is created, along with notes telling exactly where the cuts are to be
made. This could be done manually written on paper, but today a com-
puterized list of the desired takes is created in the proper sequence, along
with information telling exactly where the cuts are to be made and any pro-
cessing to be done when assembling the entire list. The resulting document
is the EDL, which the computer follows to assemble the edited result. Note
that one can control analog recorders with digital edit controllers.
Editing Intercutting of several recordings of a musical selection in order
to make an improved apparent performance.

129
EDL

The advent of magnetic tape as an original medium for making

phonograph records allowed recorded performances to be perfected by
replacing wrong notes and removing extraneous noises, etc. (Editing of a
similar type was possible in the recording of piano rolls, an activity that
many well-known pianists engaged in during the early 1900s.)
The technique for editing analog tape is quite simple: the tape is cut
at the appropriate point, selected by moving it slowly across the head by
hand in order to hear the point, and another section is attached with adhe-
sive splicing tape. If one is careful, undetectable splices may be accom-
plished, and a composite recording can be made with nobody the wiser.
digital editing is essentially the same technique, but the edited version
is a copy made from the takes to be intercut. This is all done electronically
rather than mechanically, as the software arranges the takes according to
the edit decision list. Early digital tape formats contained an analog track
to allow razor blade editing like an analog tape, with electronics attempt-
ing to compensate for the damage to the data stream at the splice point. This
was abandoned as affordable computing power became available. The pre-
cision in the timing of digital edits on some early systems was determined
by the format of the digital signals. In some digital audio tape recorders, the
signal was recorded onto the tape in blocks of data called frames, a video-
like format. There are 30 frames per second in the American (ntsc) stan-
dard system, and 25 frames per second in the European (pal/secam) video
standard. Editing points could most easily occur between frames, meaning
the precision in timing of the edit was 130th or 125th of a second respectively.
Today it is easy to edit at any point on the audio waveform.
Virtually all recorded music is edited, making performing artists infal-
lible in the ears of the record-buying public. (One performer was heard to
murmur, on hearing the playback of the edited master tape for a record he
was making, Wow, I wish I could play like that.)
Editing of any kind puts tremendous stress on musicians who appear
live before an audience as well as on record, and calls for ever more accu-
racy and care in performance. This can, it is contended by many, lead to
dull, mechanical performances, less imbued with spontaneity and musi-
cal exibility. At least one famous critic has said that all the younger pianists
sound very much alike, with nowhere near the individual differences heard
a generation or so ago among a similar group. (No less an old master than
Vladimir Horowitz has said that in any performance he expects to miss sev-
eral notes, explaining that to avoid this, he would have to play very care-
fully, without taking any chances, and that the music would thereby suffer.)
EDL See edit decision list.
EE, E to E A term more often heard in the video industry, it means moni-
toring the signal directly through a recorders electronics (electronic to elec-
tronic) and not after the recording has been made via the confidence
head. Audio people tend to refer to source monitoring as opposed to
tape monitoring during recording.
EE Tape, Extra Efciency Tape A type of improved open reel recording

130
 EIAJ

tape introduced in the 1980s. In a manner similar to the chrome cassette,

EE tape offered higher performance but required a machine with a special
EE equalization setting, 35 microseconds instead of 50. The system has essen-
tially died along with the use of open reels for home recording.
Effects Specialized audio devices found in recording studios, such as rever-
berators, flangers, phasers, and delay lines, are called effects, which
probably is derived from the sound effects of motion picture jargon. Some-
times the abbreviation FX, a shortened version of EFX, is used.
Effects Bus The mixing bus in a recording console used to mix the sig-
nal to be sent to the various effects devices. It is also called the effects
send bus.
Effects Return An input on a mixing console that receives the modied
signal from the effects devices. The effects return inputs usually have
volume controls, or faders, to control the intensity of the particular effect
in use.
Effects Send An output from a mixing console that is connected to the input
of an effects device. The effects send outputs usually have volume con-
trols to set the signal level the effects device will see, and the overall level
of all the effects send outputs may be controlled by an effects master con-
trol operating from the effects bus.
Efciency A measure, usually applied to loudspeakers, of how much of
the input electrical energy is converted to sound energy, expressed in per-
cent. The remaining energy is converted into heat. Most direct radia-
tor loudspeakers are about 1 or 2 percent efcient, while horn-loaded
units may be as much as 10 to 20 or even 30 percent efcient.
High efciency is important in commercial sound systems where high
volume levels are desired in order to reduce the amplifier power output
requirements. Also, low-efciency speakers would be damaged by heat
when driven to high power levels.
Efciency, contrary to some popular opinion, is not related to quality,
or fidelity. In fact, a good case can be made for the assertion that low-
efciency loudspeakers, in general, are more accurate than high-efciency
ones. This is because low efciency implies the existence of more damp-
ing, which absorbs energy and makes response curves smoother and
reduces the effects of resonances.
EFM See eight-to-fourteen modulation.
EFX Short for effects.
EIA Electronic Industries Alliance. Previously called Electronic Industries
Association. A private U.S. organization consisting of manufacturers for the
purpose of agreeing on industry standards in audio and video equipment.
This is the group that standardized the dimensions of the 7-inch tape reel
for consumer tape recorders, among many other things.
EIAJ Electronics Industry Association of Japan. This is the Japanese agency
involved in the setting of standards, not only for audio equipment, but for
video as well. The EIAJ also makes recommendations for test procedures,
and published a set of recommended tests for compact disc players.

131
Eigentone

Eigentone An acoustical resonance or standing wave in a room

caused by parallel wall surfaces, also called a room mode. Any set of par-
allel walls will establish a series of standing waves, the lowest one of which
has the wall spacing as a half-wavelength. A standing wave causes a nonuni-
form sound level distribution in a room, with peaks and dips at various
locations in the room. Thus, two sound-reecting walls spaced 16 feet apart
would give rise to a resonance at about 30 hertz and at all the harmonic
frequencies of 30 Hz. If the surfaces are absorbers of sound, the standing
waves are much less pronounced. The lowest frequency mode between
parallel surfaces will produce a sound pressure maximum at each surface
and a sound pressure minimum halfway between them. If the surfaces are
very reective at that frequency, the minimum will be very low, and the
sound may be almost inaudible at that position.
In a room designed for listening to music, either live or recorded, the
length, width, and height dimensions should not be equal, so that the modes
between different pairs of surfaces will not occur at the same frequencies
and reinforce each other. The relative dimensions should not be commen-
surate; that is, they should not have a simple arithmetic ratio. The golden
ratio, which is about 1:1.62, is sometimes recommended as the ideal pro-
portion for the dimensions of a music room.
In recent years, there has been quite a bit of emphasis placed on the loca-
tion of loudspeakers in rooms to avoid the excitation of certain strong
room modes, and it is true that if placed at a minimum for a particular mode,
very little energy will be drawn from the speaker at that particular modal
frequency. Also, sometimes complex equalization of a sound system will
be attempted to reduce the effect of modes. But the distribution of sound
energy in the room will still be nonuniform, for this is dictated by the modes
themselves, not the location of the sound source. The best way to correct
such a situation is to eliminate the mode rather than relocate or equalize
the loudspeaker.
One way to avoid strong room modes is to avoid smooth hard parallel
surfaces. Because low-frequency modes are the most objectionable, a lis-
tening room for recorded music should have a fair amount of low-frequency
absorption. Another technique for reducing the effects of low-frequency
room modes is active cancellation by the introduction of a sound that is of
opposite polarity to the objectionable room mode sound. This is done by
sensing the sound eld near a boundary of the room with a microphone
and reproducing this signal in phase opposition with a suitable amplier
and loudspeaker.
Eight-to-Fourteen Modulation, EFM The coding of the bits in the com-
pact disc system in order to optimize the process of reading off the disc.
Each 16-bit sample is divided into two 8-bit blocks, which are then
expanded to 14 bits each by the addition of certain zeros and ones. The
purpose of this is to constrain the pattern and thus reduce the dc content
of the code. Three extra bits called merging bits are added at the end of
the 14-bit codetwo to ensure that the code meets the pattern constraint,

132
 Elcaset

and one to minimize the power of the bit stream signal at low frequen-
cies. This procedure forces the pits and lands on the disc to be between
3 bits and 11 bits long, and their repetition rate is more nearly optimized
to the time constant of the laser detector. Thus, 17 channel bits (14 EFM
plus 3 merging) are used instead of the 8 original audio bits for each block.
Eight-Track Cartridge A slim plastic box containing 14-inch magnetic tape
in an endless loop conguration so it could be played continuously if
desired. It had four separate stereo pairs of tracks and a moving head played
any of the four stereo programs. Performance was not very good, and the
endless loop precluded rewinding or fast forwarding, but it was widely used
in cars before the advent of the cassette. It was developed by the Lear com-
pany, of William Powell Lear, the man who invented a business jet.
EIN, Equivalent Input Noise Also called input referred noise. EIN is how
noise is specied on mixing consoles, stand-alone mic preamps and other
signal processing units with mic inputs. The problem in measuring mix-
ing consoles (and all mic preamps) is knowing ahead of time how much
gain is going to be used. The mic stage itself is the dominant noise gener-
ator; therefore, the output noise is almost totally determined by the amount
of gain: turn the gain up, and the output noise goes up accordingly. Thus,
the EIN is the amount of noise added to the input signal. Both are then
amplied to obtain the nal output signal.
For example, say your mixer has an EIN of -130 dBu. This means the
noise is 130 dB below a reference point of 0.775 volts (0 dBu). If your micro-
phone puts out, say, -50 dBu under normal conditions, then the S/N at the
input to the mic preamp is 80 dB (i.e., the added noise is 80 dB below the
input signal). This is uniquely determined by the magnitude of the input
signal and the EIN. From here on out, turning up the gain increases both
the signal and the noise by the same amount.12
Elastomer A synthetic rubbery substance that has a high degree of damp-
ing and high compliance. It is used in phono cartridges to damp vari-
ous mechanical resonances. In the old days natural rubber was used for
mechanical damping of phonograph moving parts, but it gradually hard-
ened over time.
Elcaset Elcaset was a type of audio tape cassette containing 14-inch tape
and running at 334 inches per second tape speed. The standard cassette uses
tape only a little wider than 18 inch and runs at 178 inches per second.
The Elcaset was introduced in the late 1970s as a high-quality alterna-
tive to the compact cassette, and it did have relatively impressive char-
acteristics. However, it was introduced so late that the compact cassette had
already been widely accepted as a music recording medium and had been
developed to the point of very ne performance. The typical consumer was
simply not willing to invest in a different machine and in the higher costs
of the Elcasets, and the system died in the marketplace.

12. Thanks to Dennis Bohn, Rane Corp.

133
Electret Microphone

Electret Microphone A type of condenser microphone that does not

require a polarizing voltage. The electret microphone contains on the back
plate, or inherent in the diaphragm, a material called an electret, which holds
a permanent electric charge. An electret is the electrostatic counterpart
of a permanent magnet. Sometimes electret microphones are called pre-
polarized condenser microphones.
In the electret microphone, there is a slight danger of the charge leak-
ing off the electret due to high humidity or high temperatures. More recent
designs have minimized this problem. If the electret loses its charge, the
sensitivity of the microphone will gradually decrease, reducing the sig-
nal-to-noise ratio.
The impedance of the electret microphone is very high, as in the con-
denser microphone, so there must be a preamplifier close by. This may
consist of a small field effect transistor, which is built into the micro-
phone housing. A small battery may be included to power the preamplier,
or it may operate by phantom power.
Inexpensive electret microphones costing less than a dollar are found in
virtually all consumer electronics with voice input, such as cell phones.
High-quality electret microphones found in recording studios can cost sev-
eral thousand dollars.
Electricity The force causing all natural phenomena not known to be
caused by something else.Ambrose Bierce. See also charge.
Electroacoustics The science that deals with the application of electrical prin-
ciples and apparatus to acoustical phenomena. transducers, such as
microphones, loudspeakers, and phono cartridges, are electro-
acoustic devices.
Electrode A part of an electric circuit that conveys current or volt-
age from one section of the circuit to another. Examples are the plate,
cathode, and grids of a vacuum tube; the base, emitter, and collector
of a transistor; and the carbon rods in an arc light.
Electrodynamic Many years ago, certain dynamic microphones and
loudspeakers used electromagnets rather than permanent magnets to
attain the needed magnetic elds. These devices were thus called electro-
dynamic. The availability of very strong permanent magnet material has
made them obsolete.
Electrolyte Literally, a liquid that can conduct an electric current. It usu-
ally consists of water with certain chemicals dissolved in it. Most batteries
contain a liquid or paste electrolyte, as do most electrolytic capacitors.
Electrolytic Capacitor An electrolytic capacitor is a type of capacitor that
must be polarized in order to operate correctly. The name comes from the
fact that an electrolyte is used as one of the conductors. In other words,
it must have a direct voltage placed across it, and moreover, this voltage
must be of the correct direction. An electrolytic capacitor will have a very
high impedance when a positive voltage is placed on it, but will have a
relatively low impedance when this voltage is reversed. It acts somewhat
like a diode and is thus nonlinear.

134
 Electromagnetism

The electrolytic capacitor has a much greater capacitance for its size
and cost than a conventional capacitor and is therefore desirable for use
in audio circuits. As long as it has the proper voltage across it (i.e., it is
properly biased), it works correctly in audio circuits, but if the bias is
lost, the capacitor will cause harmonic distortion in the signal it is
passing.
Electrolytic capacitors are used extensively in audio power amplifiers,
especially for power supply filtering, where their very large values of
capacitance are needed. Small ones are also much used as coupling capac-
itors between the stages of ampliers.
There are available today so-called bipolar electrolytics, which have two
capacitors connected in series back-to-back in a single housing. They do
not need to be polarized and are often used in loudspeaker crossover
networks.
Electrolytic Detector The rst diode detector, invented by the American
engineer Reginald Fessenden around 1901. Fessendens detector used a
small aluminum cup containing a mixture of acid and water, and a small
silver wire that dipped into the acid. This arrangement allows current
to pass in one direction only and thus can be used to rectify, or detect,
the received radio signal. Dr. Lee DeForest also made an electrolytic detec-
tor for his wireless system, and there ensued a legal battle that was nally
resolved in Fessendens favor. All this predates the fleming valve of 1904,
which was the rst vacuum diode and was also used as a detector.
Electromagnetic Compatibility, or EMC Audio equipment that is designed
to be immune to electromagnetic interference is said to be electro-
magnetically compatible.
Electromagnetic Interference, or EMI A great many of the unwanted
noises heard in audio systems, such as hum, static, and buzz, are caused
by electromagnetic waves that are picked up and amplied by the audio
system. Common sources of EMI radiation are uorescent lamps, power
lines, computers, automobile ignition systems, solid-state light dimmers,
am and fm radio transmitters, and television transmitters. EMI caused by
very high frequency transmissions is called radio frequency inter-
ference (RFI) and television interference (TVI).
EMI can be a major problem under certain circumstances and is usually
very difcult to eliminate from audio equipment that is not designed to be
immune to it in the rst place. Properly designed equipment is said to pos-
sess electromagnetic compatibility (EMC).
Methods of controlling EMI include shielding of audio wiring and
devices, grounding and elimination of ground loops, balancing of audio
circuits and twisting of the wires in a balanced line, transformer cou-
pling, and the use of bypass capacitors and ferrite beads in low-level
circuits.
Electromagnetism The science of the interactions between electrical and
magnetic phenomena.
In the early days of physics, it was not known that there was any con-

135
Electron

nection between electricity and magnetism, and they were studied inde-
pendently. It was the brilliant Scottish physicist James Clerk Maxwell who
in 1873 rst postulated the quantitative relationship between electricity and
magnetism, and predicted the existence of electromagnetic waves. This the-
oretical understanding brought about by Maxwells famous equations led
to the development of all the electrical and electronic equipment we know
today. See also magnetism.
Electron The elementary particle that carries the smallest unit of electric
charge is the electron. An electric current normally consists of a ow
of electrons, although in some cases, a current could consist of a ow of
positive charges. In a vacuum tube, electrons are boiled off the hot cath-
ode and are attracted by the positively charged plate. They constitute the
current through the tube.
The theory of electrostatics says that like charges repel one another and
unlike charges attract. Any charge has an electric eld around it, and another
charge in this eld will feel a force. Electric elds in tubes or the semi-
conductors of transistors cause the charges to drift one way or the
other, and this constitutes the electric current. In transistors, the current is
sometimes thought of as if it consisted of positive charges, but normally it
consists of electrons.
Electronic Architecture The physical design of the various interconnections
of the components in a discrete or integrated circuit are sometimes
called its architecture. See also topology.
Electronic Crossover See active crossover.
Electrosonic See intensity stereo.
Elliptical Equalizer A special equalizer that causes the two channels of
a stereo signal to be more nearly in phase at low frequencies. The pur-
pose is to make the signal easier to cut onto a record.
Large out-of-phase, low-frequency stereo signals result in large vertical
motion of the recording stylus, sometimes to the extent that the stylus will
leave the surface of the record. This means the discontinuity in the groove
cannot be tracked by the playback stylus. The elliptical equalizer actually
works by introducing crosstalk between the channels at low frequen-
cies. Making the signals in phase only at the low end does not appreciably
reduce the apparent stereo separation, for our ears rely most on high-fre-
quency cues for the stereo effect.
The name comes from the path of the cutting stylus, which is more nearly
conned to the form of an ellipse than a circle. See also mono compatibility.
Elliptical Filter A multiple-element low-pass filter that has the steep-
est possible cutoff slope and a small amount of ripple in the passband.
Its name comes from the fact that elliptical functions are used in its design.
The elliptical lter is a low-pass or bandpass lter with one or more
notch lters added to it. The rst notch is just a little above the cutoff
frequency, reducing the response in this area. At higher frequencies
where the notch lter has less attenuation, another notch lter is
added, etc.

136
 Energy Time Curve

Elliptical lters are excellent for use as anti-aliasing filters in dig-

ital audio devices, but they are complex and difcult to design.
EMC See electromagnetic compatibility.
EMF Electromotive Force (EMF) is synonymous with voltage. The term
back emf is the usual usage, and it indicates a voltage induced in a voice
coil that is in opposition to the voltage applied externally to the coil. The
back EMF is a result of the motion of the coil in the magnetic eld of the
loudspeaker.
EMI See electromagnetic interference. Also Electric and Musical
Industries, the diverse British recording company formed by a merger of
the Gramophone Company and Columbia Graphophone (sic) in 1931. In
the same year, EMI opened the famous Abbey Road recording studio in
London, and the chief EMI scientist, Alan Dower blumlein patented a
means of recording stereophonic sound. This remarkable patent foresaw a
great deal of future products, such as stereo records using a single groove
with the stylus recording and playing the left and right channels by two
separate motions angled 90 degrees apartthe same basic system as was
used worldwide for stereo LP vinyl records rst introduced in 1958 by EMI.
Blumlein also foresaw the use of stereo sound for motion pictures, and a
method for realizing it is covered in his 1931 patent.
The Bell Laboratories in the U.S. were also experimenting with stereo
recording at the time, but there is no documentation of any communica-
tion between them and EMI.
Emitter Follower A circuit containing a bipolar transistor in such
a conguration that its voltage gain is 1. It is a solid-state version of
the cathode follower. Emitter followers have very high input imped-
ance and very low output impedance, making them suited to driving
fairly long audio lines. The emitter follower has a large amount of cur-
rent gain and can become unstable unless care is taken in its design. See
also source follower.
EMT Plate See plate reverberation.
End-Addressed, End-Fired A microphone that is positioned so its end
is pointed toward the sound source, as opposed to a side-addressed micro-
phone, which is positioned with its side toward the sound source. Condenser
microphones may be either end-red or side-red, and ribbon mics are
almost always side-red. Generally, what is considered the end or side is
merely a function of the shape of the outer housing. The transducer itself,
such as a diaphram, points toward the sound source.
Energy Time Curve The energy time curve is the envelope of the impulse
response of a device, usually expressed as the common logarithm of the
magnitude of the envelope.
If the impulse response of an audio device is examined directly, it is
not possible to see any detail where the response is at very low values
because it is displayed on a linear amplitude scale. This must be so because
the impulse response of any device has both positive and negative values,
and negative values cannot be represented on a logarithmic scale.

137
Enhancer

The envelope of the impulse response, however, is always positive, and

it can be expressed on a log or decibel scale. In this way, it is possible to
see details at low levels that are invisible in the impulse response itself.
It is instructive to investigate the energy time curve of such devices as
loudspeakers, microphones, and entire sound-reinforcement systems,
including the rooms in which they are placed. Special care must be taken
in making energy time curve measurements, especially if FFT techniques
are used, or large errors can occur.
Enhancer One of any signal processing devices for use in recordings and
performance to add special effects. An example is the aphex aural
exciter.
In general, enhancers work by adding certain types and amounts of har-
monic distortion. The type of distortion (even-order harmonics or odd-
order harmonics) added affects the subjective character of the resulting
sound. It is interesting that harmonic distortion does not sound unpleas-
ant unless it is present in relatively large amounts. This is because the har-
monics are in tune with the music itself. This is not the case with some
other types of distortion such as intermodulation, which sounds harsh and
unpleasant.
Envelope See adsr.
Envelope Delay Another name for group delay.
Envelope Follower A device used in electronic music synthesis that will
convert the envelope of a musical signal into a control voltage. That
is, the output voltage will be low when the signal is soft and high when
the signal is loud. This control voltage can then be used to control any num-
ber of parameters in the synthesizer.
Envelope Generator See adsr.
Envelopment Envelopment is the characteristic of an auditorium that
causes the listener to be surrounded or enveloped by sound, and it has
been shown to be probably the most important single attribute of good
acoustics. It is related to deutlichkeit, a term originated in Germany.
A sense of envelopment in a room is caused by the existence of fairly
early reections (within 20 or 40 milliseconds) of sound from the side walls.
Early reections from the ceiling do not provide the same benet, for they
reinforce the direct rather than the reverberant sound, and sometimes give
the room a somewhat dry and clinical sound. It has been shown that audi-
ences greatly prefer the enveloping sound to come from fairly large angles
from straight ahead. This means the room should be designed so the wall
reections move across the room, rather than towards the rear. A room with
good envelopment sounds warm and agreeable, even if the reverber-
ation is lacking. On the other hand, a room with sufcient reverberation
but lacking in envelopment is likely to sound mushy or blurred and
indistinct. One objective measure of envelopment is the interaural
cross-correlation, or iacc, which is a measure of the differences in
the sounds reaching the two ears. A relatively large value of IACC corre-
sponds to a large subjective impression of envelopment.

138
 Equalizer, Equalization

Epoxy Patent The epoxy patent is a clever method of avoiding the piracy
of proprietary circuit designs. When someone designs a particularly inno-
vative electronic circuit, it is likely that it will be copied by other design-
ers. To protect the circuit against this, it can be totally encased in opaque
epoxy plastic. Then, a would-be pirate would have to destroy the assem-
bly completely in order to get it apart. This procedure is called an epoxy
patent in the vernacular.
EQ See equalizer.
Equalizer, Equalization An equalizer, contrary to what its name implies,
alters or distorts the relative strength of certain frequency ranges of an
audio signal. In a sense, it should probably be called an unequalizer.
However, the rst equalizers were used to make the energy at all frequen-
cies equal, or to achieve at response, in telephone lines, and this is where
the term originated. Another early use of equalizers was in the motion pic-
ture sound industry, where they were used to improve intelligibility in lm
sound tracks. Later on, equalizers were found useful for creating special
sound effects in the early days of radio and movies, where they are exten-
sively used to this day. All equalizers are made up of various circuits
called filters, which are frequency-selective networks containing resis-
tors (R), capacitors (C), and inductors (L). Normally, lters attenu-
ate certain frequency ranges and do not boost them; however, some
equalizers that boost the signal are called lters.
The rst consumer-type equalizers were the tone controls on radios, the
rst one of which was simply a variable low-pass lter to reduce static
and other high-frequency noise. The familiar bass and treble tone con-
trols came later. Today, equalizers are also used in vast numbers in sound
reinforcement systems and in recording and broadcast studios for var-
ious purposes.
An equalizer can boost or attenuate a certain frequency band, but in com-
mon usage, equalize means to boost. The preferred terminology for the actual
process is boost/cut rather than equalize/attenuate. In Britain preferred
usage is lift/dip.
Equalizers that can have peaks in their response curves (such as para-
metric and graphic equalizers) are characterized by the relative sharpness
of the peaks. The q of a lter is a measure of this sharpness and is dened
as the center frequency divided by the half-power bandwidth. For
instance, a one-third octave lter centered at 1,000 hertz will be 232 Hz
wide at its half-power points. Its Q is thus 1,000/232, or 4.31. Filters with
Q values much higher than this tend to ring, distorting transients, and
call attention to themselves when used in sound systems. See also q.
Some equalizers, such as 13-octave graphics, have a constant Q regard-
less of the amount of boost or cut, and they are sometimes called constant
Q equalizers. Most parametrics, on the other hand, have higher Q values
with high boost or cut levels. These are called proportional Q equalizers.
All equalizers cause the signal to be selectively phase shifted, and the
more sharply dened the equalization, the steeper the phase shift curve will

139
Equalizer, Equalization

be. For instance, an octave-band lter will have about a 90-degree phase
shift spread over the octave band, while a 110-octave lter will have the same
90-degree phase shift occurring in the narrow 110-octave band. These
phase shifts distort transient waveforms, for they really are variable time
delays for different frequencies. In general, the more rapidly the phase is
changing as a function of frequency, the more audible the effect is.
Many equalizers are integral parts of audio components and are not
adjustable by the user. Such equalizers include de-emphasis, pre-
emphasis, riaa equalization, shelving, and nab equalization.
Many other equalizers exist as stand-alone devices, and they can be roughly
grouped as follows:
Active and passive equalizers: An active equalizer requires power to oper-
ate. It has active components in it such as transistors and integrated
circuits. A passive equalizer does not require any power to operate. Pas-
sive equalizers contain resistors, inductors, and capacitors, but no active
components. Passive equalizers are essentially noiseless in operation and
are very reliable and distortion-free, but they have insertion loss, which
sometimes has to be compensated for by an amplier. Active equalizers
are very popular but suffer from noise, reduced dynamic range, and sus-
ceptibility to rfi. Many of the following types of equalizers exist as active
or passive.
Rotary equalizer: An adjustable equalizer with several frequency bands
with stepped rotary knobs to select the frequency range and the degree of
boost or cut. The frequency bands are xed.
Parametric equalizer: Somewhat like the rotary equalizer but with added
control of the center frequencies and bandwidths of the frequency bands.
Boost and cut and the other parameters are generally continuously vari-
able rather than being stepped. An equalizer that does not allow variable
control of all the parameters is sometimes called a quasi-parametric, and
some units with slide controls rather than rotary are called para-graphic.
Some parametrics allow cut only and are then called notch equalizers, band-
reject equalizers, or cut-only equalizers. The Q values of most parametric
equalizers increase at high boost or cut levels. Sometimes these equalizers
are called proportional Q equalizers.
Graphic equalizer: A multiband variable equalizer that consists of a series
of parallel lters, usually of the same bandwidth, that are capable of boost-
ing the signal or cutting it. Each lter band is controlled by a slider knob,
and these are arranged on the front panel of the equalizer so their positions
show, at least approximately, the overall response curve for each setting.
The band center frequencies are xed for each band and are spaced pro-
portionally to the logarithm of frequency. In other words, each lter con-
trols a frequency span encompassing the same musical interval controlled
by every other lter, such as octave bands or one-third octave bands. Some
graphic equalizers do not have all the lter intervals equal, having the
low-frequency bands narrower than the high-frequency ones. This makes
a good deal of sense for applications such as the equalization of sound-

140
 Equivalent Circuit

reproducing systems in listening rooms, where low-frequency anomalies

are generally more numerous than high-frequency anomalies.
Probably the most common graphic equalizer has one-third octave band-
widths. It has been shown that lters narrower than this, especially when
used to boost the signal, are quite audible because the lters themselves
ring. The ringing is less audible for narrowband notch lters.
One-third octave graphic equalizers have a constant Q regardless of the
amount of boost or cut. They are sometimes called constant Q equalizers.
If all the knobs on a graphic equalizer are at the same level, one would
assume the overall frequency response would be at, but this is usually not
the case. With most such units, a at curve can be attained only with the
knobs set for no boost or cut. A level boost or cut causes ripple, or peaks
or dips at each lter center frequency. For this reason, any graphic equal-
izer should be operated with the controls as close to the zero boost line as
possible.
Transversal equalizer: A special and uncommon type of equalizer that uses
a tapped time-delay line as the active element. A signal fed through a time
delay and then connected back to the delay input and recirculated through
it many times will have peaks and dips in its frequency response curve
depending on the delay time and the loop gain of the delay line. By select-
ing many different delay times and judiciously connecting the various feed-
back loops, many different frequency response curves may be attained. The
transversal equalizer is an analog, not a digital, device and can use a bucket
brigade, charge-coupled device, or an all-pass lter as a delay line. A sim-
ilar result can be attained with digitized signals. See digital equalizer.
Narrowband lter: Early in the days of sound reinforcement system
equalization, very narrow notch lters were used in an attempt to reduce
the tendency for acoustic feedback. The technique was pioneered by
C. P. Boner. He used very narrow lters, sometimes as little as 5-Hz wide
at the 3dB points, and he used as many as 150 of them, all custom-tuned,
at once. It was later realized that the narrowband technique is much more
complex than need be and that similar results, without doing as much vio-
lence to the signal quality, can be achieved by one-third octave lters.
In the digital domain, including the use of daws, all the types of equal-
ization are accomplished by mathematical manipulation of the digitized
audio signal, in many cases at much less expense than using analog devices.
Digital editing can be more sophisticated than analog editing; for instance,
phase shift problems can be tamed by the use of Finite Impulse Response
(FIR) ltering which can be implemented to have no relative phase shift as
a function of frequency. However, FIR lters do have a time delay that must
be compensated for in the overall equalization setup.
Equivalent Circuit An electrical circuit that behaves like an analog of a
mechanical or acoustic system is called an equivalent circuit. It is interest-
ing that mechanical quantities such as mass and stiffness can be modeled
with electrical quantities such as inductance and capacitance. Theoretically,
any mechanical system could be represented by an appropriately designed

141
Erase Head

circuit. The important thing is that the behavior of electric circuits is well
predicted mathematically, and the equations can be used to analyze the
mechanical system. Generally it is easier to construct and analyze an elec-
trical analog of a mechanical system than to analyze the mechanical sys-
tem directly.
Erase Head A tape head that is used to erase, or demagnetize, the tape
before it reaches the record head in an analog audio or video tape
recorder. Some audio recorders allowed selective activation of individual
erase tracks, some merely erased the entire width of the tape. If erasing the
entire tape is acceptable, bulk erasing is usually more efcient and can leave
less noise on the tape, but to use this benet would require turning the erase
head off, which is possible but seldom done. Instead, bulk erasing ensures
an entirely blank tape, and users put up with the slight increase in erasure
noise from the erase head. Note that the r-dat needs no erase headdata
is simply overwritten. See also bulk eraser, degausser.
Error Concealment A technique to reduce the audible effect of a digital error
in a digital audio system when the error cannot be corrected by the tech-
niques of error correction. Error concealment usually consists of mak-
ing a smooth transition from the last good data block before the error to
the rst good data block after the error.
Error Correction In digital audio systems, the sampled amplitudes of
the signal waveform are expressed by digital codes in the binary num-
ber system. The codes are grouped into words of eight binary digits (bits)
each. If in the transmission of the digital words some bits are missing, or are
incorrect due to tape dropouts, etc., the result will be gross distortion of
that portion of the signal when it is reconstructed. Therefore, it is extremely
important to provide detection and correction of such digital errors.
In a digital system, error detection is possible if the number of allowed
words is less than the maximum. For example, in a two-bit transmission
system we might allow only 00 and 11 as possible words. If we receive 01
or 10 at the end of the transmission chain, we know there is an error. We
cant correct it, however, because we do not know whether a 00 or 11 was
intended.
Things are a little different in a three-bit system. Suppose we allow 000
and 111 as valid words. Then, we can say a received 110 should probably
be a 111, and a 100 should have been a 000. Thus we can make the substi-
tutions and correct the information. To enable this, we have added one bit
per word, and this is called a parity check bit.
compact discs are produced with two simultaneous schemes of
adding parity bits for error detection and correction. In this system, one set
of audio samples is grouped into 32 eight-bit words (or bytes), four of which
are added parity bytes. The second code has a word length of 28 bytes of
which four are parity bytes.
Although it is theoretically possible to correct up to four errors per word
by fully utilizing this method, it is not done in practice because of the very
large amount of hardware required to do it. Instead of error correction, the

142
 Fantasound

words with detected errors are replaced by words having intermediate

values between the two adjacent correct words. In other words, an inter-
polation scheme is used. If two or more adjacent words are incorrect, the
system simply mutes those words, effectively disconnecting the signal
for that instant. It is surprising how tolerant the ear is to the presence of
this mutingif it occurs infrequently and randomly, it is inaudible to most
listeners.
This procedure of detection, interpolation, and muting might be called
error concealment rather than error correction.
ET See transcription.
Etching The process by which printed circuit boards are made. The
boards are covered with copper, and the areas where the conductors will
be are covered with an acid-resistant material. Then the board is immersed
in an acid bath which etches away the uncovered copper, leaving the con-
ducting paths intact.
Euroblock Connector Short for European style terminal blocks, which are
pluggable multiconductor connectors with screw terminals for holding the
wires.
Exciter Lamp See sound head.
Exponential Horn See horn.

F
Fade To gradually change the volume, or loudness level, of an audio sig-
nal. Fades are often done to end recordings of popular music. Also called
a fade out.
Fade In An inverse fade, starting from inaudibility and rising to full vol-
ume.
Fade Out A fade out is the gradual reduction to zero of the volume near
the end of a recording, often used to end popular songs.
Fader Another name for a variable attenuator, or volume control. pan-
pots are also sometimes called faders, used for fading a signal from
one channel to another.
Fanning Strip See terminal strip.
Fanout An end of an audio snake that has all of the loose, separate con-
ductors and connectors. Fanout also refers to how many electrical devices
a given device can effectively drive, or operate. An example is a digital logic
gate which can drive 4 other logic gates, has a fanout of 4.
Fantasound An early type of stereophonic motion picture sound devel-
oped jointly by RCA and the Walt Disney Studio for the Disney movie Fan-
tasia, which opened in 1940. Fantasound is often credited as the rst com-
mercial stereophonic motion picture sound system, but the Bell Telephone
Laboratories had publicly demonstrated a similar system in 1937 in New York
and again on a much larger scale at the 1939 New York Worlds Fair, where

143
Faraday Cage

it accompanied a stereoscopic movie produced by John Norling. It was said

that ve million people saw and heard this presentation! See also ssfs.
The original recording in Fantasound was accomplished using eight
channels of optical recording on two 35-mm lms. The rst six channels
accepted the signals from individual microphones placed strategically
in the orchestra, and the seventh channel was a mix of the rst six. The eighth
channel was connected to a single microphone placed in the hall for a dis-
tant, or reverberant, pickup. For reproduction in the presentation theater,
three loudspeaker systems were placed behind the screen, and surround
speakers were placed in the sidewalls and rear of the theater. In the rst
road show presentations of Fantasia, the signals from the eight optical tracks
were mixed by an experienced operator into the theaters four reproduc-
tion channels. This entailed a tremendous amount of skill and equipment
and was a very expensive undertaking. Shortly after the initial release, a
simplied system was utilized that had a single separate 35-mm lm with
four optical sound tracks derived from the original eight and a fth track
that acted as a control for the dynamics and placement of the four sound
tracks. This control system was called togad, for Tone-Operated Gain-
Adjusting Device. The TOGAD was the predecessor of the automated mix-
down systems found in modern recording studios.
Even with this simplication, Fantasound was found to be too expen-
sive in practice, and what with the complications encountered with the
beginning of World War II, it was not used after 1941. The name Fantasound
is still a trademark owned by the Disney Studio.
Faraday Cage An electrostatic shield is sometimes called a Faraday cage,
after Michael Faraday, the English scientist who rst described its action.
Sometimes it is also called a Faraday shield. An example is the braided cop-
per shield over a microphone cable. The better the conductor (in other
words, the less resistance it has), the more effective a shield it will be.
In most audio transformers there is a Faraday shield between the
primary and secondary windings in order to prevent capacitive coupling
between them. (See capacitance.) This reduces extraneous noise such as
hum and increases common-mode rejection.
Far Field If a sound source is operating in an enclosure or room, the sound
pressure level will vary with the distance that the measuring micro-
phone is from the source. At certain close ranges, the levels will obey the
inverse square law, and at these distances there will exist approximately
a free field. At greater distances the reduction in measured level with
increased distance will be less than predicted by inverse square until
nally a region is reached where the level is constant regardless of distance;
this is called the reverberant field.
The area between the free eld and the reverberant eld is called the far
eld. Its extent is a characteristic of the directionality of the sound source
as well as of the acoustics of the room. See also critical distance.
Far-Out Field When measuring sound pressure level (SPL) in an enclosed
space, it will be found that the measured levels will increase in value near

144
 Feedback

the boundaries of the space. The increase will generally be 6 dB at the bound-
ary surface, and will decline to the reverberant level at a distance of one-
half wavelength of the sound. The reason for this is that the sound pressure
reected from the boundary adds to the incoming sound pressure, dou-
bling it at the boundary. The double pressure is a 6 dB rise in level. The
region in which this effect occurs is called the far-out eld, and its extent
depends on the frequency of the sound in question. It is important to take
care when measuring sound levels indoors to avoid taking measurements
in the far-out eld. See also far field and reverberant field.
Fat Wire System In telephone systems before the advent of digital multi-
plexing, where several incoming lines were available at one subscriber loca-
tion, the switching between the lines was done in the telephone itself, usually
by push buttons. This required that all the lines be routed to each phone
through a 25-pair cable. This is called a fat wire system, as opposed to a
slim wire system, in which the switching is done at a central location and
each phone has only three or four pairs of wires running to it. Today, it seems
that virtually all multiline systems in ofces use digital multiplexing.
Faulkner Microphones A pair of gure-8 pattern microphones about 8
inches apart and both facing the same direction. It would seem like the idea
was a modication of the classic x-y stereo pickup using coincident gure
8s angled 90 degrees apart, but the only difference between the two signals
is due to the time difference between a sounds arrival at the two micro-
phones. We suspect two omnidirectional microphones in the same con-
guration would sound much better because at least the deep bass would
be natural-sounding, and off-axis sounds would not be attenuated.
FCC The Federal Communications Commission is the government agency
that regulates all radio and television transmissions in the United States.
Broadcast stations are licensed by the FCC, including hobbyist or ham
stations, and the FCC maintains surveillance of all broadcast signals to be
sure they remain in their allocated frequency bands. Maximum power
output of transmitters is also regulated. Quality standards for broadcast
signals were set by the FCC, but in recent years enforcement has been
relaxed, to the detriment of the technical quality of broadcasting.
Feedback There are two types of feedback of interest to the audio person:
acoustic and electronic.
Acoustic feedback is the condition where a gain control is set too high
in a sound reinforcement system and the amplied sound enters the
microphone and is re-amplied until a steady howl or whistle is heard.
This is sometimes also called regeneration. The remedy is to reduce the vol-
ume control setting. Another way to reduce feedback is to equalize the sound
system so the response is smoother. The acoustic gain of a system is higher
at peaks in the response curve, and these are the frequencies where feed-
back occurs. Also, the phase shift in the entire system from microphone
through electronics, loudspeakers, and the room back to the microphone
is important. If this phase shift is a multiple of 360 degrees, the system will
be likely to regenerate.

145
Feedforward

equalization for feedback control is tricky, however. The best that one
can do is equalize the entire system for a specic microphone and micro-
phone location. The acoustic gain is dependent on the spatial relationship
between the loudspeakers and the microphone, and different microphone
locations require different equalization curves.
Devices such as the Sabine Feedback Eliminator have been developed
that constantly monitor the sound system and quickly activate a notch lter
if feedback is detected.
There are also automatic techniques by which a dual-channel fft ana-
lyzer is used to determine the system frequency and phase response
using live music as the test signal. Then the proper equalization is deter-
mined to optimize the response curve and the equalization is added while
the analysis is going on. In this way the system is continually updated and
changes in acoustics due to temperature or humidity variations are com-
pensated for in almost real time during the actual concert.
The other type of feedback important in audio is negative feedback
applied to amplifiers and some other audio devices. Negative feedback
is the insertion of a small portion of the output voltage of an amplier,
that is 180 degrees out of phase with the input, back to the input so as to
cancel part of the input signal. This reduces the gain of the amplier, but
also reduces the distortion and noise introduced by the amplier. The
lost gain must be made up, but the net effect of the feedback is still to reduce
distortion. Negative feedback also reduces the output impedance of the
amplier, improving its damping factor, and making it more suitable
to drive a loudspeaker. There is much controversy in audio design circles
about how best to apply negative feedback. It is widely thought that for
best results it should be used in small amounts, and around local gain stages
rather than globally around an entire device.
Negative feedback is of prime importance in electronic designs using
operational ampliers, or opamps, as they are commonly called. An opamp
is a device with extremely high gain, and its characteristics can be tailored
to various tasks such as microphone preamplifiers, voltage ampliers,
active equalizers, etc., by application of proper feedback.
Negative feedback as an improvement in audio ampliers was invented
by a Bell Telephone Labs scientist named Harold S. Black, and was rst
described in the Bell Labs Technical Review in 1934.
Feedforward A type of circuit to reduce distortion caused by an active
component. The feedforward amplies the signal at the input of the com-
ponent and adds this signal to the output signal of the device so as to can-
cel the distortion. Feedforward can only be used in conjunction with
feedback and is effective only when the distortion-creating mechanism
is well understood.
Feed-through Capacitor A special capacitor, usually of less than 50-pF
capacitance, designed to pass audio signals through metal partitions
or shields, but at the same time to shunt radio frequency (rf) energy to
the chassis. The purpose of using feed-through capacitors is to prevent

146
 FFT Analyzer

RF from either entering or escaping from shielded sections of electronic cir-

cuits in order to prevent or reduce radio frequency interference.
Ferrite According to the OED, a mixed oxide of ferric iron and another metal
or metals; specically, a compound of formula MFe2O4 (where M is a diva-
lent metal), many examples of which have magnetic and electrical prop-
erties, which make them suitable for use in high-frequency electrical
components.
In other words, a nonmetallic composite material that has a high mag-
netic permeability and a low electrical conductivity. The high perme-
ability means that coils that use it as a core have a relatively high
inductance. The low conductivity means it has low losses due to eddy
currents.
Ferrite is well suited to high-frequency use and is used for the cores of
transformers operating at radio frequencies. Its low losses also make it suit-
able for use in the cores of magnetic tape heads, especially erase heads.
Ferrite beads, which are small doughnut-shaped pieces of ferrite, are used
to reduce the susceptibility of audio devices to radio frequency inter-
ference. The beads are placed around the signal-carrying wires near the
input terminals of the device, and they increase the series inductance of the
wire, attenuating very high frequencies.
Ferrite Bead See ferrite.
Ferro Fluid An interesting material that is a stable colloidal suspension of
small magnetic particles in a uid. In the presence of a magnetic eld, the
uid becomes quite stiff, and when out of a magnetic eld, it resembles lubri-
cating oil.
Ferro uid is used in some loudspeakers, especially tweeters, to con-
duct heat from the voice coil to the magnet assembly. The uid is placed
around the voice coil in the magnetic gap, and the magnetic eld keeps it
in there. It is a much better heat conductor than air, and because it keeps
the coils cool, more power can be applied to the coil. The uid also acts
as a damping agent, but this effect is minimal.
FET See field effect transistor.
FFT Analyzer The Fast Fourier Transform (FFT) is an algorithm for the extrac-
tion of a frequency spectrum from a time waveform. See also fourier
analysis.
The FFT analyzer is a digital device that performs the seemingly mag-
ical transformation from the time domain to the frequency domain. It does
so very quickly, hence the name. The FFT analyzer can be, depending upon
its design, an extremely high-resolution analysis device with wide frequency
range and dynamic range.
For the highest accuracy under laboratory oconditions it is the analyzer
of choice for measuring distortion, signal-to-noise ratio, phase
response, frequency response, and flutter and wow, because it can
make the measurements over a wide frequency band simultaneously.
It is also able to measure the tef curves, familiar from tds measurements.
The dual-channel analyzer uses the input and output signals from a

147
Fiddle

device and calculates the devices characteristics from them. It measures a

system rather than a signal. This means the signal itself need not be any
special type, such as a sine wave. For instance, an ideal signal for mea-
suring the frequency response of a loudspeaker or tape recorder is
music. The FFT notices the difference between the music signal at the
speakers input terminals and the speaker output as sensed by a mea-
surement microphone and calculates the frequency response based
on this difference. It is thus a simple matter to measure response at various
levels and with various types of music. It is even possible to measure the
entire frequency response of a sound reinforcement system, including the
acoustics of the auditorium, in real time and with live music as a sig-
nal. In this way, the system can be equalized while it is being used for a
concert, and the audience is unaware that any testing is going on.
Fiddle An instrument to tickle human ears by friction of a horses tail over
the entrails of a dead cat.Ambrose Bierce
Fidelity The accuracy with which a music reproduction system will re-
create the sound of the original music. True delity would be perfect repro-
duction, which can be considered an ideal to be aimed for.
It is impossible to reproduce at the ears of a listener an exact replica of
the sounds they would hear in a concert hall. The greatest discrepancies
are probably the lack of directional realism caused by too few independent
channels, and the impressing of the local acoustics of the listening room
onto the recorded ambience of the original recording room. Some multi-
channel systems address these problems; see ambisonics.
Luckily (for the sound equipment makers), our ears are not too critical
in their evaluation of delity and they can be easily trained to accept gross
distortions as accurate reproduction.
High delity sound systems are generally designed to please the lis-
tener rather than to achieve high accuracy. It has been shown many times
over that when the average listener is allowed to adjust the tonal balance
of a sound system as he wishes, he will seldom if ever do so in a way that
maximizes the accuracy of the system; in fact it seems he will often maxi-
mize the deviation from highest delity.
Field In ntsc color video, each frame lasts 1/29.97 seconds, and is actu-
ally composed of two interlaced sequential half pictures consisting of
every other horizontal line lasting 1/59.94 seconds. Each such half picture,
i.e., the odd-numbered lines and the even-numbered lines, is called a eld.
Field Effect Transistor, FET A special type of transistor that behaves
somewhat more like a triode than a transistor. The FET is symmetrical,
and its three terminals are the gate, the source, and the drain. current
in the FET is from the source to the drain, and the magnitude of this cur-
rent is determined by the voltage on the gate.
There is almost no current in the gate, meaning the FET has a very high
input impedance, similar to that of a triode. The FET is relatively lin-
ear in its control of the current by the gate, and it is easier to design low-
distortion ampliers using FETs than using regular, or bipolar, transistors.

148
 Figure-8 Microphone

FETs are used frequently in the preamplifiers in condenser micro-

phones, where their high input impedance is ideal.
There are two main types of FET, called junction FETs or JFETs and metal-
oxide-semiconductor FETs, or MOSFETS. JFETs are the most common and
inexpensive. MOSFETs have exceptionally high input impedance and are
used where this is important. Either of these can be of one of two classes,
called N-channel and P-channel. The N-channel FET requires the gate to
have negative bias with respect to the source, and the P-channel FET
requires a positive bias. MOSFETS have the gate insulated from the drain-
source channel, and they are sometimes called IGFETs (insulated gate FETs).
They have surpassingly high, nearly innite, input impedances.
A third type of FET is the unijunction transistor, which behaves like a
nonlinear resistor. It is used in certain oscillator circuits.
Dr. Julius Lilienfeld has been credited with inventing and patenting the
eld effect transistor in the 1920s, before the invention of the bipolar tran-
sistor at Bell Labs.

Field Effect Transistor

Field Recording Any recording not done in a studio. The resultant tapes
may be called eld tapes, although as tapes become obsolete, we may see
eld coherent laser cubes. See also location recording.
Figure-8 Microphone A microphone whose directional pattern resembles
the gure 8, meaning it is insensitive to the sides but has full sensitivity at

149
Filament

the front and back. The gure-8 pattern is the shape of the mathematical
cosine curve when plotted in polar coordinates, so a gure-8 microphone
is also sometimes called a cosine microphone. Historically, the only gure-
8 microphones available were ribbon microphones, but today many
condenser microphones are made that can be adjusted to provide the
gure-8 pattern. See also variable pattern microphone.
Filament The heater in a vacuum tube, which heats the cathode to induce
it to give off electrons, is sometimes called the lament. This is analogous
to the lament in a light bulb, although it usually does not get as hot. Typ-
ical tubes are heated until the cathode glows red. The major consumers of
power in vacuum tube equipment are the heaters, and a big advantage
of solid-state equipment is the reduced power consumption and reduced
heat output. It has been shown that the failure rate of electronic compo-
nents approximately doubles for each 10-degree Farenheit temperature rise,
and increased reliability is another advantage of the fact that solid-state
equipment requires no heaters.
Fill The ambient sound between words in a production track that is used
both to replace undesirable noises on the track and to create handles
extending the track at the beginning and end. Handles enable the re-
recording mixer to crossfade smoothly between shots with differing back-
ground tones.13
Film Chain A device consisting of a motion picture projector and video cam-
era, used to copy lms onto videotape or to broadcast them directly. To rec-
oncile the 24 frames per second of the movie to the ntsc television frame
rate of a little less than 30 frames per second NTSC video frame rate, some
chains use a projector with a ve-bladed shutter, which projects each frame
of lm ve times into the video camera. The resulting 120 frames per sec-
ond are regrouped four-at-a-time into 30 video images per second. See also
drop frame and 3:2 pull down.
The lm chain is nearly obsolete today, having been replaced by a sim-
pler device known as the ying spot scanner, wherein the lm moves con-
tinuously rather than intermittently and is scanned with a tiny spot of light.
Filter A lter is a type of equalizer that is designed to reduce the energy
at a certain frequency or in a certain frequency band. Filters always act
as subtractive devices, never adding anything to a signal; at least they should
not. The most common type of lters are analog lters, which operate on
signals directly.
An example of a lter is a subsonic or rumble lter found on some
record playing equipment. This lter attenuates the very low frequencies,
typically below 15 hertz or so, and this reduces the effect of noise caused
by the mechanical vibration of the turntable.
Analog lters come in many types, but they always use reactive ele-
ments in their design, such as inductors and/or capacitors. Perhaps

13. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

150
 Fishpole

the simplest lters to build are made of resistors and capacitors, and
they are called R-C lters. Some lters use active components such as
amplifiers or gyrators in their design, and they are called active lters
as opposed to passive lters. Some lters use delay lines in feedback
circuits, and they are called recursive lters.
digital lters operate on signals that have been digitized. They are
purely mathematical, performing a series of arithmetic operations on the
digital words. In a sense, digital lters are synthesized lters; digital tech-
niques being used to emulate or simulate analog lters. Digital lters have
the advantage of being drift-free. They always do their job in exactly the
same way. They can be designed for nearly any desired characteristics in
the frequency domain and in phase response.
In some digital audio systems, notably some compact disc players,
digital lters are used as portions of the anti-imaging and anti-aliasing
lters. Of course a digital lter cannot work on an analog signal directly;
it must rst be converted into digital form. This means an analog anti-alias
lter is needed before the rst sampling of the signal. After this initial dig-
itization, digital lters can be used. See also oversampling.
FIR, Finite Impulse Response A commonly used type of digital lter. Dig-
itized samples of the audio signal serve as inputs, and each ltered output
is computed from a weighted sum of a nite number of previous inputs.
An FIR lter can be designed to have linear phase (i.e., constant time
delay, regardless of frequency). FIR lters designed for frequencies much
lower than the sample rate and/or with sharp transitions are computa-
tionally intensive, with large time delays.14
FireWire, IEEE 1394 A high-speed serial digital interface bus that supports
multiple data formats so that audio, video, MIDI, and control signals may
all be sent over a single cable. The fastest current data rate is 800 megabits
per second with a 100 meter cable limit, whereas the original FireWire was
400 Mbps with a 4.5m cable limit. Different connectors are used. FireWire
can distribute power as well as data, and the system is smart-conguring,
permitting hot-plugging of devices. IEEE 1394 is designed to be bidirec-
tional and with the ability to broadcast from a single source to multiple
receivers. One of FireWires important features is its ability to transmit
isochronous data such as digital video. FireWire has found wide use in
computer and digital audio-video devices. The term FireWire was coined
by Apple Computer, which invented it and submitted it to the IEEE where
it became standardized as IEEE 1394.
First Law of Acoustics A variation on Murphys law which can be stated,
Never make the same measurement twice, because you will not get the
same answer. The complexity and variability of acoustic phenomena make
it very difcult to make exactly repeatable measurements.
Fishpole Slang for microphone boom.

14. Thanks to Dennis Bohn, Rane Corp.

151
5.1, 6.1, 7.1

5.1, 6.1, 7.1 In soundtracks such as those on movies and dvds, 5.1 means
that the soundtracks are recorded with ve main channels: left, center, right,
left surround, and right surround, plus a low-frequency-effects (LFE) bass
channel (called a .1 channel because it covers only about 10 percent of
the frequency range of the main channels).
In his excellent book 5.1 Surround Sound (2000), Tomlinson Holman
states: In 1987, when a subcommittee of the Society of Motion Picture and
Television Engineers looked at putting digital sound on lm meetings were
held about the requirements of the system. In a series of meetings and doc-
uments, the 5.1-channel system emerged as being the minimum number
of channels that would create the sensations desired from a new system,
and the name 5.1 took hold from that time, In fact, this can be seen as a
codication of existing 70mm practice, which already had ve main chan-
nels and a low-frequency, high headroom channel.
Some movie soundtracks use a variation of 5.1 called Dolby Digital
Surround EX, which can be used via DVDs in home theaters. The Sur-
round EX format matrix encodes a third surround channel onto the left and
right surround channels of 5.1 soundtracks, and may be decoded or not at
the cinemas or home listeners option due to their inherent compatibility.
Because the extra surround information is carried on the left and right sur-
round channels, Dolby Digital Surround EX encoded soundtracks are still
regarded as 5.1 soundtracks.
With respect to home playback, the terms 5.1, 6.1, and 7.1 mean that there
are ve, six, or seven main speakers, plus a subwoofer, in the playback sys-
tem. (The subwoofer reproduces the LFE channel recorded on 5.1 sound-
tracks, plus any bass the main speakers cannot handle.) The difference is
in the number of surround speakers: two in a 5.1 system, three in a 6.1 sys-
tem, and four in a 7.1 system.
5.1-channel soundtracks can be played on a 5.1-speaker system, but they
can also be played on a 6.1- or 7.1-speaker system. To do this, the two sur-
round signals on the 5.1 soundtrack are spread across the three or four sur-
round speakers. This distribution can be accomplished by a Dolby Digital
EX decoder, a THX Surround EX decoder, or other proprietary methods
provided in home theater equipment by various manufacturers.
The number (i.e., 5.1) describing the soundtrack does not have to match
the number applied to the speaker system. It is even possible to play two-
channel stereo content over these multi-speaker systems by using a matrix
surround decoder such as Dolby Pro Logic II. The delivery format and
the speaker conguration are independent, and it is the decoders job to
bridge them effectively.
Fixed Bias A circuit arrangement where the bias of an active component
is supplied by a separate power supply and is thus constant even though
the current in the component may vary.
Flag A digital bit or group of bits that is inserted in a digital code to accom-
plish a specic task different from that which the code usually does. Orig-
inally, ags were put into computer programs at certain points to call the

152
 Flatbed Editor

programmers attention to certain aspects of the program. These ags have

no effect on the execution of the program. In digital audio, however, ags
may be used to disable the recording circuits of digital tape recorders to
prevent unauthorized copying. An example is the anticopy ag found on
most CDs.
Flanging A special effect made popular in the 1960s where a delayed ver-
sion of a signal is mixed with the signal, creating a swooshing sound.
Flanging was rst done by recording a signal on two similar tape
recorders, playing them back simultaneously, and mixing them together.
The record-playback sequence on the tape recorders results in a small time
delay of perhaps a tenth of a second. Both output signals are delayed by
the same amount if the tape recorders are similar and they add together in
the mixer, and the sound heard is essentially the same as the signal at the
input to the tape recorders.
To achieve the effect of anging, one recorder is slowed down a little,
increasing its time delay. This is done by pressing ones thumb against the
ange of the tape recorder supply reel, hence the name anging. When
the time delay is different for the two combined signals, there will be fre-
quencies where the phase shift is 180 degrees, and the signals will can-
cel, causing deep dips or holes in the frequency response curve. This
is called the comb filter effect. As the speed is varied, the frequency of
the dips is swept across the frequency range, giving the swooshing sound.
Attaining the most desirable effect requires an educated thumb. The best
effect is obtained when the signal being anged contains frequencies over
a wide range.
This technique was thought primitive by many, which led to the devel-
opment of all-electronic angers. These units contain an electronic delay
line that is adjustable. If the time delay is very short, the effect is called
phasing, and does not sound nearly as convincing as true anging.
It is interesting to note that a sound similar to anging is familiar to many
people. As a jet airplane passes overhead, the direct sound is heard rst,
and the sound reected from the ground, buildings, etc., is heard later. This
time delay causes a cancellation of the combined sound at the ear at some
frequency, and this frequency varies as the airplane moves and the path
lengths change. This is the reason anging reminds one of the sound of a
jet plane.
Flash Excess vinyl compound squeezed out around the stampers when
records are pressed. It is removed with a special cutter.
Flat Lower in pitch, as opposed to sharp, which means higher in pitch.
In reference to musical scales, at indicates one-half tone lower; for exam-
ple, B at is a whole tone lower than C.
Flatbed Editor A machine for editing motion picture lm and its sound
track. It is essentially a table with several horizontal pairs of reels of lm
and sprocketed magnetic sound stock. The picture lm is routed through
an apparatus that projects the image onto a small screen for viewing, and
the sound stock is routed through a playback head assembly. The sound

153
Flat Frequency Response

and picture lms are synchronized to run at the same speed, and this speed
is variable so the editor can hear the sound and see the picture at low speeds
and in forward or reverse to allow her to select the editing points. Both lms
are then cut and spliced to maintain synchronization of the sound and pic-
ture. Such an editor may be referred to by a brand name such as moviola
or Steenbeck, however, such manufacturers also make other editing prod-
ucts that may also be called by the brand name.
Flat Frequency Response An amplifier, loudspeaker, microphone,
etc., is said to have at frequency response, or at response, if its output
is at the same level for all frequencies of interest, provided its input also
has uniform amplitude over the same frequency range. In other words,
a at system has the same gain or sensitivity at all frequencies of inter-
est. A graph of the gain versus frequency will be a straight line, hence the
name. In reality, a tolerance is allowed, typically plus or minus 3 dB. See
also frequency response.
Fleming Valve A vacuum diode patented in 1904 as a detector for use
in wireless telegraphy by the English professor Ambrose Fleming. Flem-
ing did not build the rst diode; he made use of the edison effect bulbs
brought to England in 1884 by William Preece, who had obtained them from
Edison. Fleming was working with the Marconi Wireless Company and was
trying to nd a better detector than the coherers then in use. He realized
the Edison bulbs would rectify low-frequency AC signals, and he experi-
mented to verify that they could be used at radio frequencies.
When news of Flemings patent reached Edison, he immediately led
suit because of his 1884 patent of the Edison effect, but the courts ruled,
after a prolonged dispute, that Edisons patent did not cover the use of the
device as a detector.
It is interesting that Fleming resisted the use of the term diode (rst pro-
posed by Dr. Eccles of Manchester University in 1919) and insisted that it
be called an oscillation valve.
Fletcher-Munson Effect In the early 1930s, Fletcher and Munson under-
took to measure the sensitivity of the human hearing mechanism at dif-
ferent frequencies. They rst generated sounds at very low levels at many
frequencies over the audible range to determine the threshold of hearing,
or the softest sound that could be heard. They made a plot of these levels
versus frequency, and found the curve was not uniform, but varies dras-
tically with frequency. The most sensitive range of human hearing is
between 3 kilohertz and 4 kHz; the sensitivity falls off rapidly at lower
frequencies and somewhat more slowly at higher frequencies. In other
words, very soft sounds must be more powerful at frequencies lower and
higher than 3 to 4 kHz in order to be heard. This result was expected from
previous experience, but the effect had not been precisely measured before.
Next, the experimenters chose 1,000 hertz as a reference frequency and
increased the strength of the softest audible sound at that frequency by ten
times (10 decibels). They then generated other tones at lower and higher
frequencies and strengths and asked the subjects to judge when these other

154
 Flutter

tones were as loud as the reference tone at 1,000 Hz. The strengths of these
tones were plotted on the same graph as the threshold was plotted to form
a contour of equal loudness and the surprising result was that this curve
was not parallel to the threshold curve. It was not as steeply curved, indi-
cating that our ears hear different frequency tones more uniformly in loud-
ness when they are stronger than the threshold levels. The experiment was
repeated with 1,000-Hz reference tones in increments of 10 dB all the way
up to 120 dB above threshold.
These equal loudness contours become atter as the level increases, indi-
cating that our ears have much more uniform sensitivity at high levels than
at low levels.
The results of this experiment were widely reported and were thought
to be important in understanding how we perceive music. One outcome
of this was the so-called loudness control found in many music-repro-
ducing systems. The loudness control attempts to compensate for the
Fletcher-Munson effect by boosting the lowest and highest frequencies of
the music signal by an amount that increases as the volume is reduced.
Thus, as one turns down the volume control, the low and high frequencies
are reduced less than the frequencies between 3 and 4 kHz.
The loudness control, however, was not as great a success as the design-
ers expected it to be, and most music listeners ignored it, or used it very
little.
The reason for this is a further complexity in the way we perceive com-
plex sounds. The Fletcher-Munson experiment used sine waves, or pure
tones, to measure the equal loudness contours, but music never consists of
pure tones. All music waveforms are complex in that they contain energy
at many frequencies in a harmonic series, and such complex sounds do
not obey the equal loudness contours, at least not the ones measured for
pure tones. Most rock listeners seem to love the loudness switch, as it greatly
boosts the bass and treble, which they associate with rock music.
Floating Floating is a circuit conguration in which neither wire carry-
ing the signal is connected to the system ground. See balanced;
Appendix 6.
Fluff Either a verb meaning to garble or miss a word or phrase when speak-
ing into a microphone or a noun meaning the result of this action. Some-
times the word blooper is used as a synonym for the noun, but uff is
more common.
Flutter In an analog tape recorder, if the tape speed varies, the pitch of
the recorded music will vary. If the rate of variation is fairly high, typically
above 5 hertz or so, it is called utter. If the speed varies at rates below
several hertz, it is called wow. Flutter is actually frequency modula-
tion and imparts a tremolo-like or vibrato-like character to the music.
It is very annoying with certain types of music, such as piano, which never
has any natural pitch variation. On the other hand, recordings of stringed
instruments are much more tolerant of utter, string players not being able
to sustain a steady pitch.

155
Flutter Echo

According to the din and iec standards, utter is measured by record-

ing a steady 3,150-Hz tone and then passing the resultant signal through
an fm detector. The Japanese standard (JIS) calls for 3,000 Hz rather than
3,150 Hz. The demodulated portion of the signal is then passed through a
weighting lter, which discriminates against frequencies above 15 Hz or
so, and then the rms value may be determined. This is expressed as a per-
centage and is called the RMS weighted utter. The DIN standard calls for
peak rather than RMS detection and so is more strict than the American
standard.
An acceptable value for utter in a tape recorder is 0.1 percent or less.
Record turntables also have utter, but the speed variations are generally
much lower in frequency and are called wow. The measurement tech-
nique is the same, with a different weighting curve.
It is said that the Flying Karamazov Brothers juggling team has a cat
named Flutter.
Flutter Echo An acoustic effect in some rooms where sound is reected back
and forth between two parallel surfaces, such as opposite walls. In order
to qualify as a utter echo the reections must be fewer than about 15 or
so per second. This happens when the walls are more than about 25 feet
apart. Parallel surfaces closer than this give rise to standing waves that
result in a nonuniform distribution of sound between the surfaces.
Flutter echoes are to be avoided in auditoria, and this is usually done
by making large wall surfaces nonparallel.
Flutter Roller See scrape flutter.
Flux The total number of lines of magnetic induction threading through a
surface is called the magnetic ux through the surface. The unit of ux is
the weber, abbreviated Wb, which is a relatively large unit. The nanowe-
ber (nWb), which is one billionth of a weber, is used to measure magnetic
ux levels that exist in analog recording tape. Before the adoption of the
weber, the Maxwell was used, which was one tenth of a nWb. One may see
a reference to a uxivity of 20 milliMaxwells/mm, equal to 200 nWb/m.
A magnetic tape head, when replaying a tape, is sensitive to the rate
of change of ux in its iron pole pieces, not to the ux itself. See tape head.
Flux also refers to a cleaning agent used on metals before they are sol-
dered. Rosin is a typical soldering ux and is commonly packaged with
the solder as a small core running down the center of the solder wire. This
is called rosin core solder. For soldering heavily corroded metals, an acid
ux is often used. Acid core solder is available but is never used in elec-
tronics work because the acid is very corrosive and damages electronic cir-
cuit components.
Flux Density A measure of the strength of a magnetic eld. The unit of ux
density is the tesla, named after Nikola Tesla, the Serbian scientist. In a
loudspeaker, the ux density supplied by the magnet is closely related
to the overall quality, transient response, and efciency of the unit, as well
as its cost. The traditional material for loudspeaker magnets has been a
cobalt-nickel alloy called Alnico, but in recent years ceramic magnets have

156
 FM Stereo

been used in large quantities. Some relatively exotic materials such as samar-
ium cobalt and neodymium can produce extremely strong magnetic elds
and are sometimes used for high-frequency drivers.
FM See frequency modulation.
FM Stereo A method of FM broadcasting that allows stereo receivers to
receive the full stereo signal. It is also compatible with monaural FM
receivers, which simply sense the sum of the two stereo channels.
The system works by rst forming sum (L+R) and difference (LR) sig-
nals from the stereo input signals (L and R). This is sometimes called ma-
trixing. It can be shown that all the stereo information resides in the
difference signal; in other words, if the difference is zero, then the original
signals were identical, and it therefore was not stereo. The LR signal, which
is normally lower in level than the L+R signal, is used to modulate a 38-
kilohertz subcarrier, which is then added to the L+R signal and used to mod-
ulate the main FM carrier. In addition to this, a low-level 19-kHz pilot signal
is added to the composite signal. This pilot, at one-half the frequency of
the subcarrier, is used by the receiver to generate a 38-kHz subcarrier with
which to demodulate the LR signal. It is interesting to note that the stereo
indicator lights on most FM tuners use the existence of the 19 kHz pilot to
identify stereo broadcasts. If the station is broadcasting monaural source
material, but the pilot is still transmitted, the resulting audio signal will
still be monaural even though the stereo indicator will still light up.
A standard monaural receiver is not sensitive to the subcarrier and sim-
ply demodulates the L+R signal. A stereo receiver senses the pilot tone and
uses this to demodulate the LR signal, which is then matrixed with the
L+R to obtain the original L and R signals.
The advantage of this system is that it is compatible with existing mon-
aural FM receivers. Its disadvantage is that the subcarrier, which carries all
the stereo information, is much more sensitive to signal degradation from
multipath distortion and long-distance reception. This means that
good-quality stereo reception is limited to an area relatively near the
transmitter.
The 19-kHz pilot signal must also be ltered out of the signal, because
it will cause interference with the bias frequency of analog tape recorders,
causing audible whistles to be recorded. See also birdie. It can also confuse
noise reduction systems such as Dolby, which is why cassette decks with
Dolby NR must have a 19kHz, or multiplex notch lter switch.
Over the years, there have been many attempts to broadcast stereo pro-
grams, and they competed ercely for fcc standardization in the late 1950s.
Probably the rst of these was a system designed by Murray Crosby and
William Halstead and experimentally broadcast in 1950 under the call let-
ters KE2XKH in New York City. This system used a 35-kHz subcarrier that
was modulated by the second stereo channel, and then this signal was mixed
with the other channel, and the combination modulated the main carrier.
The system was called Stereosonic, from stereo using an ultrasonic sub-
carrier. (See stereosonic for another, more common use of the term.)

157
FM Synthesis

Later, in 1958, Murray Crosby patented the Crosby multiplex system,

which used sum and difference matrixing to achieve compatibility with stan-
dard FM mono reception, as in the present-day system. The Crosby sys-
tem was very well designed, boasting complete compatibility, 15-kHz
frequency response on both channels, and no degradation of signal-to-noise
ratio. It used a 41-kHz subcarrier to transmit the LR stereo information.
The subcarrier was also used to modulate the main carrier, unlike the
present-day system.
Experimental broadcasting was undertaken by WBAL in New York
(operating under the call letters KE2XXT) and by KDKA in Pittsburgh.
Many engineers at the time expected the FCC to standardize on the
Crosby system because the competing systems were of demonstrably
lesser quality, but the FCC chose the system in use today because it allows
another channel to be simultaneously broadcast (see sca) with another sub-
carrier at 67 kHz. The result is the degradation of the signal-to-noise
ratio of the stereo program and reduced listening range from the station.
Among the competing stereo broadcast systems were the Halstead,
which used no matrixing and modulated the 41-kHz subcarrier with the
right channel. The left channel modulated the main carrier. This system was
not compatible with monaural reception and had differing frequency
responses on each channel. The Burden system tried to overcome these
difculties by using sum and difference matrixing in transmission like the
Crosby system. However, no de-matrixing occurred in the receiver, pre-
sumably to reduce cost. Burden argued that the requisite de-matrixing
would occur acoustically in the listening room. How this was supposed to
happen was not satisfactorily explained, and the system never saw the light
of day. The Phantodyne and Calbest systems were also contenders, but they
suffered from poor frequency response and were not taken seriously.
It is interesting to note that in 1958, three-channel stereo was broadcast
experimentally in Boston by WGBH-TV, WGBH-FM, and WBUR-FM. The
center channel was carried by the TV station, and the two FM stations broad-
cast the other two channels. The listener needed a TV set and two FM sets
to hear the full effect, and it was said to be very convincing.
Stereo broadcasting was also done experimentally in England in the 1950s
and as early as 1926 using two AM stations. An interesting note is that the
BBC, in its two-station broadcasts in 1958, reasoned that it had no control
over the polarity of the sound output of the receivers in use and that there
was only a 50% chance that the phasing would be correct. Applying impec-
cable British logic, it reversed the phase of one channel halfway through
each program, ensuring that each listener would hear the program under
the proper conditions half the time!
FM Synthesis The generation of complex signal waveforms in electronic
music by frequency modulation of one or more sine wave signals by
other sine waves. This method is an alternative to additive synthesis.
FM synthesis as a method of generating complex musical waveforms
was pioneered by John Chowning of Stanford University, and he has shown

158
 Foldback

that an extremely wide variety of waveforms may be made this way. The
method also requires signicantly less hardware than other methods, such
as additive synthesis.
One of the rst commercial synthesizers to use FM synthesis was the
Synclavier, produced by New England Digital Corp. There are now many
other such units available.
FMX A proprietary system of FM transmission and reception to improve
the signal-to-noise ratio in stereo broadcasting. FMX was developed
by the now-defunct CBS Technology Center in the mid-1980s. It is well
known that FM stereo reception is much noisier than monaural recep-
tion in low-signal, or fringe, areas. This is a result of the method by which
the stereo information is encoded on the transmitted signal (see fm stereo).
CBS claimed that quiet FMX stereo reception extends as far from the trans-
mitter as quiet monaural reception of conventional FM. This more than dou-
bled the area in which stereo could be received from a given station.
FMX added another subcarrier to the transmitted signal that was 90
degrees out of phase with the stereo subcarrier. This new subcarrier was
said to be in quadrature with the main subcarrier. The stereo information
in the input signal, consisting of the LR, was compressed and used to mod-
ulate the new carrier. In the FMX receiver, the quadrature carrier was sensed
and demodulated, and the LR signal expanded to restore the dynamics
of the original signal. In order to properly restore the quadrature LR sig-
nal, the receiver used the level of the stereo LR signal from the standard
subcarrier as a reference. Even though this signal may be noisy, its level
was indicative of the amount of stereo separation in the original signal. The
quadrature LR signal was then used, after expansion, to restore the orig-
inal stereo signal.
FMX was compatible with regular FM stereo because a standard stereo
receiver does not detect the quadrature subcarrier. There was some con-
troversy over whether the FMX system would work properly in the pres-
ence of multipath reception. multipath distortion in FM is analogous
to ghosts on television, and is especially severe for stereo transmission.
FOH, Front Of House. The location within the audience area of an audi-
torium or arena for the mixing console and its operator that controls the
sound the audience hears, or the house sound. In many cases, especially
in large arenas, there will be another control console location at one side or
the other of the stage. This console is called the monitor mixer, and con-
trols the sound into the monitor speakers through which the performers
hear themselves. Both the FOH and the monitor mixer inputs will be con-
nected to all the microphones on stage for maximum exibility. The sig-
nal fed to the stage monitors is usually not the same as is sent to the house
sound reinforcement speakers. This is because the performers may want
to emphasize certain microphones, such as vocals, so they can better hear
themselves.
Foil The copper conducting path on a printed circuit board.
Foldback Foldback is the general term for the part of a sound reinforce-

159
Folded Horn

ment system in an auditorium that supplies amplied sound to the per-

formers so they can hear themselves. Foldback loudspeakers are often
placed on the stage in front of the performers, although they may be above
or behind the performers. They are commonly called monitors by the per-
formers. Another use of the term is the foldback earphone often used by
so-called talking head television personalities. The earphone allows the
person to hear his own voice along with directions from a technician or pro-
ducer. See also ifb.
Folded Horn A type of horn loudspeaker where the horn is folded or
wrapped around itself to save space. Mostly, folded horns are used for
low frequencies, but there are some high-frequency folded horns.
There have been many designs for folded horns over the years, some
which use a corner of a room as an extension of the horn itself. The Klip-
schorn, patented by Paul Klipsch in the 1940s, was of this type. It is inter-
esting to note that the rst commercially available exponential horn
reproducer was a folded horn. This was the Orthophonic Victrola acoustic
phonograph introduced in 1925. The exponential horn was rst described
by A. G. Webster of the Bell Telephone Laboratories in his article that
appeared in the Proceedings of the National Academy of Sciences in 1919 titled
Acoustical Impedance and Theory of Horns and Phonograph.
This was one of the key papers in the early history of sound reproduc-
tion, introducing the concepts of acoustical and mechanical impedance
and noting the similarities to electrical impedance.
Foley Effects Sound effects in a motion picture or video tape that are pro-
duced by various mechanical devices operated by hand and other sounds
such as footsteps. Examples are slamming doors and creaking oors and
hinges. The recording of Foley effects is carried out in a dedicated studio
called a Foley stage. Some of the contraptions used for Foley effects are quite
complex and ingenious. Purportedly named after an early practioner of
these effects, Jack Foley.
Foley Stage See foley effects.
Foley Walkers Foley technicians that work on a Foley stage to record Foley
effects. They use a variety of objects and/or surfaces to generate realistic
sound effects. On-screen footsteps are one of the most commonly used
effects, hence the term walkers.
Follower An active circuit element, such as an opamp or a tube, which
is connected so that its output voltage will be the same as its input volt-
age. In other words, it will have a gain of 1. Followers act as impedance
converters, and this is the reason they are used. A follower will have a high
input impedance (sometimes extremely high) and will have a low out-
put impedance. It is thus suited to driving long lines, or low-impedance
circuits, while it does not act as a signicant load on the device connected
to its input. One use of the follower is the rst stage in the preamplier for
a condenser microphone. See also cathode follower; emitter
follower.
Forensic Audio The practice of techniques that analyze recordings for valid-

160
 Forward Voltage Drop

ity. This is often done for legal reasons, such as to determine if a wiretap
recording is genuine or altered. A famous instance of forensic audio was
the analysis of the gap in the Watergate tapes.
Formant A frequency band in the spectrum of a voice or musical instru-
ment that contains more energy or amplitude than the neighboring area.
Formants are the distinguishing characteristics of the vowel sounds of the
human voice and, for any vowel, are relatively xed in frequency, even
though the pitch of the voice may be changing, as in singing. The overall
shape of the spectrum of a musical or vocal sound is called the spectral enve-
lope, and may not change much as different pitches are sounded. The for-
mants determine this shape in large part.
A musical instrument has a denite set of formants, and it is they that
impart its tone color, or timbre. The fact that the timbre does not change
too much as the instrument sounds different notes allows us to recognize
the instrument regardless of the pitch being played. It is interesting that
our hearing mechanism is very sensitive to formants, no doubt because
speech intelligibility depends heavily on recognition of vowels. If the sound
of any musical instrument is prolonged without changing the pitchfor
instance, if it is recorded on a tape loopit soon begins to lose its subjec-
tive timbre and begins to sound like a buzzer. It is the unvarying formant
frequencies in the presence of the varying pitch of the instrument that allow
us to recognize the timbre.
If a sound-reproducing system has peaks in its frequency response, these
peaks will be heard in the presence of music as a denite timbre or col-
oration in the sound. Such a sound system can be said to have a tone color
of its own that it impresses on the music. The same thing happens with
many other audio devices, especially microphones and loudspeakers.
This is why peaks in the response curve of any device are to be avoided.
However, the great adaptability of the human hearing mechanism is such
that if one listens for a long time to a particular sound system, one will
become habituated to that particular sound. This is the reason so many
people put up with very poor quality sound systems and seemingly dont
mind. This is also the reason one should compare the reproduced sound
of a system with the actual sound itself, rather than trusting to memory,
when evaluating components such as loudspeakers.
Former See coil former.
Forward Resistance The actual resistance of a diode or other semi-
conductor junction when conducting current in its normal direction.
The forward resistance of a semiconductor junction is not as simple and
linear as normal, or so-called ohmic, resistance, but depends on the volt-
age across the junction and the current through it, especially at low values
of voltage and current. See also reverse resistance.
Forward Voltage Drop The voltage across a diode or other semicon-
ductor junction when it is passing a current in the normal direction.
Most solid-state junctions will have a nearly constant forward voltage
drop regardless of the current. For silicon junctions, it is about 0.7 volt.

161
Foster-Seeley Detector

Foster-Seeley Detector A type of FM detector developed to provide rejec-

tion of amplitude modulation so it would be less susceptible to radio
frequency interference. It was an improvement over its predecessors
but not as good as the ratio detector.
Fourier Analysis Baron Jean Baptiste Joseph Fourier was a gifted nineteenth-
century French engineer who developed the mathematical techniques to
allow the transformation of a mathematical function from one domain
to another domain. Fourier did this, interestingly enough, while working
on problems of heat transfer in artillery for Napoleon. The Fourier trans-
form allows a function that represents an audio signal (signals are in
the time domain because they exist in time) to be transformed to another
function that represents the same signal in the frequency domain. The
signal in the time domain is called a waveform, and the same signal in
the frequency domain is called a spectrum.
The Fourier transform performs a frequency analysis, assigning peri-
odicities in the waveform to specic frequencies in the spectrum. For
instance, a waveform that repeats itself 100 times per second is said to be
periodic at 100 repetitions per second. This periodicity means that the spec-
trum of the signal will have a component at a frequency of 100 hertz, or
100 cycles per second. Similarly, other periodicities will correspond to other
frequencies in the spectrum. No information is lost when transforming to
the frequency domain from the time domain, and in fact the inverse trans-
form may be applied to go back to time from the frequency domain.
Sometimes the spectrum is called the frequency response function,
but this is not exactly rigorous, for the frequency response is a transfer func-
tion and is a characteristic of a device, whereas a spectrum is simply a sig-
nal expressed in the frequency domain. See frequency response.
The reason we want to transform a signal from the time to the frequency
domain is that our hearing mechanism seems to do a similar thing. We tend
to analyze signals into their frequency components, which are related to
musical pitches. (See pitch.) When we look at a waveform on an oscil-
loscope, it is very difcult to predict what it will sound like. It is almost
impossible to tell the pitch of a complex signal by looking at the waveform,
but when it is transformed into frequency we know that specic frequen-
cies correlate (in a fairly complex way) to pitch sensations. The spectrum
is like looking at the signal from another angle, which reveals detail not
otherwise visible.
Fourier did the mathematics but he did not build a spectrum analyzer.
In fact, it was not until 1967 that a technique (called the Cooley-Tukey algo-
rithm, after the discoverers) was devised that permits a digital computer
to perform the Fourier transform at a very high speed. (It probably should
be called the Lanczos algorithm, for it was actually described by Cornelius
Lanczos much earlier.) This rediscovery led the way to the digital fast
Fourier transform, or FFT, analyzers that are common today.
There are many far-reaching consequences of Fourier analysis, one of

162
 Free Field

which is that any repeating (i.e., periodic) waveform can be represented

by only a fundamental frequency and a series of harmonics. The fun-
damental frequency is the reciprocal of the period of the wave, and the har-
monics occur at integral multiples of this frequency. For instance, a signal
whose period is 100 milliseconds will have a repetition rate of 100 Hz, and
will also contain energy at 200 Hz, 300 Hz, 400 Hz, etc.; no matter what the
shape of the wave is, it will still consist of only this fundamental and har-
monics. The level and phase of the harmonics will vary with the waveform.
Fourier Transform See fourier analysis.
Fox-Hole Sprockets See cinemascope.
Frame In motion pictures or video, a frame is one complete picture. The
standard frame rate in sound movies is 24 pictures per second, and in ntsc
television in the United States, it is 29.97 frames per second. Special effects
such as slow motion are lmed at higher frame rates, and the frame rate
for silent movies was around 16 per second, although it varied widely.
In digital audio, a frame is a slice of digital information. In the com-
pact disc, a frame covers six sampling periods, or 136 microseconds.
Each frame ends with a synchronization code. When digital audio was
recorded onto a video tape recorder, the digital frame was tted into a video
frame, hence the name. The synch pulses of the video format were then
used to synchronize the digital audio data. It was found to be technically
convenient to use a sampling rate of 44.05594 kHz to t the data into the
video signal.
The European standards for video (pal/secam) of 625 lines and 25
frames per second are different from the American standards (NTSC) of
525 lines and almost 30 frames per second.
Free-Cone Resonance The frequency at which the cone of a dynamic
loudspeaker will resonate when in free air; that is, when not mounted
in a baffle or cabinet. The free-cone resonance is the mass of the cone
assembly resonating with the springiness of the cone surround and spi-
der and is always lower in frequency than the mounted resonance. It is
one of the thiele-small parameters.
Free Field A sound source radiating into three-dimensional space where
there are no reecting surfaces is said to be radiating under free eld con-
ditions. The sound pressure level as measured at various distances from
the source would obey the inverse square law precisely. There is no
such thing as a true free eld, but it is approximated in the anechoic
chamber.
Because all rooms have at least a small amount of reverberation, the
sound eld from a source is always contaminated with reected sound. This
makes accurate measurements of sound levels difcult because the micro-
phone is not able to discriminate between the direct and reected sounds.
This is the reason that anechoic rooms are used for much acoustical testing
the reected sound is of sufciently small amplitude that it is negligible
for most work.

163
Free-Field Microphone

For certain small sound sources in large rooms, the sound eld fairly
close to the source will approximate a free eld, but as the distance
increases, the reverberation becomes relatively stronger. The sound eld
closer to the source than about one wavelength of the sound is not a free
eld, but is referred to as the near field; meaningful measurements this
close to a sound source are very difcult to accomplish. See also far field.
Free-Field Microphone A type of omnidirectional microphone that is
designed to have flat response when in a free field and pointed at the
source of sound, as opposed to a true pressure microphone, which will
have an increased high-frequency sensitivity under the same conditions.
The true pressure microphone has a rising high-frequency response
due to reection of sound from the microphone diaphragm, which causes
a pressure increase. The effect is due to the disturbance of the sound eld
by the microphone itself, and only occurs when the wavelength is short
compared to the diameter of the microphone. The free-eld microphone
corrects its response for the disturbance it introduces.
Free-eld microphones are useful when recording is done at close range
to the sound source, where they will give flat frequency response. Pres-
sure microphones are better suited to more distant pickups, where the rise
in high-frequency response helps to compensate for high-frequency losses
due to absorption by the air. Most measurement microphones, as found on
sound level meters, are of the free-eld type.
Free-Field Response The frequency response of a microphone in a free
acoustic eld. Manufacturers will make omni microphones with various
response curves at different angles of incidence to satisfy market demands.
The free-eld responses at different angles will accurately compare the
characteristics.
Frequency Frequency, in its simplest form, is a measure of how often (how
frequently) an event repeats itself. A sound source, such as a tuning fork,
which vibrates back and forth 1,000 times per second, is said to have a fre-
quency of 1,000 hertz.
Frequency used to be stated in cycles per second, or cps, but there has
been international agreement that hertz (Hz) will be used to indicate fre-
quency. The term is in honor of Heinrich Hertz, a German pioneer in the
transmission of radio waves.
Most sounds are complex and cannot be described by a single frequency.
The sound of a bell or a cymbal, for instance, contains very many frequencies
at the same time. The best we can do is to perform a fourier analysis
on the signal and thus determine what individual frequencies are present
in the complex sound.
Frequency of a signal in large part determines its subjective pitch,
although the correlation is not one-to-one, especially for simple tones. Com-
plex tones that are periodic, i.e., that repeat their waveform over and over,
do have a rather precise correlation between frequency and pitch although
the pitch will vary somewhat with the amplitude of the sound. Such a
sound will consist of many component frequencies arranged in a harmonic

164
 Frequency Response

series, but it is usually said to have the frequency of the lowest, or fun-
damental, component. For instance, an oboe sounding a tone repeating
at 440 times per second will sound like A above middle C, and we say it
has a frequency of 440 Hz, even though it has energy at many higher (har-
monic) frequencies as well.
Frequency Deviation See modulation index.
Frequency Discriminator A type of FM detector. See discriminator.
Frequency Doubling An effect sometimes heard with low-frequency
loudspeakers where the second harmonic distortion becomes as
strong or stronger than the fundamental. The effect is that the bass sounds
an octave higher than it really is. Frequency doubling is usually the result
of overloading a speaker.
Frequency Extender A device used to extend the usable frequency range
of a telephone line by about two octaves. A standard telephone line has
a frequency response extending only from about 300 to 3,300 hertz. This
lack of low-frequency response can be troublesome and results in most
male voices sounding thin, or tinny. The frequency extender uses the
frequency shifter principle to shift all frequencies up by 250 Hz on the send-
ing end and back down at the receiving end. This extends the response
down to 50 Hz on the low end and sacrices only the upper 250 Hz on
the upper end.
Frequency Modulation, FM Frequency modulation is the instantaneous
changing of the frequency of a carrier in response to a modulating sig-
nal, usually an audio waveform. As the signal voltage varies up and
down as it follows the waveform, the frequency of the carrier varies up and
down from its nominal unmodulated value. In commercial FM radio
broadcasting, the carrier frequency is in the band from 88 MHz to 108 MHz.
The FM receiver is tuned to the carrier frequency, and the received signal,
after suitable conditioning, is applied to a special circuit called an FM
detector, or discriminator, which recovers the audio signal.
FM transmission is relatively quieter than AM transmission because the
discriminator is not sensitive to amplitude variations caused by atmos-
pheric interference, and it permits wider frequency response because
the FCC has allocated wider bandwidths to be transmitted in the FM band
than in the AM band. The bandwidth of AM transmission is limited to 10
kilohertz to prevent adjacent stations from interfering with one another,
whereas a bandwidth of 100 kHz is allocated in each FM channel.
In music, vibrato is a form of frequency modulation because it is a peri-
odic variation in frequency. FM is also used in the synthesis of musical tones
in some forms of electronic synthesizers. It is possible to achieve a very
wide range of harmonic and nonharmonic effects by this means.
Frequency modulation distortion is a type of distortion produced by
a loudspeaker. See doppler distortion.
Frequency Response Frequency response is a shortened way of stating the
amplitude response versus frequency characteristic. It is a complex func-
tion that describes the way in which the gain and phase of a system or a

165
Frequency Shifter

device vary with the frequency of the stimulus. Frequency response is usu-
ally presented as a graph or plot of the output of a device on the vertical
axis versus the frequency on the horizontal axis. In audio work, it is com-
mon to use the logarithm of frequency as the horizontal axis and to use the
decibel scale for amplitude in the vertical axis. This is sometimes called
a log-log plot, and it correlates better than a linear plot with our hearing
mechanism. The frequency response consists of two parts, called the mag-
nitude and the phase. The magnitude is the part most often seen.
Another way to graph the frequency response is to divide it into real
and imaginary parts. This presentation includes the phase information
as well as the magnitude. The real and imaginary parts are simply the pro-
jections of the complex function onto two orthogonal axes, which are arbi-
trarily called real and imaginary. There is nothing imaginary about the
imaginary part, and it is also sometimes called the quadrature part, indi-
cating that it is a projection on an axis 90 degrees (a quarter circle) from the
real part. The magnitude is the square root of the sum of the squares of the
real and imaginary parts.
It is important to realize that frequency response is dened to be a char-
acteristic of a system or a device, not a characteristic of a signal.
Frequency Shifter A device that linearly shifts all the frequencies of a com-
plex input signal. The amount of shift can be varied, and the shift can be
up or down in frequency. Also sometimes called a spectrum shifter.
The key word here is linear, which means all the frequency components
are shifted by the same number of hertz, in contrast to a pitch change
caused by changing the speed of a tape-recorded signal. In such a pitch
change, all the frequency components are shifted by a constant percentage,
and therefore, high frequencies are shifted proportionally more than lower
ones. A pitch shift by speed change thus preserves all the musical intervals
between the components. A true frequency shifter, on the other hand,
destroys the harmonic relation between the components. The sound of a con-
sonant musical tone becomes dissonant or bell-like or harsh, depending on
the amount of shift. Frequency shifters are used in electronic music synthe-
sizers for special effects that would be almost impossible to attain otherwise.
Some years ago, it was conjectured that if a small amount of frequency
shift was added to the amplied signal in a sound reinforcement system,
the result would be a reduction in the tendency for acoustic feedback,
because the amplied sound picked up by the microphone comes out of
the loudspeaker at a higher frequency. Small amounts of frequency shift,
about 4 or 5 hertz or so, are not very audible with speech sounds. Frequency
shifters for this use were made and sold in the 1960s. Unfortunately, the
theory did not hold in practice, and where there was feedback before, there
were little chirps at the ends of the words instead. This proved to be about
as annoying as the feedback!
Frequency-to-Voltage Conversion See pitch tracking.
Fringe Area A location, remote from the transmitter, where radio or tele-
vision reception is marginal due to low signal strength and high noise level.

166
 Fundamental

Sometimes a special high-gain antenna can pull in weak stations in fringe

areas.
Fringing Fringing is a rise in the level of low frequencies when a record-
ing is replayed by an analog tape head with a narrower track width than
the one used to record the tape.
The replay head is sensitive to the magnetic ux on the tape that is out-
side the actual track width of the head. Therefore, flat frequency
response is only maintained when the recorded track is the same width
as the reproduce head track. The effect only occurs at low frequencies, where
the recorded wavelength is long compared to the dimensions of the head.
Frontal-Incidence Microphone A microphone having an essentially
constant sensitivity over a wide frequency range when in a free acoustic
eld and pointed at the source of the sound. Also known as free-field
microphone.
Front End In a radio receiver, the amplication stages before hetero-
dyning takes place are collectively called the front end. The front end must
amplify signals from the antenna that cover a wide frequency range and
a very wide dynamic range. It is common to experience front-end over-
load with very strong stations at close range, and this causes distortion
and tuning difculties.
Fullcoat Motion picture lm, either 16 mm or 35 mm wide, that is coated
with magnetic material instead of a photographic emulsion. It is also known
as magstock. It is used for magnetic recording of lm sound tracks, and
because it has sprocket holes, it can be easily synchronized with the picture
when the nal sound track is assembled. See also double-system sound.
Full Duplex A communications channel, such as a telephone line, that
allows simultaneous transmissions in both directions is called a full duplex
channel.
Full Track An analog tape head conguration in which the entire width
of the tape is used for the recorded track. This, of course, applies to mon-
aural recorders only. Full-track recording achieves about 3 decibels bet-
ter signal-to-noise ratio than half-track recording.
Fundamental The lowest frequency component in a complex periodic
waveform. Any sound waveform that is perceived as having a musical
pitch is periodic, i.e., it has a shape that repeats itself. This sound will have
a series of harmonics, sometimes called overtones by musicians. (Actu-
ally, the rst overtone is the second harmonic, etc.) The lowest of these har-
monics, called the rst harmonic, is also called the fundamental. The
musical pitch of such a sound generally is determined by the fundamen-
tal frequency.
An interesting psychoacoustic effect of our hearing mechanism is that
we perceive the pitch of a complex wave as being that of the fundamental,
even if the fundamental is actually missing. In other words, as long as the
higher harmonics are present, the ear will reconstruct the missing funda-
mental. It does this by sensing the difference in frequency between the suc-
cessive harmonics. This difference frequency will be constant because the

167
Fundamental Tracking

harmonics are all at integral multiples of the fundamental frequency. This

phenomenon is called fundamental tracking, and is very important to
our subjective appreciation of music.
Most low-pitched musical instruments, such as the bassoon, bass viol,
tuba, etc., produce very little energy at the fundamental frequencies of the
lowest tones. It is easy to see how this must be, simply by the physical size
of the instruments. To radiate signicant energy, a sound source usually
must be large compared to the wavelengths involved. For instance, low A
on the piano has a frequency of 27 hertz, whose fundamental has a wave-
length of over 30 feet, and negligible energy is produced at this frequency.
The same is true for most instruments, although the organ is a notable excep-
tion. There are specially designed pipes in most organs that do produce large
amounts of very low frequency energy, and this is the reason the organ can
cause noticeable vibration of oors and walls in buildings.
This fundamental tracking allows us to perceive satisfying bass notes
emanating from relatively small instruments, and it also eases the demand
on sound reproducing systems. This explains the difculty of reproducing
the sound of the organ at realistic levels. Fundamental tracking also allows
us to enjoy music reproduced from very small loudspeakers and still hear
the bass tones. The sound of a symphony orchestra reproduced on the three-
inch speaker in a transistor radio bears very little resemblance to the orig-
inal wide-range sound; yet we are able to perceive the musical pitches very
well. Of course, we will recognize the fact that the low pitches are weak
and the sound is thin, but the pitches are still heard correctly.
Fundamental Tracking See fundamental.
Fuse A protective component for an electrically operated device, often con-
sisting of a glass-enclosed wire that will melt, or blow, and break the elec-
trical connection if the current through it is higher than its rated value.
There are many types of fuses, and they are rated for their current-carry-
ing capacity as well as the speed with which they will open up when over-
loaded. A fast-blow fuse may blow in 1 ms, whereas a slow-blow fuse
may take many seconds to open up with a small overload current.
Fuses operate because they have a relatively high resistance compared
with the rest of the circuit they protect, and they are nonlinear devices,
whose resistance changes with temperature. Some loudspeaker systems
are protected by fuses, usually slow-blow types, and they can affect the
damping of the system as well as increase the distortion.
Fusion In stereophonic music reproduction, sounds from two or more
loudspeakers can be perceived to be coming from some point between
the speakers. The multiple sounds are sometimes said to fuse into a sin-
gle sonic image. Probably the most common example of this fusion is the
apparent location of a solo instrument or voice in the center between two
speakers. The precision of the image location has a great deal to do with
the uniformity and similarity of the loudspeakers.
Fuzz Box A device that makes a virtue of distortion. It consists of a pre-
amplifier that is overdriven to produce clipping and consequently large

168
 Gain Riding

amounts of harmonic distortion. It is used as an effects device, most often

with electric guitars.
Fuzzy Distortion A type of distortion rst described by Prof. Malcolm
O. J. Hawksford, fuzzy distortion is caused by the fact that only a very few
electrons will enter a transistors base and contribute to signal trans-
mission when amplifying very low level signals. This is analogous to quan-
tization distortion in digital systems. It may be minimized by certain
circuit topologies.
FX Short for effects, meaning any form of audio signal processing or a
device to produce reverb, delay, chorusing, echo, flanging, phasing, etc.

G
Gaffer An electrician working on a lm or video stage is called a gaffer.
Gaffers Tape A very strong cloth adhesive tape used to temporarily secure
things such as cables to the oor. Gaffers never refer to this as duct tape.
There are a surprising number of opinions about the best characteristics of
such tape.
Gain The amount of increase in the power of a signal by an amplifier is
called the power gain. It is simply the ratio of the output power to the input
power and is conveniently expressed in decibels.
An amplier also usually increases the voltage and/or the current
of a signal, and these increases are called voltage gain and current gain,
respectively. They should be expressed in decibels only if the input and out-
put impedances of the amplier are the same, but this is seldom the case
in practice. Voltage and current gains should properly be referred to as sim-
ple numbers, that is, a voltage gain of 10 or 100, etc. Nevertheless, most engi-
neers express voltage gains of various types of ampliers in decibels
regardless of the fact that the input and output impedances are very differ-
ent. As long as they understand that the true power gain cannot be expressed
this way, no confusion seems to result. For more information, see decibel.
A passive device, such as a transformer, can have voltage gain or
current gain (but not both), even though it cannot amplify the signal.
Gain Bandwidth Product A gure of merit for certain active electronic
devices such as opamps. It is a numerical multiplication of the gain times
the bandwidth.
Gain Riding Gain riding is the variation in the volume control setting while
making a recording in order to prevent overload and distortion at loud
levels and to avoid noise problems at low levels.
If the gain is adjusted to maintain the signal level near the top of the
dynamic range of the recording device, the signal-to-noise ratio will
be maximized, but the dynamics, or loudness variations naturally present
in the signal, will be minimized. For most music, this is undesirable, but gain
riding still must be done to some extent in order to avoid overloads.

169
Gain Structure

limiters, compressors, and automatic gain controls are devices that

perform gain riding in specic ways without operator intervention.
It is important to note that gain riding will not prevent signal overloads
that can occur in circuit components that precede the gain control. A
common place of signal overload in recording consoles is the input stage
of the microphone preamplifier. Well-designed recording consoles
have overload detection circuitry built in, and red LEDs in each channel
warn the user of overloads in the preampliers. The user must adjust the
microphone preamplier so it is not overdriven, usually with a control that
attenuates the signal to the preamp. Any amplier can also have this prob-
lem, and the solution is always to attenuate the signal to not overdrive the
amplier, or reduce its gain.
It is also worth noting that the terms gain and volume control or
attenuation are often used interchangeably, as the effect often seems the
same.
Gain Structure The settings of all of the sequential amplication (gain)
stages in a device or system. A single device such as a mixer may have many
sequential gain stages within it, although some may not be user adjustable.
A large sound system may have many separate audio processors in its chain,
and each may be considered a gain stage in itself. Good gain structure is
the result of the analysis and setting of all gain stages in a device or system
to ensure that the system has full dynamic range, that no stage distorts
from clipping (overloading), and is high enough above the noise oor for
good signal-to-noise ratio. If a power amplifier and speaker is the
end of the chain, the amplier must have its sensitivity adjusted so that full
power is available without noise from excess sensitivity (gain). In a multi-
component audio system that has its gain structure adjusted correctly, if you
put a signal in at its input and gradually raise the level of that signal, all the
gain stages in the system will reach their clipping level at the same time.
Gain structure also refers to a devices internal gain stages that are part
of its design and not user adjustable.
Gang To mechanically couple two or more controls, such as potenti-
ometers, faders, or variable capacitors, is to gang them. The combined
assembly is then called a ganged fader, and so on. For example, a stereo
volume control consists of two ganged single-channel potentiometers.
Gap Scatter In analog tape recorder heads with more than one track,
the gaps associated with the tracks are positioned as nearly as possible in
a straight line, which ensures that the tracks are recorded and reproduced
at the same time. This is important in stereo recording to preserve local-
ization of sound sources.
Any deviation from perfect alignment is called gap scatter, and can be
measured by recording the same sinusoidal signal on both tracks of the
tape. The tape reels are then reversed and the tape reproduced backward.
The phase difference between the two reproduced signals is an indication
of the magnitude of gap scatter. For this test to be accurate, the head must
be in perfect azimuth alignment.

170
 Golden Ear

Gassy A condition resulting from an imperfect vacuum that can afict vac-
uum tubes, reducing their performance. When operating, a gassy tube will
usually show a purplish glow in the neighborhood of the plate. A gassy
tube is also said to be soft or at.
Gate A circuit that performs like a switch, allowing a signal either to
pass or not, is called a gate. The position of the gate (open or closed) is con-
trolled by an applied voltage, which can come from a number of differ-
ent places. If the level of the signal itself determines the gate opening, it is a
noise gate, closing when the signal level is so low that the noise would be
audible.
One of the three terminals of the field effect transistor is also called
the gate. The voltage on the gate acts as a control and determines the cur-
rent through the fet.
Gauss Aunit of magnetic ux density, equal to one ten-thousandth of a tesla.
Many loudspeaker manufacturers use this unit in their specications, but
the tesla is more often used by the scientic community. The gauss was named
in honor of the famous eighteenth-century German mathematician and physi-
cist Karl Friedrich Gauss, who investigated magnetism.
Getter A small metal tray or cup placed inside a vacuum tube and containing
a small quantity of metallic barium. When the tube is evacuated, the bar-
ium is heated and it quickly combines with the residual oxygen to make
barium oxide. This prevents the oxygen from gradually oxidizing the hot
elements in the tube, especially the heater.
Gibson Girl See splicer.
GIGO, Garbage In, Garbage Out A slang term meaning defective or noisy
data sent to the input of a device will result in defective or noisy output
from the device.
Glass Master A glass disc with a light-sensitive coating, whose surface is
to be etched with pits by a laser beam that is modulated by digital data that
represents the audio signal. This surface is then plated with a coating of
silver, and is then used as a master for the stampers from which cds are
eventually pressed.
Glitch An undesired voltage excursion found in a dac. See deglitcher.
Gnats Nut A vanishingly small distance or quantity.
GND Abbreviation for ground.
Gobo A movable panel or bafe used in a recording studio to isolate cer-
tain instruments from others in a group. The gobos are usually about three
or four feet high and may be made of wood with berglass on the surfaces,
or sometimes they are of solid foam polyurethane plastic.
Golden Ear A so-called Golden Ear is a person to whom is ascribed (usu-
ally self-ascribed) the ability to discern and appreciate subtleties and to iden-
tify defects in recordings and sound systems that ordinary people nd
elusive. Generally, Golden Ears are forever modifying and changing their
sound systems, much to the delight of the equipment industry. We have
been in search of the true Golden Ear for many years, and numerous can-
didates have come to light. Alas, they have all come up short when put to

171
Good Acoustics

the test. For instance, many such have failed to tell the difference between
stereo and monaural, or failed to detect when the channels were
reversed in a stereo presentation, etc.
This is not to denigrate the importance of careful and critical listening,
for our ears are truly marvelous measuring instruments when properly
trained and accustomed to evaluating sound. For instance, it has long been
known that objective measurements on sound systems may not correlate
with what one hearsan amplifier may seem nearly perfect when
measured and still sound faulty. Also, components that look identical to
measuring instruments often sound quite different. For this reason, mea-
surement techniques are constantly being rened, and new types of dis-
tortion are being discovered all the time. It is generally agreed that no
measurement technique thus far discovered can match the well-trained ear
in sensitivity to certain types of distortions.
The human hearing mechanism operates quite differently from micro-
phones and signal analyzers, and its physiological and psychological
aspects are far from being fully understood.
Good Acoustics See Appendix 4.
Gooseneck A exible, spiral, metal pipe about 14 inches long used to attach
a microphone to a microphone stand. The gooseneck allows the micro-
phone to be oriented in almost any direction. It would be nice if goosenecks
would allow the microphone to be moved without introducing any noise,
but they generally squeak when bent.
Gradient Microphone A gradient microphone is sensitive to the pressure
gradient (or variation in pressure over a distance) of a sound eld. The rib-
bon microphone is an example. Gradient microphones have a polar
pattern resembling a gure 8. This curve is the polar plot of the math-
ematical cosine curve, and gradient microphones are sometimes called
cosine microphones. See also figure-8 microphone.
Grain A subtle type of distortion found in some audio devices, mostly
digital devices but sometimes also power amplifiers. See also
granulation.
Grammy A recording industry award presented by the Recording Acad-
emy, also known as the National Association of Recording Arts and Sci-
ences, NARAS.
Gramophone The Gramophone was the rst recorder-reproducer to use a
at disc rather than a cylinder as the medium. It was invented by Emile
Berliner and commercially produced in 1893.
The Berliner disc introduced important differences from the Edison cylin-
der phonograph. It used lateral rather than vertical modulation, and the
discs were recorded with a stylus scratching through a soft wax layer on
a zinc disc, reducing the work the stylus had to do. The groove in the disc
was then formed by acid etching. These techniques increased the ampli-
tude of modulation, making the records sound louder than Edisons. The
introduction of the disc meant also that records could be easily mass-
produced by stamping, which was much cheaper than the complex mold-

172
 Ground

ing procedure Edison used to reproduce cylinders. No doubt the explosive

growth of the recording industry was due to the Berliner innovations.
The rst Berliner discs sounded like a partially educated parrot with
a sore throat and a cold in the head, according to Eldridge Johnson, founder
of the Victor Talking Machine company.
Granulation distortion in a digital audio system due to the uncertainty
in the level of the samples is known as granulation; also called quanti-
zation distortion.
If the sampling rate is an exact multiple of the input tone frequency,
granulation results in harmonic distortion, i.e., the distortion compo-
nents are at multiples of the input frequency. If not, the granulation resem-
bles random noise, with a broad nondiscrete spectrum, in which case it
may properly be called quantization noise.
Gravitation The tendency of all bodies to approach one another with a
strength proportional to the quantity of matter they contain, the quantity
of matter they contain being ascertained by the strength of their tendency
to approach one another. This is a lovely and edifying example of how sci-
ence, making A the proof of B, makes B the proof of A.A. Bierce
Gravity Cell An essentially obsolete type of voltaic cell that uses zinc
and copper electrodes and copper sulfate and zinc sulfate as the elec-
trolyte. The cell is arranged with the copper electrode on the bottom sur-
rounded by a concentrated solution of copper sulfate, and the zinc electrode
on the top surrounded by the lighter solution of zinc sulfate. Gravity keeps
the two sections of electrolyte from mixing, hence the name gravity cell.
The zinc electrode usually looks like a crows foot, so the cell is also some-
times called a crowfoot cell. The gravity cell is unusual in that it becomes
discharged when no current exists through it but does not become dis-
charged under full-current conditions. For this reason, it is called a closed
circuit cell and is kept with a shorting strap across it when not in use. It
was very well adapted for use in telegraphy, and this is where it saw most
of its use.
Green Book See red book.
Grid In a tube, the element that controls the electric current is the grid.
It is usually made as a series of ne wires, and is placed between the cath-
ode, or emitter of electrons, and the plate, or collector of electrons. A
tube may contain several grids, but normally only one, called the control
grid, actually controls the amount of current in the tube. See also audion.
Grip A stagehand working on a lm or video stage is called a grip.
Groove Guard The result of shaping the cross section of stereo records
so the groove area is somewhat protected by raised edges and centers. If
records are stacked one on top of another, as is the case with a record changer,
the groove areas are prevented from coming into contact.
RCA Records invented and trademarked the term Groove Guard.
Ground In an amplifier or other audio component, the metal chassis
is called the ground, whether or not it is connected to the actual ground.
One of the signal leads is usually connected to ground, as is one of the

173
Ground Lifter

power supply leads. Ground is therefore sometimes called a common.

In Britain, the term earth is used instead.
An electronic chassis is made of metal because metal is a conductor of
electricity; the chassis therefore acts as a shield, keeping external elec-
trostatic elds out of the interior. Such electrostatic elds would induce
noise in the circuits, and shielding greatly reduces this. The shielding
effect of the chassis is extended to the signal wires by using shielded cables,
which have an outer conductor completely surrounding the inner conductor.
A coaxial television cable is an example. The outer conductor, or shield, is
connected to the chassis ground. If two or more components are connected
together, such as a preamp and an amplifier, the shield in the cable usu-
ally serves to connect the two chassis together, making all the grounds com-
mon. If this ground connection is broken, very large amounts of 60-hertz
noise, or hum can be introduced. This is because the 60-Hz power lines,
which are nearly everywhere, radiate strong electric elds at 60 Hz, just
like radio transmitting towers.
Ground Lifter A switch found on some audio adapter boxes, or direct
boxes, interrupts the continuity of the shield connection between the out-
put and input. This disconnects the chassis ground of a device like a gui-
tar amplifier from the ground of the recording console or tape recorder,
reducing the possibility of hum induced by ground loops.
Ground Loop See shield.
Group Delay The slope of the phase-versus-frequency curve of a fre-
quency response function, i.e., the rate of change of phase of the response
as a function of frequency. Group delay is a property of a device or a system.
A pure time delay, equal at all frequencies, gives a constant slope of phase
versus frequency. If in an audio component this slope is not constant, but
varies with frequency, the component is said to produce group delay dis-
tortion. This is equivalent to a time delay that varies with frequency. For
instance, an anti-aliasing filter will typically have a phase response
curve that slopes sharply down at high frequencies, which means that the
high-frequency components will be delayed longer in their passage through
the lter. The audible result is a loss of precision in musical transients;
they are spread out, or smeared, in time, and a more diffuse stereo image
results.
By using an all-pass network, it is possible to correct the group delay
distortion of such lters, but this is seldom done in practice. (See all-pass
filter.) The most effective way of measuring group delay is to use the dual-
channel fft analyzer, which displays the phase response directly as a
function of frequency.
Group Velocity Group velocity is the speed with which a group of waves
progresses through a medium.
For some transmission media, it is found that signals of different fre-
quencies travel at different speeds. Examples of this are ocean waves, where
shorter wavelengths (higher frequencies) travel faster than longer ones;
and light waves traveling in glass, where shorter waves travel slower. This

174
 Haas Effect

phenomenon is called dispersion, and such media are called dispersive.

Dispersion is what causes a prism of glass to separate the various frequencies
of light into a spectrum.
Consider a signal made up of a short burst of a single-frequency tone.
The tone frequency will be associated with a relatively short wavelength,
while the burst itself constitutes a disturbance of lower frequency. The tone
will then travel at a different speed than the burst, and the individual waves
of a tone will be seen to move through the burst. This happens with ocean
waves; the small waves can be observed as they begin at the rear of the
group and peter out at the leading edge of the group.
This, luckily, does not happen with sound waves in air. If it did, music
would be aural anguish, even more than some music already is. However,
some audio devices are dispersive to electrical signals, and this does cause
distortion of the signals. A transient, for instance, is necessarily made
up of many different frequencies, and if they travel at different rates in a
device, the transient will become spread out in time (or smeared) as it
progresses through the device, and it will not sound the same coming out
as it did going in. The distortion produced by such a device is sometimes
called group delay distortion. All filters produce this type of distor-
tion to some extent, although there is no agreement among audio people
about the audibility of small amounts of it.
Grunge A rather ill-dened term for a type of distortion in audio equip-
ment that is difcult to quantify and measure. It is also sometimes called
mid-range smear, which is at least more descriptive. It is caused by the inter-
action of the many frequencies present in a music signal. The result is a broad
band of nonlinear distortion products.
Guard Band A narrow unrecorded area between the recorded tracks of a
magnetic tape. The guard band reduces crosstalk between the chan-
nels of the tape recorder. See also half-track.
GUI Graphic User Interface. A computer operating system that uses sym-
bols on the screen and a pointing device to move objects around and acti-
vate and manipulate les without using the keyboard. The Apple Macintosh
operating system and the various Microsoft Windows operating systems
use GUIs. The principal of the GUI was probably invented by the Xerox
Corp., who sold the rights to use it to Apple Computer.
Gyrator A gyrator is a circuit topology that uses only capacitors and
resistors along with amplifying stages, but acts like an inductor.
In some cases it is less expensive to build a gyrator circuit than to use
an actual inductor, depending on the value of the required inductance.

H
Haas Effect The Haas effect, also called the precedence effect, is related to
the localization of the apparent sonic image when the same signal is pre-

175
Half Normalled

sented to the two ears at slightly different times. If a short signal such as a
click is presented over earphones to one ear and then to the other ear a few
milliseconds later, the human hearing mechanism will judge the sound to
be coming from the side of the head where the earliest sound arrived. For
no delay between the sounds reaching the ears, the sound is localized
straight ahead, or sometimes within the head itself, and as the delay is
increased, the image moves farther and farther to the side where the earli-
est sound occurred. This is true up to about 25 to 35 milliseconds, after which
more delay will result in two distinct sounds being heard.
The Haas effect also means that if a sound arrives at a listeners ears from
two locations, as is the case with sound reinforcement systems with sev-
eral loudspeakers, the sound will be localized at the loudspeaker that pro-
vides the earliest arriving sound, and the other speaker will not be heard
at all. This is true even if the delayed sound is stronger than the rst sound.
The difference in level may be as much as 10 decibels, and the delayed
sound will still be inaudible.
The Haas effect is an example of sensory inhibition, where the response
to a stimulus causes the response to another stimulus to be inhibited. Sen-
sory inhibition, as the name implies, is a characteristic of other senses besides
hearing. The science of psychophysics deals with these phenomena.
Half Normalled See patch bay.
Half-Power Bandwidth A standardized method of stating the band-
width of a bandpass or band reject filter. The half-power bandwidth
is the higher frequency where the response is 3 dB lower than the max-
imum minus the lower frequency where the response is 3 dB lower than
the maximum. A 3-dB reduction in level is, of course, a power reduction
of one-half. See decibel.
Half-Step The musical interval of a minor second in a diatonic scale. In
just intonation, the minor second has a frequency ratio of 15/16, and
in the equal tempered scale, the minor second has a frequency ratio of the
12th root of 2, or about 6%. See Appendix 8.
Half-Track An analog tape head conguration that assigns half the tape
width to each audio channel. In monaural recorders, the single chan-
nel uses one-half the tape width, allowing the other half to be recorded when
the tape reels are ipped over and reversed. Strictly speaking, the track
widths are a little less than one-half the tape width. There is a narrow band
down the center of the tape that is not recorded in order to reduce
crosstalk. This is called a guard band.
Handshaking The initial exchange of binary data between two digital
devices or systems that establishes proper communications between them.
A common example is the noises made by the computer modem when rst
establishing a link with another modem.
Hamster Switch A control found on professional disk jockey performance
mixers that reverses fader action. For example, if a fader normally is off
at the bottom of its travel and on at the top of its travel, then activating the
hamster switch reverses this, so off is now at the top and on is at the bot-

176
 Harmonic Distortion

tom of travel, or alternatively, it swaps left for right in horizontally mounted

faders. It is used to create the most comfortable (and fastest) fader access
when using either turntable, and to accommodate left-handed and right-
handed performers. Credited to, and named after, one of the original
scratch-style crews named The BulletProof Scratch Hamsters.15
Hangover When a signal into a loudspeaker suddenly stops, as hap-
pens with musical transients, sometimes the loudspeaker cone will con-
tinue to move and produce sound. This is called hangover. It is both a
low-frequency and a high-frequency phenomenon, and is reduced by
adding damping to the system. One way of doing this is to increase the
damping factor of the amplifier so it has better control of the cone.
Hangover causes poorly damped woofer systems to sound boomy, and
poorly damped tweeters to sound hissy.

Hangover

Hard Clipping clipping where the edges of the waveform are very sharp,
producing the maximum amount of high harmonic content.
Hardware Key See dongle.
Hard-Wired Permanently connected with wire, rather than being connected
by a removable cable or plugs and jacks.
Harmonic Distortion In a perfect audio device, such as an amplifier or

15. Thanks to Dennis Bohn, Rane Corp.

177
Harmonic Series

tape recorder, the output signal would be a replica of the input signal with
no changes except maybe power level. The perfect device does not exist, how-
ever, and the output signal will always have some distortion when com-
pared to the input signal. The simplest form of this distortion consists of
harmonics of the input signal being added to the output signal. This is called
harmonic distortion, and is caused by the system not being perfectly linear.
Harmonic distortion is usually measured as a percent. An amplier put-
ting out 10 volts at 1,000 hertz and adding 1 V of 2,000 Hz is said to have
10 percent of second harmonic distortion. Harmonic distortion is best mea-
sured with a spectrum analyzer, where the amounts of the various har-
monics are shown. The summed levels of all the added harmonics is called
the total harmonic distortion, or thd, and is usually expressed as a percentage
of the level of the signal being measured. This signal is typically a sine wave.
The percentage of one harmonic, typically the third, may also be reported
by itself, and would be denoted as HDL3 for 3rd harmonic distortion level.
Various devices contribute different types of harmonic distortion; for
instance, an analog tape recorder adds odd-order harmonics (primarily
third) almost exclusively, while a typical amplier will add both even- and
odd-order harmonics. tube-type ampliers add lower-order harmonics,
and transistor-type ampliers tend to add higher-order harmonics. This
is at least partially responsible for the differences in sound between dif-
ferent types of audio devices. When the THD consists of second harmon-
ics, the sound is loud and brassy. Some ampliers and preamps cause
this type of distortion when overloaded, and this has often been deliber-
ately exploited to increase the punch of certain recordings.
Third harmonic distortion, on the other hand, creates a somewhat cov-
ered sound, generally considered undesirable. See also aphex aural
exciter.
Harmonic Series A harmonic series is a group of frequencies, each one of
which is an integral multiple of the frequency of the lowest one, or fun-
damental. See also overtones.

Harmonic Series Spectrum

178
 Head Bump

HATS Head And Torso Simulator, or a dummy head but with the torso.
It is used to more accurately simulate the acoustics near a human upper body.
HDCD High Denition Compatible Digital. An encode/decode system for
music cds and dvds that encompasses up to 176.4-kHz sampling rate and
24-bit resolution. The system was introduced by Pacic Microsonics, before
it became a subsidiary of Microsoft, and it is compatible with standard red
book CDs. HDCDs can thus be played on conventional CD players and
are claimed to attain lower noise levels and somewhat less distortion.
It is also claimed that conventional CDs can be played on HDCD players
with improved noise and distortion gures.
HDR Hard Disc Recorder. An audio recording device that digitizes the
audio and records directly onto a computer-type hard disc. Several
congurations are available, usually with multichannel capability. HDRs
are often included in Digital Audio Workstations, daws.
HDTV, High Denition Television Television that has about twice the res-
olution or more of the original analog systems like ntsc, pal, and secam,
with multichannel audio. HDTV could theoretically be analog, but consumer
HDTV systems are digitally transmitted. Display of the picture can be ana-
log (similar to a computer CRT display) or digital (similar to a digital at
panel computer display). The analog NTSC system transmits pictures of
525 interlaced scan lines, of which about 480 are visible. The ATSC HDTV
system has several modes, and includes 1080 interlaced scan lines (or
vertical pixels), called 1080i, and 720 progressive scan lines (720p). ATSC
can also transmit non-HDTV at 480 lines.
Head See tape head.
Head Amplier, Head Amp Synonymous with pre-preamplifier; mostly
British usage.
Head Bump A series of irregularities, or bumps, in the low-frequency
response of analog tape recorders sometimes can be over 2 decibels in

Head Bump Effect

179
Headphones

amplitude and are worse at higher tape speeds. They are essentially non-
existent in cassette recorders because of the low tape speed.
Head bumps are caused by the fact that the playback head as a whole
responds to very long wavelengths recorded on the tape, while only the
gap in the head responds to the shorter wavelengths. At certain frequen-
cies (corresponding to specic recorded wavelengths), the overall head
response will be in phase with the gap response, and the overall response
is increased. At other frequencies, the two responses will be out of phase
and will partially cancel, causing a dip in the overall response. As frequency
increases, the effect becomes less and less, and the bumps fade into a smooth
frequency response curve.
The effect is actually dependent on the recorded wavelength rather than
the absolute frequency, and since low tape speeds mean short recorded
wavelengths, the effect is only apparent below the audible frequency
range. As tape speed increases, the effect occurs at higher and higher fre-
quencies, and can be a real problem at 30 ips.
Careful head design and attention to the amount of tape wrap around
the head can reduce the effect.
Headphones Headphones are miniature loudspeaker-like sound repro-
ducers designed to be worn over the ears for private listening to audio
signals. Also known as earphones and headset.
There are many types of headphones, but probably the most common
is the dynamic headset, which is similar to a dynamic loudspeaker in oper-
ation. Another type is the electrostatic, which is basically similar to the elec-
trostatic loudspeaker. A newer type is the planar dynamic, which uses a
perforated magnet assembly and a diaphragm with a metallic conductor
embedded in it. It is similar to the leaf tweeters in operation and can
provide very extended, smooth response. Unfortunately, the planar type
of headset is seldom seen today, probably because it is expensive to man-
ufacture.
Some headphones require a good seal between the earpiece itself and
the head to achieve good bass response. Such a headset is called circum-
aural, meaning it ts over the ear. Dynamic headsets are usually in this
category. Other units, sometimes called open air phones, are designed
to rest on the outer surface of the ears and do not require a seal for at bass
response. These sets are often very light in weight and can be of the planar
type, or dynamic. This category of headset is called supra-aural, mean-
ing upon the ear. They do not reduce the interference from ambient noise
as do the circumaural units.
It is an unfortunate fact that the frequency response of a headset is
a function of the anatomy of the ear on which it rests due to reections in
the ear and variations in the volume of the cavity between the ear and phone.
This means each person hears a different tonal balance when listening to
the same headset. The effect is measured by inserting a tiny probe micro-
phone next to the eardrum of the listener and performing a frequency sweep
with various headphones. The variations in frequency response can exceed

180
 Heads Out

10 decibels at certain frequencies. Perhaps this is a reason that various

people prefer different types of headphones.
It is interesting that audiometer testing for hearing loss almost always
uses headphones to present the sounds to the subject. The phones are stan-
dardized, and must be measured with a specic acoustic coupler with the
measurement microphone, but this does not ensure that the subject will have
a uniform sound level at the ear drum. This means audiometer testing is
subject to at least a 10-dB variation. It would be much better to do audiome-
ter testing under free-field conditions; when this is done, almost all sub-
jects have a lower measured hearing threshold.
Headroom Headroom in an audio device is the difference in level between
the highest level present in a given signal and the maximum level the
device can handle without noticeable distortion. Music, of course,
exhibits very wide variations in level, even within a single selection (except
perhaps rock music, which seems to have a dynamic range of 3 decibels
or so), and any audio device must handle the maximum expected level in
order to sound distortion-free.
Less obvious is that on an instantaneous basis, music waveforms have
short peaks that are much higher in level than the average signal level. This
is especially true of percussive sounds in music. These short peaks are not
registered by most audio level reading devices, such as the vu meters often
seen on analog tape recorders. If a vu meter is used to set the recording
level, the temptation will be to set the level too high, allowing the peaks to
overdrive the tape, causing distortion. In most audio equipment of high
quality, the maximum signal level without overload will be at least 10 dB
above the maximum level shown on the meter, i.e., there will be 10 dB of
headroom. Nearly all digital recording and signal processing equipment
of recent vintage uses solid-state circuitry with a string of leds to indi-
cate the signal level. They usually show the peak level of the signal rather
than the average level, and are much better at letting the engineer know
how close to overload the signal is. However, there is no agreed-upon stan-
dard for the response rise and fall times and peak-hold times of these dig-
ital meters, so the engineer must be sure to educate himself in exactly how
his meters behave.
Headshell The plug-in holder for a phono cartridge. The headshell plugs
in to the tonearm and the cartridge is mounted in the headshell with
small screws.
It is a painstaking job to properly mount a phono cartridge and be sure
that all the distances and angles are correct for proper tracking of the record
groove. Small errors in vertical tracking angle and perpendicularity of the
stylus cantilever with the record surface will cause signicant distor-
tion of the reproduced sound.
The p-mount standard greatly simplies the installation of cartridges
by making everything pre-measured and preset.
Heads Out A reel-to-reel analog tape recording that has been rewound
and is ready to play. It is generally considered best for long-term storage

181
Head Stack

to leave recordings in the un-rewound condition (tails out) for minimum

print-through.
Head Stack The assembly of tape heads in an analog magnetic recorder.
The head stack normally consists of an erase head, a record head, and a
playback head. They are mounted close together in a metal housing that
provides magnetic shielding for them.
Hear Out To hear, by careful listening, the various pitch components of
a complex tone. It takes a good deal of practice to be able to do this. Also
called analytic listening.
Heater In an electronic tube, the cathode must be heated to a red heat
in order to boil electrons out of it. There is a small resistance wire inside it
to accomplish this. Sometimes, the heater is the cathode itself; in this case,
it is called a directly heated cathode. The rst tubes were of this type.
The heaters in a tube-type device are responsible for most of the power
dissipation in the device and are the reason the device must be designed
for adequate cooling.
Heat Sink A metal structure, usually aluminum, on which power tran-
sistors are mounted. The heat sink conducts heat away from the transis-
tors, effectively reducing their temperature. The heat sink is actually more
of a heat radiator than a true heat sink, and in some ampliers it occupies
a prominent place in the sides or back of the chassis.
Helical Scan A common method of tape recording where a rotating head
drum writes and reads slanted tracks of information on magnetic tape.
This greatly increases the amount of information that can be recorded. The
head drum is set at an angle relative to the tape travel, hence would draw
a helix, or 3-D spiral. All modern videotape recorders, r-dat, DTRS and
adat tape machines use this technique.
Helmholtz Hermann von Helmholtz was a nineteenth-century German
physician and acoustician who wrote a ne book on psychoacoustics
called On the Sensations of Tone. It is still used as a reference today. See also
helmholtz resonator.
Helmholtz Resonator A type of acoustical resonator consisting of an
enclosed volume of air connected to the atmosphere by a short channel or
pipe. The natural springiness of the enclosed air reacts with the mass of air
in the pipe, resulting in a sharply tuned resonance, whose frequency
depends inversely on the enclosed volume. Helmholtz resonators are found
in some loudspeaker systems. See bass reflex and ducted port. Her-
mann von Helmholtz, the great German acoustician of the nineteenth cen-
tury, invented the Helmholtz resonator to detect the individual component
frequencies in complex musical sounds. He made them in a variety of sizes
and, by holding them to his ear one after another, was able to hear out
the harmonics produced by musical instruments and speech. This was prob-
ably the rst scientic attempt at frequency analysis of sound, although the
principle of the Helmholtz resonator was discovered at least as early as 79
b.c. by the Roman architect Vitruvius. See also sounding vessels.
Hertz, Hz The internationally agreed-upon symbol to indicate frequency

182
 Heyser Spiral

in cycles per second. It is named after Heinrich Hertz, the famous nineteenth-
century German physicist who rst investigated radio waves.
In 1935, the iec proposed that hertz (Hz) be used instead of cycles per
second or cps, and some European countries did so. In 1948, the Gen-
eral Conference on Weights and Measures adopted the term into the si met-
ric system. Many people continued to use cps into the 1960s.
Heterodyne Another name for amplitude modulation. The process of
amplitude modulation is actually the instantaneous multiplication of one
signal by another. This results in the formation of sidebands that con-
tain the same information as the original signals but translated upward and
downward in frequency.
The term heterodyne is used for a frequency translation circuit in am
and fm receivers. The signal from the station is amplied and multiplied
by a signal from a local oscillator, and this translates the received fre-
quency down to a relatively low frequency called the intermediate fre-
quency, or IF. This signal is amplied and detected to recover the original
audio. By varying the local oscillator frequency, any radio station signal
can be translated, or heterodyned, to the IF, meaning the amplifiers and
detector need to operate only at this frequency. A receiver using this prin-
ciple is called a supersonic heterodyne, or superhet, receiver. The prin-
ciple was patented in 1920 by Edwin Armstrong, who also invented the fm
discriminator.
Heterodyning is also used in some spectrum analyzers and to a cer-
tain extent in electronic music, where it is called frequency shifting.
Heyser Spiral Named for Richard Heyser (19311987), who rst used time
delay spectrometry (TDS) for analysis of audio signals. The Heyser Spi-
ral is a graphic representation, or plot, of a so-called analytic signal,
which is one result of a tef measurement. The three dimensions of an ana-
lytic signal are time, the real part, and the imaginary part of the amplitude
in a three-dimensional curve. The Heyser Spiral plot of an analytic sine wave
or cosine wave is a constant-diameter spiral, shown below.

Heyser Spiral Display of a Sine Wave

183
High-Cut Filter

The same type of plot is often used to visualize the electrical imped-
ance of audio devices such as loudspeakers. The impedance of most
audio devices produces phase shift of the audio signal that varies with
frequency, and the actual phase shift angle is decomposed into two parts,
the so-called imaginary (or quadrature) component and the real (or in-
phase) component. The Heyser Spiral of an impedance is usually a two-
dimensional polar plot of the real amplitude versus the imaginary
amplitude as a function of frequency. The frequency axis goes through the
origin of the plot and is 90 degrees from the plane of the paper. If an audio
device has a resonance at a particular frequency, the spiral will show a
circle, and if several resonances exist, they will each have a circle in the plot.
The size of the circles indicates the strength of the individual resonances.
High-Cut Filter A low-pass filter, according to Dennis Bohn, of the Rane
Corp.
High End, The The segment of the consumer audio/video industry directed
to hard core audiophiles and videophiles is called the high end, pos-
sibly in reference to the prices asked for the equipment.
High Fidelity, Hi-Fi A term that became popular in the 1950s meaning a
relatively high quality sound system for consumer use. The term stereo
started to supplant it in the late 1950s.
High-Pass Filter A high-pass lter uniformly passes signals above a cer-
tain frequency, called the cutoff frequency. The cutoff frequency is where
the lter response is 3 decibels below the nominal response. The response
rolloff in the stopband may be gradual or sharp.
The rumble lter found in many record player systems is a high-pass
lter. Another example is the bass cut lter on mixing consoles, used to
reduce boomyness in certain mics and low-frequency room noise.

High-Pass Filter

184
 Holophonics

High-Speed Dubbing The copying of analog tape recordings at tape

speeds higher than that used for the original recordings. If a tape is played
at twice normal speed, all the frequencies will be doubled, and everything
will sound an octave too high as well as twice as fast. If this signal is
recorded onto another tape and then replayed at normal speed, the fre-
quencies all come out correct, but it will take only half as long to make the
copy. In like manner, if the original is played at eight times normal speed,
it will only take one eighth the time, etc.
When high-speed dubbing is done, it is important that the player and
the recorder be able to handle the higher frequencies without distortion,
and the pre-emphasis and de-emphasis curves must be different from
normal-speed machines. Virtually all commercially available cassettes and
tapes have been made this way using special dubbing equipment, and the
quality can be very good; however, it is thought that the very best quality
copies can only be made at the normal speed.
Hill and Dale A descriptive term for vertical, rather than lateral, modula-
tion of the groove on a phonograph record or cylinder. Thomas Edison was
the champion of vertical modulation, even though it was well established
quite early that lateral modulation results in less distortion because the
cutting stylus experiences more uniform mechanical resistance when cut-
ting the record. Stereo records, of course, are cut with a combination of ver-
tical and lateral motion.
Hi-Pass Filter Short for high-pass filter.
Hold-Back Tension Also called back tension. The hold-back tension is the
tension in the recording tape that holds it in contact with the tape heads
or drum. Hold-back tension can be supplied by simply applying a small
braking force to the supply reel on the tape recorder, or it can be imparted
by passing some current through the rewind motor, providing a torque on
the reel.
The hold-back tension should be fairly constant in any tape recorder to
ensure that the tape speed and the tapes contact with the heads will be
uniform, but constant torque on the supply reel gives greater tension as the
tape pack is reduced in size, causing the tape speed to gradually slow down
toward the end of the reel. More sophisticated tape recorders use one of
several techniques to provide constant tension regardless of the size of the
tape pack on the reel.
Hole in the Middle A phenomenon of two-channel a-b stereo that
causes a performer in the center of the reproduced sound stage to sound
farther away than she should. It is caused by the fact that the reproduced
level of a centrally located performer is a little too low and a little too rever-
berant because of relatively longer distances to the two microphones. The
phantom center channel was an attempt to reduce the effect, but the
proper way to do it is to use three independent channels. This was shown
conclusively by Harvey Fletcher in his stereophonic transmission experi-
ments in 1933. The effect has also been called the recession of the middle.
Holophonics An acoustical recording and broadcast technology claimed

185
Hookup

to be the aural equivalent to holography, hence the name. Holophonics is

an encode process that occurs during the recording session using a special
listening device named Ringo. It is claimed that playback or broadcast
is possible over headphones or any existing mono or stereo speaker sys-
tem, with various levels of spatial effect. Optimal effects occur when two
tracks (stereo) are played utilizing digital technology over headphones and
minimal effect when played over a single mono speaker (two tracks merged
into one and played over a single speaker).16
Hookup The interconnection of audio devices is sometimes called the
hookup, and the wiring inside the chassis of a device is also called the
hookup. Two components in a circuit that are connected with a wire are
said to be hooked to each other.
Horn In reference to acoustic phonographs, horns have been used from the
start to couple vibrations of the diaphragm to the air; in this, they resem-
ble the ear trumpet.
In reference to loudspeakers, a horn is a smoothly tapered tube con-
necting the loudspeaker diaphragm at the small end to the air in the room
at the large end. Most horns of this type have a cross-sectional area that
doubles for each unit of distance along its length. Such a horn is said to
have an exponential are rate.
Another type of horn is the conical or straight-sided horn, which has an

Horn Loudspeaker

16. Thanks to Dennis Bohn, Rane Corp.

186
 Horn

area that increases in a linear fashion with distance. Conical horns were
sometimes used in old acoustic phonographs. Exponential horns are much
more efcient and uniform in their characteristics than conical horns, and
although they are more difcult to build, they are the type generally used.
The theory behind the exponential horn was rst expounded by J. P. Max-
eld and H. C. Harrison of Bell Telephone Labs, and was rst used in the
Orthophonic Victrola in about 1925.
High-frequency loudspeakers benet from the use of horns because the
horn acts as an acoustical transformer, coupling the relatively high mechan-
ical impedance of the diaphragm to the low acoustic impedance of the
air. The horn, by improving the coupling of the diaphragm to the sur-
rounding air, greatly improves the efciency of the loudspeaker. Such a horn
operates well over a frequency range above a certain frequency called
the cutoff frequency, where the horn no longer presents an acoustic
load to the diaphragm. Without an acoustic load the diaphragm cannot
transfer any energy to the horn, and it is free to move without restriction.
The lower the cutoff frequency, the longer and larger the horn must be. This
means that for low-frequency loudspeakers (below 300 hertz or so) horn
loading is seldom used, except in large theater systems. In some cases, the
horn may be folded, or coiled around itself to make it more compact. Var-
ious ingenious designs of this type have been introduced over the years,
and some are still in use.
Because a horn usually is called upon to radiate frequencies whose wave-
length is short compared to the dimensions of the mouth of the horn, its
radiation will be quite directional at high frequencies. This characteristic
is desirable in sound reinforcement systems, where the energy must be
aimed, or beamed, at the audience to increase efciency and to reduce
the excitation of the reverberation in the room.
For home use, the high efciency of horns is seldom needed, and their
directionality is usually undesirable, so their use is somewhat limited,
although this point is hotly debated by some sound system acionados.
High-frequency horns also do not have quite as smooth a frequency
response as do direct radiators, and they produce signicantly
greater distortion as well, especially at high sound levels.
Another use of the term horns is in connection with the making of phono-
graph records. When the master record is cut into the soft acetate of the
disc by the hot cutting stylus, the slight melting of the acetate gives rise to
tiny mounds, or horns, along the sides of the groove. These horns are usu-
ally curved over at the top, and they cause difculty in separating the acetate
from the plated metal master or, in the third step, separating the stamper
from the mother. If left on the stamper, the pressings retain them to some
degree. They then often break off, causing debris to be left in the groove,
which makes for noisy playback of the record.
There has been a mild controversy going on about what to do about
horns. It is possible to remove the horns from the mother by carefully pol-
ishing with a cotton ball and a mild polishing agent, but some people in

187
House Mix

the industry say that a high-frequency loss is the inevitable result. Reduc-
ing the cutting stylus temperature will reduce the formation of horns in
the acetate, but this also increases the noise. The making of records is as
much an art as it is a science.
House Mix An output on a sound reinforcement control console that is
used to feed the power ampliers for the loudspeakers in the main audi-
torium, or house. The house mix will typically be quite different from
the signal used to feed the stage monitors, or foldback speakers. See also
monitor mix.
House Mixer See foh.
House Synch A master timing reference signal used to ensure synchronized
operation of timing-sensitive devices such as television equipment and dig-
ital audio. Original analog television facilities realized that unless cam-
eras, switchers, effects generators, and so on were all synchronized from a
common signal, switching the video signal around between devices would
cause a glitch. The solution is a precision signal used by all devices so they
all know exactly when the video frame starts, and this is called House Synch
because it is used by all devices in the facility, or house. The format of
signal used in the U.S. is usually black burst, which is a composite video
(meaning it contains picture, horizontal and vertical sync signals, and the
color burst or reference waveform) with a black picture.
In the digital world there can be a similar need for synchronization of
multiple devices in a facility, as well as more accurate clocking. Audible
pops and clicks can result from transfering digital data among devices whose
clocks are not quite in sync, and many feel that distortions from clock jit-
ter are exacerbated by multiple reclockings at slightly varying rates. Digi-
tal clock devices usually offer several options, such as word clock (the same
rate as the sampling frequency in use) and integer multiples of word clock.
The old standby, black burst, can also be used.
There is a mild controversy about the term sync versus synch, and both
spellings are seen.
Howlback, Howlround British terms for acoustic feedback.
HPS-4000 A multichannel high-powered sound system custom-designed
and installed in theaters by John Allen. The system is said to use three-way
horn-loaded loudspeakers exclusively and many more surround speak-
ers and subwoofers than are commonly used. The HPS-4000 specications
state each screen speaker will produce 112 dB Sound Pressure Level at 35
feet from the speakers.
HRTF Head Related Transfer Function, a mathematical model of the acoustic
response of the human head. It is used for research in sound elds, often
using dummy heads or hats.
HT Acronym for high tension, which is British usage for high voltage.
Hub The circular center section of the reel on which magnetic tape is wound;
the remainder of the reel consists of the anges. In recording studios, tape
was often purchased wound on hubs without the anges, resulting in
signicant savings. See also pancake.

188
 Hum Bucking

Hum A 60-hertz or 120-Hz frequency component in an audio signal

is called hum. One of the most common types of noise encountered in audio
systems is hum from the 60-Hz power-line frequency, and its elimina-
tion or reduction to inaudible levels is sometimes quite difcult.
Hum is introduced into audio signals by several different mechanisms.
The power lines delivering electrical energy around the country behave like
giant transmitter antennas and radiate electromagnetic waves at 60 Hz.
Sometimes, the strength of these emissions is quite large, especially in loca-
tions adjacent to high-voltage power lines. This radiation induces 60-Hz
currents in audio cables by electrostatic and electromagnetic induction,
causing the familiar humming sound at a musical pitch a little below low
B on the piano keyboard. To prevent electrostatic induction, audio cables
have an electrostatic shield surrounding them; and to prevent electro-
magnetic induction, long audio cables are balanced and the two signal-
carrying wires are twisted together. See also Appendix 6.
Another route for hum to enter audio signals is via the power supply
found in virtually all audio devices. The power transformer radiates a local
60-Hz magnetic eld, which can cause electromagnetic induction, especially
into nearby tape heads or phono cartridges. Sometimes it is possible
to reduce this type of hum by orienting the power transformer in a differ-
ent direction, for its hum eld is not uniform in all directions around it.
The power supply itself must convert the 60-Hz line power into a smooth
direct current for use by the active components in the device, and all
traces of ripple must be removed from it. This is done by power supply
filtering. The ripple caused by faulty power supplies is generally at 120
Hz and higher harmonics of 60 Hz rather than at 60 Hz, meaning the
audible hum is one octave higher in pitch than hum directly induced by
the power lines. It is thus easy to distinguish between the two simply by
listening to the signal.
Another type of hum often heard in audio systems is caused by inter-
ference from television broadcast stations and is called TVI, or television
interference. It is characterized by a buzzing sound rather than a
smooth hum, and it changes when the audio cables are moved around.
In most countries outside North America, the power-line frequency is
50 Hz rather than 60, and the resulting hum is correspondingly lower in
pitch.
Removal of the various kinds of hum is a constant challenge and fre-
quently a frustration to the audio technician. It is said that a certain tech-
nician, when asked by a customer, Why does my tape recorder hum?
answered, Probably it doesnt know the words.
Hum Bucking The cancellation of induced hum in an audio system by
the introduction of a 60-hertz component equal in amplitude and of
opposite polarity to the offending hum.
Hum bucking used to be common practice in the days of the electro-
dynamic loudspeaker. The signal to the voice coil was routed
through a hum-bucking coil that was placed by the large coil, which gen-

189
Hum Switch

erated the magnetic eld in the magnet assembly. The current in the eld
coil was not perfectly smooth due to power supply ripple, hence the need
for hum bucking.
Hum bucking is often used today in magnetic pickups for electric gui-
tars, and is used in some telephone lines to reduce 60-Hz noise.
Hum Switch A switch found on some audio equipment, such as ampli-
fiers for musical instruments, which reverses the neutral and hot leads
of the power cord. The ground lead of the power cord remains connected
to the chassis.
There is no agreement among manufacturers on which wire is consid-
ered the neutral, and in general the circuits are not perfectly symmetri-
cal with respect to capacitive and magnetic coupling from the two sides
of the power line. Sometimes reversing the leads on one or more of a series
of interconnected pieces of equipment will reduce the noise level, especially
hum. See also isolation transformer.
Hum Tone The sound produced by large, traditional tower bells, such as
church bells, consists of many overtones. The lowest, caused by vibration
of the whole bell, is called the hum tone and is one octave below the fun-
damental or prime tone for which the bell is named. It is interesting that
the hum tone of a bell is generally not audible at allthe perceived pitch
of the bell (called the strike tone) is one octave higher than the hum tone,
and there is no component in the sound spectrum of the bell correspon-
ding to the strike tone.
HVAC Heating, Ventilation, and Air Conditioning. HVAC systems can be
major contributors to noise in auditoriums, churches, and other large pub-
lic spaces.
HX A circuit developed by Dolby Laboratories for use in cassette recorders
to reduce the effects of self-erasure of high-level high-frequency sounds.
It involved the reduction of the record bias and a change in the high-
frequency pre-emphasis dependent on the signal level. HX has been super-
seded by hx pro, which is more effective and simpler.
HX Pro A clever circuit, developed by Dolby Laboratories, found in some
cassette recorders. It varies the high-frequency bias signal in the record
mode to reduce the tendency toward self-erasure. In magnetic tape
recording, loud high frequencies in the signal look like bias to the tape
and will tend to erase the signal as it is being recorded. The effect is called
high-frequency compression and is a fault of magnetic recorders in gen-
eral and cassette recorders in particular. The HX Pro system senses the level
of the high frequencies and reduces the level of the bias accordingly. It oper-
ates only in the record mode, and cassettes recorded with HX Pro can be
played on a non-HX machine without any degradation.
Hybrid Amplier An amplier that uses a combination of transistors
and tubes, supposedly combining the best characteristics of each.
Hybrid Transformer A special transformer with the windings con-
nected in such a way that a receiver can accept signals from and a trans-

190
 Hysteresis

mitter can send different signals down a single transmission line without
the signals interfering with each other. If the impedances of the circuits
and the transmission line are properly balanced, there is a high degree of
attenuation between the two local circuits. Because the hybrid allows one
pair of wires to simultaneously carry independent signals in both directions
without interference, it is sometimes called a two- to four-wire converter.
The induction coil in a standard old-fashioned telephone is actu-
ally a hybrid transformer, but it is deliberately unbalanced a little to pro-
vide side tone. Side tone is the sound of ones own voice one hears in the
receiver when speaking on the telephone. The induction coil in modern
phones has been replaced by electronic circuitry.
Hyper-Cardioid The hyper-cardioid is a microphone with a pattern some-
what like a cardioid, but less sensitive at the sides. It is used when it is
desired to minimize the reverberant sound picked up when some distance
must be maintained between the microphone and the sound source.
The hyper-cardioid will pick up the same direct-to-reverberant sound
ratio as an omnidirectional microphone at twice the distance from the
source. Examples of its use are in television and motion pictures, where the
microphone must be kept out of sight.
Hysteresis When a device or a system is presented with a stimulus that is
increasing in value, the response typically also increases. Then, if the stim-
ulus is gradually decreased, the response also decreases. Ideally, the
response at a given stimulus level would always be the same, but in most
real systems the response to a rising stimulus is different from the response
to a falling stimulus for the same value of stimulus. This effect is called hys-
teresis and is very common in mechanical, magnetic, and electrical systems.
If a graph of response versus stimulus is plotted for a rising stimulus and
the plot is continued as the stimulus falls, the resulting curve will be a loop
called the hysteresis loop. The area of the loop is a measure of the amount
of hysteresis. Friction causes hysteresis in mechanical systems. In analog
audio, by far the most important hysteresis effect is that of magnetic mate-
rials. In a magnetic tape head, the magnetizing force is the current in
the coil, and the amount of magnetization accepted by the tape is always
a little lower when the current is increasing than when it is decreasing. The
audio signal should magnetize the tape in a pattern that matches that of
the coil current, but due to hysteresis, the match is not perfect and the result
is distortion of the recorded waveform. bias is used in magnetic record-
ing to reduce the effect of hysteresis. See also anhysteretic.
If the magnetization of the tape is plotted against the coil current, the ideal
curve would be a straight line. The actual curve is somewhat S-shaped, and,
moreover, it has a slightly different shape when magnetizing in the positive
direction than when magnetizing in the negative direction. This difference
causes the hysteresis curve to have an enclosed area, and hence it is some-
times called a hysteresis loop. The most desirable magnetic materials have
a very small area within the loop, and this area is a measure of the quality

191
Hz

Hysteresis

of a magnetic tape, especially for analog audio recording. digital and fm

audio recording generally magnetize the tape at saturation, and distortion
caused by hysteresis does not appear in the demodulated audio output.
Hz See hertz.

I
IBOC In-Band On-Channel, a system of digital am and fm radio broad-
casting developed by iBiquity Digital Corporation. On October 10, 2002,
iBiquity announced that the Federal Communications Commission (fcc)
approved the IBOC system for immediate implementation, and the Inter-
national Telecommunications Union (itu) has approved In-Band On-Chan-
nel (IBOC) as a worldwide digital radio system.
iBiquity calls their version of the system HD Radio, and on April 24,
2003, Harris Corporation, presumably a licensee of iBiquity, announced that
it sold transmitting equipment to ten radio stations in the U.S. under the
name of DEXSTAR AM/FM High Denition (HD) Radio exciters.
It is interesting to note that HD radio is broadcast by existing AM and
FM transmitters using existing rf spectrum allocations, and the conven-
tional analog AM/FM signals are also broadcast at the same time, mak-
ing the system compatible with existing AM/FM receivers.
IC See integrated circuit.
Idler Any small rotating guide in a tape transport.

192
 Imaging

Idler Wheel A small wheel made of rubber or similar material that is used
in a friction drive system for turntables and some tape recorders. Some-
times also called a puck, due to its slight resemblance to a hockey puck.
IEC The International Electro-technical Commission (IEC) is an organization
based in Europe that is involved in the setting of standards. The IEC has rec-
ommended pre-emphasis and de-emphasis curves for tape recorders, and
most European models follow the recommendations. See also ccir.
IF Intermediate Frequency. See heterodyne.
IFB Interrupt Foldback. An addressable, listen-only intercom, often seen
as an earpiece worn by television newscasters, that can play the program
audio and be interrupted with comments or cues from the director or other
off-camera personnel. Also known as a cue system.
IFPI International Federation of the Phonographic Industry is a nonprot
association based in Switzerland that internationally promotes the rights
and interests of producers of phonograms and videograms (music videos).
IGFET Insulated Gate Field Effect Transistor, preferred term for the MOS-
FET. See also field effect transistor.
IHF, The Institute of High Fidelity The IHF was an organization of U.S.
manufacturers engaged in the setting of standard measurement techniques
for audio equipment. The IHF merged with Electronic Industries Associ-
ation (eia, which is now called the Electronic Industries Alliance) in 1979.
The setting of standards in audio has largely been taken over by the Audio
Engineering Society (aes).
IIR Filter Short for Innite Impulse Response Filter. A type of digital lter
frequently used in audio applications. It requires fewer DSP calculations
than an equivalent Finite Impulse Response (fir) lter, but is more difcult
to design.
IM See intermodulation distortion.
Imaging The ability to localize the instruments when listening to a stereo
recording is called imaging, and a great deal of nonsense has been written
about it. Accurate imaging of musical instruments on a stage by use of two-
channel stereo has been shown to be difcult at best even under the most
ideal laboratory conditions. To achieve any kind of accuracy, the channels
must have precisely the same gain, the frequency response of each
loudspeaker must be identical within 1 decibel or less, and the phase
response of the two channels must be identical. The listener also must be
precisely between the loudspeakers. These conditions are impossible to meet
in practice.
Three independent channels have been shown to be the minimum
number required for any kind of consistently accurate imaging. The fact is
that for most stereo systems, imaging consists of localizing the instruments
at one loudspeaker position or the other.
One must take care, however, to discriminate stereo spread from accu-
rate imaging. Most stereo systems provide an impression of diffusion and
spread of the sound, which can be pleasing to the ear and can lead the lis-
tener to think he is correctly imaging the sound sources. The reverberant

193
Impedance

sound in stereo recordings can sound quite diffused and spacious, but
again this is a far cry from imaging in the true sense.
Impedance In an electric circuit containing direct current, the mag-
nitude of the current is determined by the voltage across the circuit
divided by the resistance of the circuit. This is known as ohms law.
In a circuit containing alternating current, the situation is more
complex; the resistance presented to the current is a function of fre-
quency. This AC resistance is called impedance and is also measured
in ohms. Impedance is the sum of resistance, capacitive reactance,
and inductive reactance. (See Appendix 11 for more information on
the denition of impedance.) Alternating currents are affected by resistance
the same way as direct currents, and Ohms law can be used for AC if the
reactances are zero, that is, if there are no capacitors or inductors in
the circuit.
In audio circuits and components, many different impedances are
encountered. A loudspeaker, for instance, is a low-impedance device, usu-
ally about 8 ohms. This means that a given voltage across it will result in
relatively large amounts of current in it. The power accepted by the
speaker is equal to the voltage multiplied by the current. A condenser
microphone, on the other hand, is a very high-impedance device, gen-
erally several billions of ohms. The voltage generated by a condenser micro-
phone results in vanishingly small amounts of current because of the high
impedance. In general, impedances are relatively low where large amounts
of power are being transferred, and are relatively high when power levels
are low. An exception to this is found in low-impedance microphones, such
as dynamic microphones, where power levels are also very low.
Low-impedance circuits are less susceptible to electrical interferences
such as 60-hertz hum than are high-impedance circuits, and they are used
to transmit audio signals over cables. Most audio transmission lines used
in the broadcast industry are of 600 ohms impedance, except for speaker
lines, which are much lower in impedance. It is interesting that 600 ohms
would be chosen as a working impedance for commercial sound and broad-
cast work.
Four reasons can be stated:
1. Shunt capacitance found in long cables has negligible effect on
high-frequency response.
2. Line resistance in long lines is not an appreciable fraction of 600
ohms, so losses are low.
3. Higher impedances are much more sensitive to electrostatic
interference, and lower impedances are more susceptible to mag-
netically induced interference.
4. The capacitors and inductors needed for building equalizers at
600 ohms are of modest size and low cost.
There is a common misconception that impedances of interconnected
audio equipment must be matched, and one is constantly hearing about

194
 Impulse Response

problems caused by an impedance mismatch. The fact is that impedances

are almost never matched in audio systems. For instance, an amplifier
designed to operate with an 8-ohm loudspeaker will have an output
impedance of a small fraction of an ohm, in fact as low an impedance as
the designer can manage. There are two reasons for this, one having to do
with efciency and one with damping factor. If the amplier driving
an 8-ohm load had 8 ohms of output impedance, one half of all the power
generated by the amplier would be dissipated in the amplier itself and
would never reach the load. This is because the output current supplied
by the amplier passes through the amplier output stage as well as the
load, and the same power would be dissipated as in the load. For the same
reason, the wires connecting the speaker to the amplier should be of very
low impedance to reduce power loss in them as well. When the output ter-
minals of an amplier say 8 ohms, this simply means that its maximum
current and voltage capabilities are such that it will deliver maximum power
to an 8-ohm load. See damping factor for a discussion of the other effect.
In radio frequency and video transmission, impedances are matched to
avoid signal reections from the ends of the lines. Such reections cause
double images, or ghosts. This effect is due to the relatively short wave-
lengths of the signals involved. Such reections could be a problem at
audio frequencies if the transmission lines were hundreds of miles long.
They were sometimes heard in analog long-distance telephone circuits as
an echo, but the effect is of no consequence to normal audio circuits. Today
long distance telephone lines are almost exclusively ber optics, and they
are not susceptible to this type of echoing. See also Appendix 11.
Impulse Response An impulse is a signal or sound that has a very short,
in fact, vanishingly small, duration. A true mathematical impulse has zero
duration and innite amplitude, but still a nite amount of energy. Of neces-
sity, the energy in an impulse is spread evenly over a wide frequency
band, and this means that it can be used as a test signal to measure the
characteristics of an audio device. Simply input the impulse to the device
and record its output. The resulting impulse response is in the time
domain, and it contains all the information about the behavior of the device
if the device is linear. Any linear system can be characterized by its
impulse response; for instance, the reverberation time of a room is actu-
ally its impulse response. The frequency response of the device can be
obtained by inputting its impulse response into a spectrum analyzer such
as an fft analyzer.
Of course it is impossible to generate a true impulse, and even if it could
be done, it would not be a good test signal because its innite height would
overload any device. However, we can produce an approximation of an
impulse by generating a very short pulse with amplitude below the over-
load level of the device being tested. An example of this is a test signal on
some special compact discs that consists of a single sample at maximum
amplitude and all other adjacent samples set to zero. This signal has a per-
fectly at frequency and phase response from DC to 20 kilohertz.

195
In-Band Gain

Impulse Response

In general, such impulses are not very good for measuring audio devices
because they are so short they contain very little energy, and the signal-
to-noise ratio is quite poor when making measurements.
The dual-channel FFT analyzer is able to calculate the impulse response
of a system from the spectrum of the system input and the spectrum of the
system output regardless of what the input signal consists of, and this tech-
nique is generally better than true impulse testing because of the improved
signal-to-noise ratio. Sometimes music is used as a test signal for this type
of measurement. The impulse response can also be calculated by a process
called deconvolution using a small computer program, if the input sig-
nal is a perfectly known signal, such as pseudorandom noise. See also
maximum-length sequence.
In-Band Gain The gain, or amplication factor, in the passband of a
device, such as a subwoofer. For instance, in multichannel movie theater
sound systems, the standard for adjusting subwoofer response is such that
the low-frequency output of the speaker within its operating range is
higher in level than a full-range screen speaker in the same frequency range.
The reason for this is to increase the subjective loudness of the subwoofer,

196
 Inductive Reactance

whose low frequency output is in the range where human hearing is not
very sensitive. All modern digital lm sound formats use 10 dB of in-band
gain for the subwoofer.
lnductance Inductance is the quantitative measure of the effect of an induc-
tor. The unit of inductance is the henry, after Joseph Henry, a nineteenth-
century American physicist. An inductor allows constant electric current
to exist through it, but it resists any change in current in proportion to the
value of its inductance. The voltage across an inductor is equal to its induc-
tance in henrys multiplied by the rate of change in current through it. The
current is measured in amperes. Since high-frequency signals have cur-
rents that are changing faster than low-frequency ones, an inductor has an
increasing impedance as frequency rises.
Many types of inductors are used in audio circuits, including equal-
izers, crossover networks, and filters. Sometimes inductors are
called chokes, especially when they are used to lter out unwanted high-
frequency interference. Many interconnecting cables used in computer and
some audio equipment have small ferrite beads resembling little dough-
nuts around them near the connectors. They increase the inductance of the
cable, reducing the transmission of radio-frequency interference into the
equipment and the transmission of rf energy produced inside the equipment.
Induction The electromagnetic process by which a varying magnetic eld
causes an electric current to exist in a conductor. The current is called
the induced current, and its strength is proportional to the rate of change
of the magnetic eld. Induction is the basic principle by which magnetic
transducers, such as dynamic microphones and magnetic phono car-
tridges, operate. It is also the process by which electricity is generated
and by which transformers operate.
Magnetic induction was extensively studied by James Clerk Maxwell,
an English scientist of the nineteenth century. He was the rst to formulate
the precise relationship between electricity and magnetism in his famous
Maxwells equations, which stand as a major scientic discovery.
Induction Coil See hybrid transformer.
Inductive Reactance Inductive reactance is that portion of impedance
which is due to inductance. Examples of inductors are coils of wire and
transformer windings. The inductor behaves as if it had inertia with
respect to electric current. A current in an inductor resists changes in its
magnitude, and to make rapid changes in the current requires rather large
voltages across the inductor.
The dening equation of an inductor says the voltage across the induc-
tor is equal to the inductance times the rate of change of current. An induc-
tor will have resistance, which behaves in accordance with ohms law,
but with alternating currents, the inductance must be taken into
account. Inductive reactance is measured in ohms and is numerically equal
to 2 times frequency times the inductance in henrys.
The impedance of an inductor rises with increasing frequency, so with

197
Inductor

a constant AC voltage across it, the current will fall as frequency rises. There-
fore, it can be used (and is used) as a frequency-dependent element in
equalizers. An example of this use is in crossover networks.
An inductor also establishes the condition whereby the current and volt-
age are out of phase by 90 degrees. The current lags behind the voltage by
90 degrees; therefore, circuits containing inductance exhibit phase shift,
which can be an unfortunate consequence. In a way this is the opposite of
a capacitor, in which the current leads the voltage by 90 degrees.
Inductor See inductance; inductive reactance.
In-Ear Monitors Musicians performing on stage may use earplug-style ear-
phones instead of stage monitor speakers. They have the advantage of
excellent sound isolation, no possibility of acoustic feedback, and lower
sound pressure levels for the performer than monitor speakers. The dis-
advantage is that the performer is completely dependent upon the mon-
itor mix and must adapt to performing this way. The system can be
damaging to the performers ears unless careful limiting is used.
Innite Bafe Theoretically, the innite bafe is a loudspeaker mounted
in a wall of innite extent. In practice, it is an enclosure that is totally sealed,
so that it completely separates the front sound produced by the speaker
from the rear sound. In theory, such an enclosure would have to be innitely
large if it were not to inuence the operation of the loudspeaker itself. For-
tunately, this is not a requirement, for the inuence of the enclosure on the
speaker can be benecial to its operation.
When a loudspeaker is placed in a closed box, its resonant frequency
is raised due to the stiffness of the air in the box. The speaker cone must
alternately compress and rarify this air as it moves in and out. Therefore,
a speaker designed for use in an innite bafe must have a very low reso-
nant frequency in free air.
The design of innite bafes was revolutionized in 1954 with the advent
of the AR-1 made by Acoustic Research, Inc. This unit was based on an early
idea by Harry Olson of RCA Research Laboratories. See also woofer; baf-
fle; bass reflex.
Infrasonic Refers to sounds or signals whose frequencies are below the
normal human hearing range, generally considered to be 20 hertz and
lower.
The lowest audible frequency is not easy to dene, for it depends
strongly on level. Some experiments have found that hearing can extend
to 10 Hz and below at very high sound pressure levels. It is a common
misconception that infrasonic signals can be ignored because they are
inaudible. See bass intermodulation.
Sometimes the term subsonic is wrongly used to mean infrasonic.
Initialization The process of reading the table of contents from a compact
disc after it is inserted into a CD player. The player then displays the track
numbers recorded on the disc. Also, most digital recording processes
require an initialization step before actual recording can commence. This
is somewhat like formatting a disk for storing les in a computer.

198
 Intelligence

Input Impedance The input impedance of a circuit or device is the imped-

ance actually experienced by a signal that is connected to it. An ampli-
fier with a 100,000-ohm input impedance looks like a 100,000-ohm resistor
to the signal coming into it.
In general in audio work, input impedances of components are several
times higher than the output impedances of the components connected
to them. If a device with a relatively high output impedance is connected
to a low-impedance input, the signal level will be reduced because the low
impedance will effectively short out the signal.
In video systems, impedances of inputs and outputs are the same, or
matched, to avoid signal reections because of the high frequencies and
short wavelengths inherent in the video signal. See impedance and Appen-
dix 11.
Insert On many mixing consoles, each input channel may incorporate a 14
TRS jack for the purpose of accessing the signal ow in that channel only,
after the mic preamp and before the fader. This allows putting a signal
processor, such as a reverberator, in that channel, or accessing the channel
output directly. A typical conguration is to use unbalanced wiring with
the jack Tip the channel output and the jack Ring the signal return back in.
This can create some interesting possibilities when using various Tip/Sleeve
and Tip/Ring/Sleeve plugs, depending upon how far they are inserted. A
Tip/Sleeve plug inserted to the rst click will access the signal at the jack
Ring contact without activating the switching that would interrupt the sig-
nal continuing back through the Return. A Tip/Sleeve plug inserted fully
will access the signal normally at the Tip contact and activate the signal
interrupt switch, while grounding the Return contact through the plug
sleeve. A TRS plug inserted fully accesses the Send and Return normally
to allow for an effects loop.
Insertion Loss When a passive audio device is placed in a sound system,
it will, depending on the nature of the device, sometimes reduce the sig-
nal level. This is called insertion loss, and it is measured in decibels.
Insertion loss can be considered the inverse of gain. Passive devices that
typically have some insertion loss are crossover networks and equal-
izers. Insertion loss is compensated, or made up, by additional gain in
the amplifiers.
Integrated Amplier An audio component consisting of a preamplifier
and power amplifier all on one chassis.
Integrated Circuit An integrated circuit, or IC, is a miniature one-piece
device containing many transistors and resistors. Some ICs may con-
tain hundreds of transistors in a space smaller than the head of a pin. Most
audio ICs are amplifiers of various types, especially opamps.
Integrated circuits are far less expensive than equivalent circuits
made from discrete components, and this is the primary reason for their
popularity. Early ICs were not optimized for audio use and were not well
thought of, but many newer designs are of excellent quality.
Intelligence In the earlier days of broadcasting, the information impressed

199
Intensity Stereo

on the carrier by the process of modulation was called the intelligence. Per-
haps the term was more appropriate then than now; in any case, today it
is usually called the information.
Intensity Stereo Intensity stereo is an unfortunate linguistic misnomer that
has come to mean the recording of stereophonic signals that are distin-
guished only by level differences. These level differences have been called
intensity differences, but sound intensity is a specically dened quan-
tity and cannot be sensed by a simple microphone, nor would it be valu-
able in music recording if it could.
The term coincident microphone stereo would be a better choice, the two
microphones being very close together. This is also called x-y stereo, with
X and Y representing the left and right channels. In X-Y stereo, the signals
in the two channels are in phase because the two microphones used are
coincident or very nearly so.
In localizing sound sources, the human hearing mechanism uses level
differences, spectral differences, and time (phase) differences between the
signals heard by the two ears. X-Y stereo deprives the listener of spectral
and time differences, and results in a sound eld that is less spacious and
more compact than what other stereophonic methods provide.
The rst investigator into X-Y stereo was Alan Blumlein, an Englishman
who conceived the idea in the early 1930s. He postulated that two figure-
8 microphones angled at 90 degrees to each other and 45 degrees to the
centerline of the stage would allow localization of the individual instru-
ments on playback. (See also blumlein, alan.) At present, there are at
least two microphone congurations that qualify as X-Y stereo. One is the
classic Blumlein method with two gure-8 microphones; the other is two
cardioids, each angled at 45 degrees from the centerline. The dual car-
dioids result in a less reverberant, or drier, sound because of reduced
pickup from the rear of the room. The gure-8 method results in more
ambience from the room. The gure-8 method is also sometimes called
the electrosonic method, and some people prefer to restrict the term X-
Y to crossed cardioids.
Incidentally, Blumlein recorded his two signals on discs in the 1930s,
using the same groove format as is used today in stereo records. His tech-
nique was all but forgotten when modern stereo records were introduced
in 1958. The Bell Telephone Labs also experimented with a similar type of
stereo record in the 1930s, but made no attempt to commercialize it.
It has been said that X-Y stereo is easier to cut into a record because there
is no out-of-phase information in the signal, resulting in less vertical
motion of the stylus. This is true, although modern mastering and play-
back equipment can easily handle signals that are out of phase, except per-
haps at the lowest frequencies.
It has also been said, mostly by broadcasters, that X-Y stereo is prefer-
able for fm stereo broadcasters because of mono compatibility. The
argument goes that since the two signals are in phase, they will combine
with each other without cancellation to form a monaural signal when

200
 Intermodulation Distortion

received by a monaural FM receiver. This is theoretically true, but in prac-

tice few listeners with monaural FM radios are concerned with the best qual-
ity reproduction. It has also been repeatedly shown that a-b stereo
recordings can be mixed to mono with very little effect on tonal balance.
See also ambisonics.
Interaural Cross-Correlation, IACC The IACC is a measure of the differ-
ence in signals received by the two ears. Its value varies from negative 1,
meaning the signals are equal and out of phase, through zero, meaning
the two signals have no similarity at all, to 1, meaning the signals are
identical.
The IACC will be 1 for sound sources directly in front of a listener or
located anywhere on the plane bisecting the head vertically. As a source
moves laterally from dead center, the IACC decreases from 1 to a minimum
value when the angle is about 60 degrees off axis. It then rises a little as
the angle is increased, becoming 1 again when the source is directly behind
the listener. The shape of the outer ears (pinnas) and the geometry of the
head largely determine the variation in IACC with angle.
The IACC is measured using a dummy head that has articial ears, ear
canals, and precision microphones located where the eardrums would nor-
mally be. Such a dummy head, also sometimes called hats for head and
torso simulator, is placed in the sound eld to be measured, and the
microphone outputs are fed into an instrument such as a dual-channel
fft analyzer, which measures the cross-correlation.
It has been found by several researchers, especially Ando and Schroeder,
that small values of IACC correspond to greater degrees of envelopment
in auditoriums, and that listener preferences are for smaller IACCs. Such
ndings have important applications in the art and science of room
acoustics.
Intercut An analog magnetic tape editing technique where a section
of the master tape is physically cut out and replaced with another take.
Adhesive-backed plastic splicing tape is used. Today, most audio tape edit-
ing is done digitally, and the original source is not physically cut. The process
is still often called intercutting.
Interlaced Scan Video In traditional analog television broadcasting,
each video frame is divided into two elds, each having one half of the hor-
izontal scanning lines. The odd-numbered lines are scanned rst, then the
even numbered lines are added. This is called interlace scanning. ntsc stan-
dard color television displays 59.94 interlaced elds per second and is done
to reduce visual icker in the picture. Each pair of elds constitutes one
video frame. This means the frame rate of video is 29.97 frames per second,
whereas the motion picture frame rate is 24 per second. This complicates
the transfer of movies to video (see 3:2 pulldown). See also progressive
scan video.
Inter-Layer Transfer See print through.
Intermodulation Distortion Intermodulation (or IM), as the name implies,
is the amplitude modulation of one signal by another. As an exam-

201
Intermodulation Distortion

ple of how this happens, consider a recording where a piccolo is sounding

a steady high-frequency note, and a bass drum is then struck. The loud low-
frequency drum tone is added to the piccolo tone, and the combined wave-
form has the small uctuations of the piccolo riding up and down on the
relatively slowly changing drum signal. The amplifier is called on to
amplify each tone equally, but if the gain of the amplier varies with sig-
nal level (i.e., if it is nonlinear), the piccolo will be amplied by differ-
ent amounts depending on whether the drum signal is near zero or near
its maximum excursion. Therefore, the piccolo signal will undergo changes
in amplitude at the rate of the drum signal. The drum amplitude modu-
lates the piccolo, and its sound will be heard to flutter in the presence
of the drum.
Strictly speaking, any two tones A and B when passed through a non-
linear device will give rise to intermodulation tones having frequencies of
A plus B and A minus B. These distortion components are sometimes called
upper and lower sidebands. Often the upper sidebands are above the audi-
ble range, but the lower sidebands are always in the audible range and are
the most objectionable. It is possible, in fact common, to have tones that
are ultrasonic modulating each other to produce audible sidebands. This
is possible with phono cartridges that have extended high-frequency
response. The ultrasonic noise frequencies intermodulate to produce audi-
ble distortion if the preamplifier is nonlinear at ultrasonic frequen-
cies. This is an example of the fact that audio equipment must be linear
over a much wider frequency range than the audible range if it is to sound
as clean as it measures.
There are several methods of measuring intermodulation. The oldest
standardized method was devised by the smpte, and it consists of pass-
ing a mixture of a high and a low frequency tone through the device. The
output signal is passed through a filter that eliminates the lower tone.
The high tone is then amplitude demodulated to recover the sidebands that
contain the energy in the modulation, which is then expressed as a per-
centage of the amplitude of the high tone. Alternatively, the output signal
may be examined on a high-resolution spectrum analyzer to measure
the level of the sidebands independently of the tones.
The smpte method was devised many years ago, and it suffers from the
disadvantage that only one combination of tones is used as a test signal. A
newer and better method uses two tones that are swept through the audi-
ble and into the ultrasonic range while their difference frequency is held
constant at some low value, such as 50 hertz. Intermodulation will cause
a tone at the difference frequency to appear in the output, and its level can
be plotted as a function of the frequency of the two sweeping tones. The
subjective quality of sound equipment is rather well correlated with meas-
urements of this type.
A third way to measure intermodulation is to pass an octave band of
pink noise through the device. The octave band from 10 kHz to 20 kHz
may be used. IM components existing between all the frequencies of the

202
 Interval

noise will appear in the 0 to 10 kHz band, and can be measured with a high-
resolution spectrum analyzer such as an FFT analyzer.
IM specications of audio equipment are meaningless unless the method
used is stated. There is no universally agreed-on method, but the SMPTE
method is generally regarded as obsolete.
International System of Units, SI In 1960, an international meeting called
the Eleventh General Conference on Weights and Measures was held; some
small adjustments to the metric system were agreed upon and it was given
the new name Systme Internationale, abbreviated SI (the abbreviation
was to be used in all languages). In the SI, all physical units are derived
from seven primary standards. All electrical units are derived from the
meter, the kilogram, the second, and the ampere.
In addition to standard measurement units, the SI also recommended a
great many other standards, such as nut and bolt diameters, numbers of
threads per centimeter, wire sizes, etc.
Perhaps someday the United States will cease to be the last holdout for
the picturesque English system and will seriously adopt the SI.
Interpolating Response A term adopted by Rane Corporation to describe
the summing response of adjacent bands of variable equalizers using
buffered summing stages. If two adjacent bands, when summed together,
produce a smooth response without a dip in the center, they are said to inter-
polate between the xed center frequencies, or combine well. [Historical
note: Altec-Lansing rst described their buffered equalizer designs as
combining and the terminology became commonplace. Describing how well
adjacent bands combine is good terminology. However, some variations of
this term confuse people. The phrase combining lter is a misnomer, since
what is meant is not a lter at all, but rather whether adjacent bands are
buffered before summing. The other side of this misnomer coin nds the
phrase noncombining lter. Again, no lter is involved in what is meant.
Dropping the word lter helps, but not enough. Referring to an equal-
izer as noncombining is imprecise. All equalizers combine their lter out-
puts. The issue is how much ripple results. For these reasons, Rane adopted
the term interpolating as an alternative. Interpolating means to insert
between two points, which is what buffering adjacent bands accomplishes.
By separating adjacent bands when summing, the midpoints ll in smoothly
without ripple.]17
Interpolation Mathematically, interpolation is the estimation of the
unknown values of some variable quantity from the known adjacent val-
ues. This is done when plotting a smooth curve from a set of discrete points,
for example. In audio, it is the estimation of erroneously read digital sam-
ples in a digital audio system from the adjacent correct samples. See error
correction.
Interval The difference in pitch between any two musical notes. The inter-

17. Thanks to Dennis Bohn, Rane Corp.

203
Intonation

val corresponds to a frequency ratio, but intervals are referred to by name

and are based on the octave. For instance, the fth is the interval between
the tonic and the fth note of the scale, etc. Some audio frequency
response measurements are made in terms of intervals, such as octave bands,
one-third octave bands, and one-tenth octave bands. The one-third octave
band and one-tenth octave band do not correspond exactly to musical inter-
vals but are dened in terms of frequency ratios.
Intonation The tuning of a musical scale. A musician who plays out of
tune is said to have poor intonation.
Inverse Feedback See feedback.
Inverse Square Law A small sound or light source radiating energy into
three-dimensional space produces an intensity that falls off in inverse pro-
portion to the square of the distance from the observer to the source. This
means a reduction of sound pressure of 6 decibels for each doubling of
distance if the source is in a free field.
A completely free eld is never experienced in practice, however, and
sound levels in rooms fall off at less than 6 dB per doubling of distance,
although in most rooms there will be a range of distances that approximate
a free eld. See reverberation.
Ionophone A type of loudspeaker invented by the French inventor Sig-
mund Klein in 1946. A high-frequency electrical discharge ionizes the air in
a small quartz tube. The audio signal is applied to the ionized air through
a step-up transformer. The ionophone has no moving parts except the air
itself. The ionophone was produced commercially for a short time by Elec-
tro-Voice as the Ionovac tweeter. It was capable of excellent high-frequency
response, but it was efcient only at very high frequencies and suffered from
reliability problems, leading to its demise. Originally called the Ionophone,
it was sold under the name of Ionovac by the DuKane Corporation, St.
Charles, IL, and IonoFane by Fane Acoustics Ltd., Batley, Yorkshire, England.
IPS See tape speed.
Iron Pejorative term for outboard equipment in the context of its effect
on sound quality. He has so much iron in his chain its a wonder that we
can distinguish between men and women on his dialog premix.18
ISDN, Integrated Services Digital Network A worldwide digital telephone
network that uses standard subscriber lines to telephone exchange build-
ings and ber optic lines for long distance communications between ex-
changes. ISDN provides the subscriber 2 digital channels at 64 kilobits per
second bandwidth each. For transmitting stereo audio at high quality, a
special modem is used to encode the audio signal with a compression
scheme known as MPEG Layer 2. The recipient of the signal needs a sim-
ilar decoder to recover the stereo signal. The system is capable of near CD
quality sound.
Isochronous Literally, having a constant frequency, such as a clock pen-

18. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

204
 ITU

dulum. In the audio/video world isochronous usually refers to the trans-

mission of a DV (Digital Video) or digital audio stream over FireWire (IEEE
1394). The amount of data is high enough that IEEE 1394 allows a guaran-
teed bandwidth for the data, which is sent isochronously over the serial
bus. The data is thus sent at a continuous rate, regardless of any minor prob-
lems such as a data error. This is in contrast to other data communication
methods such as synchronous and asynchronous, which typically have
request/acknowledge protocols and may wait for enough bandwidth on
the bus. This was deemed unacceptable for digital video.
Isochronous may also refer to two or more signals that are transmitted
or recorded at the same time and will stay synchronized, such as the sound
and video in television broadcasting. If not truly isochronous, the two sig-
nals will eventually drift out of synchronization, even if started at precisely
the same time. Non-isochronous signals are a problem when attempting to
play back audio, video, and/or control signals that require time codeprecise
alignment for the very accurate long-term synchronization of sound and
vision.
Isolation Booth Sometimes shortened to iso booth. A small windowed room
or booth within a recording studio where a performer, such as an announcer,
soloist, or drummer can be situated with her microphone(s) while record-
ing and be isolated from the sound made by the other musicians. The micro-
phone(s) in the booth are also effectively isolated from the sound in the
studio proper.
Isolation Transformer A transformer that is used to isolate an audio
system or device from the power line. The third wire in the modern power
outlet is the so-called power line ground, and is connected to the earth at
some point in the power distribution system. This ground wire often car-
ries some current when many types of equipment are connected to the
same power line. Because the wires have resistance, the ground connec-
tions at various power outlets in a building will be at different potentials,
and ground loops will result if various parts of an audio system are
plugged in at different places. This can be a real problem, for instance in
recording studios, where the control room receives its power from one source
and the studio has power outlets from another source. If someone plugs in
an electric guitar in the studio and makes a connection between its chas-
sis and a microphone shield, hum can very easily be induced.
An isolation transformer breaks the connection from the signal ground
to the power line ground, eliminating ground loops.
Audio transformers are frequently used to avoid ground loops between
various devices, and they are also called isolation transformers, or repeat
coils. See also balanced power.
ISRC The International Standard Recording Code (ISRC) appears on com-
pact discs and contains information about the country of origin, the owner,
the serial number, and the date of recording. See also control and dis-
play symbols.
ITU International Telecommunications Union. The ITU is an international

205
Jack

organization located in Geneva Switzerland. The ITU is the leading pub-

lisher of telecommunication technology, regulatory, and standards infor-
mation. The ccir was subsumed by the ITU in 1992 and became ITU-R,
the division dealing with radio-communications.

J
Jack A female connector, frequently mounted on the chassis of an audio
device, that serves as a receptacle for the male connector on the end of an
audio cable, such as a microphone cable, etc. Most commonly used for
the one-quarter-inch classic telephone-type receptacles, which are called
phone jacks. One may wonder why it is not called a Jill.
Jackeld A patch bay in England. See also patch cord.
JAES Journal of the Audio Engineering Society. A technical journal published
by the Audio Engineering Society, which contains research papers on all
aspects of the science of audio.
JASA Journal of the Acoustical Society of America. A respected technical jour-
nal, which publishes the latest research papers on all types of acoustical
subjects.
Jewel Box The original hinged plastic container for cds.
Jitter In an analog-to-digital converter, jitter is uncertainty in the
exact timing of the sampling of the signal. It is also called sampling off-
set uncertainty. Jitter introduces some distortion to the sampled signal.
See also aperture time errors.
JND See just noticeable difference.
Johnson Noise Johnson noise, named after John Bernard Johnson of Bell
Laboratories, also called thermal noise, is random white noise produced
by thermal agitation of the charges in an electric conductor. The noise is
proportional to the absolute temperature of the conductor. Johnson noise
manifests itself in the input circuits of audio equipment such as micro-
phone preamplifiers, where the signal levels are low. The noise power
is independent of the resistance of a component, and this means that noise
voltage is proportional to its resistance, and low-impedance circuits are
thus quieter than high-impedance ones. The Johnson noise level is the lim-
iting minimum noise any circuit can attain.
Jolly Green Giant Effect See proximity effect; Appendix 3.
Joystick Sometimes a special type of panpot that divides one input sig-
nal among four output channels is called a joystick. In a quadraphonic
sound system, a joystick could move the apparent position of a sound from
front to back and side to side and in combinations of these moves, all with
a single control. The usage probably stems from the control stick used to
y an airplane.
Jug Slang for vacuum tube, mostly used by amateur radio operators in ref-
erence to power output tubes. See also bottle; beam bottle.

206
 Kilocycle

Just In reference to a musical scale, a just interval is a frequency ratio

determined by the frequencies of a harmonic series of the frequency of the
tonic. Purely tuned just intervals are almost never used in music because
of their incompatibility with the octave, which is the most fundamental
of musical intervals. See Appendix 8.
Just Noticeable Difference, JND The term is used mostly by researchers
into musical perception and psychoacoustics who like to measure JNDs
of frequency of tones, loudness of various sounds, and so on. The meth-
ods used for determining JNDs are sometimes quite complex, and there is
a certain amount of controversy about their interpretation. For instance, in
some cases musicians are able to detect pitch differences much smaller
than JND data predict.

K
KDKA Generally credited as being the rst commercial radio station in the
U.S., KDKA was built by Dr. Frank Conrad of Westinghouse in Pittsburgh
in 1920. Its studio and transmitter were in the upper oors of the Westing-
house building. KDKA was a very innovative and inuential station and
was the rst to broadcast a church service, a presidential inauguration, a
remote broadcast of any sort, baseball scores, stock market reports, and time
signals.
KEMAR Acronym for a dummy head and torso simulator made by
Knowles Electronicsthe Knowles Electronics Mannequin for Acoustics
Research.
Key In music, the key is the pitch of tonic of the musical scale used. In
equal temperament, which is almost universally used today, the same music
played in different keys has a different pitch level but all the same inter-
vals. Nevertheless, many musicians, especially those who possess absolute
pitch, claim that the various keys have quite different characters, some
sounding bright, some somber, etc.
Key West Audion An early vacuum diode detector, essentially the same
as the fleming valve, independently invented by Lee DeForest about
1906. The name audion was given to it by DeForests assistant C. D. Bab-
cock. The audion was rst used in the Navys wireless station at Key West,
Florida. It was in the same year that DeForest patented the rst triode
and called it the audion detector. DeForest did not realize at rst that the
triode could be used as an amplifier; when he showed it to the engineers
at the Bell Telephone Laboratories, they immediately recognized its poten-
tial as an amplier and licensed the rights to use it from DeForest. See also
triode.
Kilocycle, kc Before 1948, an accepted term for frequency, representing
1,000 cycles per second. Since 1948, the term kilohertz (kHz) has replaced
it. See also hertz.

207
Kilohertz

Kilohertz, kHz frequency is measured in cycles per second, and the units
are hertz, abbr. Hz. One thousand Hz is one kilohertz, abbr. kHz. Some-
times the k is capitalized, but lowercase is correct. Note that it is exactly
1000 Hz, whereas a kilobyte of data is 1024 bytes due to binary counting.
A recent accommodation for binary counting is reected in the term
Kibibyte. See Appendix 10.
Kintek A company afliated with the dbx company that marketed movie
theater sound equipment, and was involved in motion picture sound track
experiments including comtrak and colortek/quintophonic. Some-
times seen as Kintek/Cinesonics.
Klangfarbe Literally, tone color in German. This can be considered a syn-
onym of timbre.
Klipschorn A famous design for a full-range loudspeaker system
invented by Paul W. Klipsch in 1941. The low-frequency section of the Klip-
schorn was an ingenious design for a relatively compact, folded horn that
used the corner of the listening room as an extension of the horn itself. It
was capable of low-distortion, low-frequency output down to about 30 Hz,
which was nearly unprecedented in those days
The Klipsch company was probably the oldest continuously operating
audio manufacturer in the country when Paul Klipsch retired in 1989. He
died in 2002, but the company lives on.
Kludge Slang for a poorly executed solution of a problem. It typically refers
to a hardware design of any sort that is not very elegant or well constructed
and may be composed of oddly designed or thrown together subsystems.
Sometimes spelled kluge, and usually pronounced klooge.

L
L The symbol for inductance or inductor, usually used in combination
with R and C (for resistance and capacitance), as in RLC lter. The
unit of inductance is the henry, abbr. H.
Labels Special non-audio information encoded along with the audio in
digital recording systems. They may be used to encode information about
the recording session, number of microphones used, dates, etc. See also
control and display symbols.
Labyrinth See acoustic labyrinth.
Lacquer Master See acetate.
Land The area between the grooves on a phonograph record and between
the pits in a compact disc is called the land. (Actually, there is only one
groove on one side of a record.) The name probably comes from analogy
with the land between the plowed furrows in a eld.
Lapel Microphone See lavalier.
Laser The light source used in the compact disc system to read the pat-
tern of pits in the surface of the disc is an aluminum gallium arsenide semi-

208
 Lathe

conductor diode, which emits light at a wavelength of 0.78 microme-

ter. When the beam strikes a land on the disc, it is reected full strength
to the detector, but if it strikes a pit, destructive interference occurs
because of the half-wavelength path difference between this and the land
reection. The output of the detector then drops, and thus the detector out-
put is a replica of the land/pit pattern on the disc. Lasers that produce light
with shorter wavelengths are used in dvd players because the pits on the
DVD are much smaller and closer together than those on a CD. Players
designed to play both CDs and DVDs contain two lasers.
The term laser is an acronym for Light Amplication by Stimulated Emis-
sion of Radiation. The laser was invented at the Bell Telephone Laboratories.
LaserDisc, LD, Laser Video Disc, LVD, LaserVision Introduced in 1978
as MCA Disco-Vision, the LaserDisc was developed to bring home theater
to the masses and compete with the then-emerging videocassette movie
with much higher quality. Its technologies would lead to all other optical
storage discs such as the cd and dvd, and it is claimed to be the rst con-
sumer device to use a laser. The LaserDisc was all analog. A helium-neon
gas laser (in a glass tube) would read pits on the 12 inch disc, but these rep-
resented an analog fm TV signal (with analog stereo sound) rather than
digital data. Later models would use laser diodes. cav mode limited play-
ing time to 30 minutes until clv mode was perfected. Later improvements
would add 2 channels of PCM digital audio of CD quality, and the limited
introduction of Dolby and DTS digital 5.1 surround sound data streams.
Unfortunately, years of early technical problems, the inability to record lim-
ited acceptance to all but videophiles, and the DVD nally killed the
LaserDisc. It might be said that the LaserDisc gave the original impetus to
the concept of the high quality home theater. Pioneer Electronics had bought
all the LaserDisc rights in the early 1980s and still administers optical disc
patents through its Discovision Associates division.
Laser Rot All but the earliest laser video discs were manufactured of two-
playing sides glued together, and after time, many discs were found to
develop noise in the picture and/or sound. The blame was placed on
reective coating deterioration, glues, and manufacturing, and likely had
little to do with the laser playback system. However, the name was con-
venient, and it has come to describe deterioration problems with any opti-
cal storage disc.
Latency Generally, the time elapsed between a stimulus and a response.
Humans have latency between the time they hear something and the time
they respond. Computer hardware and software can also have a delay that
can affect audio phase and monitoring time.
Lathe, Recording Lathe The recording lathe consists of a powerful turntable,
a cutterhead, and a cutter advancing mechanism that is used to cut the
master acetate in the rst step of producing phonograph records. Record-
ing lathes are built to very high standards of accuracy, to operate with very
low flutter and wow and negligible rumble, all of which, if present,
would be recorded onto the record being cut.

209
Lavalier

Lavalier A small microphone designed to be worn on a strap around a

persons neck. The lavalier has a built-in low-frequency rolloff to improve
the intelligibility of speech. It is also sometimes called a lapel or tie-tack
microphone and may be clipped to a lapel or other piece of clothing. Pro-
nounced, lav-a-leer by most English speakers. Sometimes called simply a
lav for short.
Layback In video editing, layback is transfer of the nished audio mix
onto the video edit master. This allows for traditional multitrack audio
production techniques while providing a videotape that a TV network can
play. In analog days, a special audio recording head with tracks in the
videotapes audio track areas would be installed on a multitrack audio
recorder and the actual videotape would be loaded and recorded onto.
Unfortunately, videotape is somewhat different magnetically from audio-
tape and this was not optimal, but it was better than the video recorder
sound track.
LBR Laser Beam Recorder, the device for making the glass master used to
produce metal parts for CD stampers.
LC Circuit A circuit usually found in equalizers or filters that uses both
inductors (Ls) and capacitors (Cs).
LCD Liquid Crystal Display, a display technology used for computers,
watches, and other indicators. Liquid crystals in displays are materials
whose light transmission can be changed when an electric current is
applied. Filters can provide colors.
LDA, Light-Dependent Attenuator. See light-dependent resistor.
LDR See light-dependent resistor.
Lead-Acid Battery An electric cell containing a lead dioxide anode,
a lead cathode, and sulfuric acid electrolyte.
The common car battery is composed of lead-acid cells. Standard lead-
acid cells need to be vented to the atmosphere, for they produce hydrogen
when being charged, but some newer types are sealed. An example of the
latter is the gelled electrolyte type, whose trademark is GelCell. Such sealed
lead-acid batteries are found in emergency lights and uninterruptible
power supplies because of relatively high-energy storage capacity, favor-
able voltage-versus-discharge curves, and good recharge ability. They
were also commonly used in devices such as camcorders before the adop-
tion of lithium-ion and NiMH batteries and the greatly reduced power con-
sumption of modern electronics.
Leadering In audio editing, the process of removing the outtakes, count-
offs, and noises between takes in a magnetic tape (and by extension, digi-
tal) recording. In analog magnetic tape editing, this process also involves
inserting leader tape between songs.
Leader Tape Nonmagnetic plastic or special paper tape that is spliced onto
analog magnetic tape between musical selections and at the beginning and
end of the magnetic tape. Leader protects the magnetic tape and identies
the beginnings and ends of selections, allowing the recordist to nd the
appropriate point in the tape when editing or mastering a record. Some

210
 LEDE

leader is timed, that is, it has marks every 7.5 inches and 15 inches to allow
the tape editor to insert the desired time between selections.
Leaf Tweeter The leaf tweeter is a miniature version of a planar loud-
speaker. Its diaphragm is a small sheet of plastic lm (the leaf), and it
has a ne grid of wires embedded in it. These wires form the voice coil
of the tweeter. There is an array of small high-strength magnets in front
of and behind the lm to provide the magnetic eld for the coil to react
with.
The leaf tweeter has very extended high-frequency response, some-
times to 40 kilohertz with very little rolloff, and its response is uni-
form and smooth over its range of operation. It is not effective below 5 kHz
or so, and must be crossed over at or above this frequency. Today, it is one
of the highest-quality, high-frequency drivers in existence. See also
crossover frequency.
Leakage The pickup of unwanted off-axis sounds by a directional micro-
phone due to the fact that its directional pattern is not ideal, or the pickup
of unwanted sounds by microphones that are supposed to be isolated from
one another, as in a multitrack studio recording. See also room leakage.
Leclanche Cell The cheapest type of cell. It is used for powering ashlights,
radios, small cassette recorders, and so on. It uses a carbon anode and a
zinc cathode with an electrolyte of ammonium chloride and is com-
monly called a zinc-carbon cell. Its output voltage is 1.5 V. It does not have
a very long shelf life or capacity, especially with heavy loads. Its capacity
is greater when used with an intermittent duty cycle. It is capable of lim-
ited recharging, but care must be used not to overcharge it or it can explode.
The Leclanche cell is essentially obsolete, having been replaced by the alka-
line cell.
LED Short for Light-Emitting diode, which is a solid-state diode that
glows a certain color when a current is passed through it. The emission
may or may not be visible light; many LEDs operate in the infrared region
of the spectrum. So-called white LEDs actually activate a chemical uores-
cence that emits a broad spectrum of light. LEDs are fairly efcient, rugged,
and long-lived, and most are inexpensive. Much research has been per-
formed to achieve different color outputs and light levels.
LEDs are frequently used as indicator lights. They can, for instance, be
arranged in a series with varying sensitivity so that they will glow in
succession as the audio level varies. This assembly can then be used as a
substitute for a vu meter or modulometer.
Infrared LEDs are used in a novel system pioneered by Sennheiser to
transmit stereo signals from a transmitter panel to battery-powered
receiving headsets in a room such as an auditorium. This permits partially
deaf persons to hear a performance at an amplied level and still be free
to move about the room with no wires attached.
An example of an LED is the laser diode that provides the beam of light
that reads the pattern of the pits in the compact disc system. See also ded.
LEDE Short for Live End, Dead End, which is a commercial trademark

211
Leq

used to indicate a particular acoustical design of a recording studio con-

trol room. In this design, the area around the monitor loudspeakers is
made acoustically absorbent, or dead, while the area behind the listeners
position is made reective, or live. This conguration is said by the design-
ers to increase the accuracy of the reproduction.
Leq Equivalent sound level. The Leq of a sonic event is that constant sound
pressure level which has the same amount of energy as the actual event.
Thus, the Leq is a long-term average, or integration, of a sound pressure
level. Leq is a convenient way of accurately measuring the level of a uctu-
ating sound. The duration of an Leq measurement can be any amount of
time, but usually ranges from a few seconds to several hours.
Leslie Speaker A biamplied speaker-amplier system for use with an elec-
tronic organ where there is a rotating bafe in front of the woofer, and
where the tweeter horns also rotate. (See bi-amplification.)
The Leslie was introduced as an adjunct to the Hammond organ. The
rotating parts provide an amplitude and frequency modulation to
the sound in a manner which cannot be duplicated with stationary loud-
speakers. The effect is a strong vibrato. The FM is caused by the dopp-
ler effect.
Level The term level is loosely used when the magnitude of a signal is
meant, usually voltage. Strictly speaking, the term should be reserved
for the value of a power in decibels when a particular reference power is
specied or understood. For instance, two sound pressures can be compared
in decibels, but when the reference is 20 micropascals of sound pressure,
the number of decibels is called the sound pressure level and is a specic
quantity of sound pressure rather than a simple ratio.
Leveler An electronic device used in certain sound systems to maintain
the level of background music so it can be heard above the background
noise. In a quiet ambience, the music stays at a low volume level, but as
the background noise increases, the leveler effectively turns up the vol-
ume, increasing the overall sound level, and in many cases, the annoyance
level.
Leyden Jar The rst capacitor, invented by Pieter Van Musschenbroek,
of Leyden, Holland, in 1745. Van Musschenbroek thought that electricity
could be bottled and experimented with bottles of water, which he charged
to high potentials with friction-induced static electricity. Through a series
of experiments, one of which put his assistant in bed for two days, he nally
did away with the water, substituting a layer of tinfoil on the inside and
outside of the jar. The Leyden jar uses glass as a dielectric and is capa-
ble of storing charges at very high voltages; it has been used in electrostatic
experiments for many years.
LFE Low-Frequency Effects, one of the audio channels in motion picture
and home theater surround sound, designated by the .1 in 5.1 sound. LFE
sounds are routed to the subwoofer in the system, and can consist of sound
effects such as thunder and explosions, etc., and also the lowest frequency

212
 Light-Dependent Resistor

Leyden Jar

octave of music. The channel is not capable of full range sound, having only
about one tenth of full-range bandwidth.
LFO Low-Frequency oscillator.
Life A spiritual pickle preserving the body from decay. We live in daily
apprehension of its loss, yet when lost, it is not missed.Ambrose Bierce
Lifter See tape lifter.
Light-Dependent Resistor, or LDR A type of photocell that has a resist-
ance that varies with the amount of light striking it. The active element in
such a cell is usually cadmium sulde. LDRs have been used in some audio
control consoles as components of attenuators, which are then called
light-dependent attenuators (LDAs). The LDA is operated by a standard
potentiometer, which controls the current through a light bulb that illu-
minates the LDR. This type of circuit is suited to remote control, where the

213
Lightpipe

audio signal does not have to be routed over long cables to the attenuator,
and contact noise in the control pot is seldom a problem.
LDAs have some disadvantages, including fairly slow response time. It
is also difcult to make them track so that several of them controlling var-
ious signals provide the same attenuation at all positions of the control
potentiometers. To control a stereo signal the channels must track in gain
very accurately to preserve the stereo image as a function of volume. The
LDR is essentially obsolete. See also opto-isolator.
Lightpipe A serial eight-channel multiplexed interface for digital audio
transmitted by a single ber-optic cable, terminating in a proprietary con-
nector. The Lightpipe was invented by Alesis to connect its ADAT modu-
lar digital multitracks.
Light Valve One type of device used for printing optical soundtracks on
movie lm. Pairs of metal ribbons in a magnetic eld are modulated by
the audio signal, which open and close to vary a light beam which exposes
the optical soundtrack on the lm. Care must be taken to avoid valve
clash, when overmodulation causes distortion in the valves as the rib-
bons hit. Another kind of light valve may also be used in additive color
lm printers. Several types of video projectors often describe their tech-
nology as light valve type if it involves some sort of transmissive con-
trol of a light source.
Limiter A special type of compressor that prevents the signal from
exceeding a certain preset level, no matter what the input signal level may
be. Limiters are sometimes used for special effects in popular recordings,
especially vocals. A vocal with limiting will be essentially at the same level
regardless of the effort put out by the singer, from a soft voice to a shout.
The shouting will sound subjectively louder, however, because of the
increased harmonic content of the sound. The dynamic range of a singer
at close range to a microphone is far greater than that of any instrument
or musical ensemble, and when recording a vocal with an ensemble with-
out limiting, a great deal of gain riding must be done to maintain musi-
cal balance.
Limiters are sometimes used in front of power amplifiers in sound
reinforcement systems or radio transmitters to prevent unexpected high-
level signals from causing overloading and large amounts of distortion.
Line Amplier Originally, a line amplifier was a special amplier
designed to amplify telephone signals for transmission over telephone lines.
The term is now used to indicate any amplier with a line level output
and an output impedance of 600 ohms or so.
Line ampliers are used in the broadcast industry for sending signals
from place to place and in recording studios to send signals between audio
devices such as reverberators.
Linear A system, circuit, or component is said to be linear if it meets the
conditions of proportionality and additivity, that is, if its output level
changes smoothly in proportion to input level changes, and if input x causes
output X and input y causes output Y, then x + y at the input must cause

214
 Linearity

X + Y at the output. This is a very important concept, adding predictabil-

ity to any system. Most tests we employ in audio (frequency response,
gain, phase, impulse response, etc.) are based on the assumption of
linearity.
Note that this does not imply that the output is identical to the input,
or that it even resembles the input. It is a common misconception that lin-
earity implies absolutely faithful music reproduction. Probably this arises
from the fact that a ruler-flat (linear) frequency response and a linearly
decaying phase of the frequency response mean that the output will be a
replica of the input. This is but one of a multitude of possible linear sys-
tems. filters, delay lines, reverberators, rooms, and loudspeak-
ers, etc., are all in principle linear, but do not have at frequency or phase
response, and their outputs do not necessarily resemble their inputs. In prac-
tice, all these elements possess some nonlinear properties that cause more
or less nonlinear distortion, as opposed to linear distortion, caused by
nonat frequency response.

Linear

Linear Distortion Any type of distortion that a linear system is capa-

ble of producing, as opposed to nonlinear distortion. Some types of lin-
ear distortion are frequency response errors and time delay errors such
as phase shift.
Linearity The degree to which a system or device is linear. There is no
simple way to express the degree of linearity; it cannot be measured in terms
of percentages or any objective quantity. Even though an ideal, perfectly
linear system might be said to be 100% linear, it is possible to characterize
less-than-perfect linearity only by the various distortions that it causes.

215
Line Array

Line Array An evolution of the column loudspeaker, the modern line

array speaker typically consists of modular full-range speaker cabinets that
can be attached in a hanging, curved column.
Line Input Any set of input terminals of an audio device designed to accept
line level signals, or signals above 25 mV rms. Most consumer devices
are usually intended to work with line levels on the order of 0.5 V rms, and
may be marked with terms such as AUX, CD, TAPE PLAY or such. How-
ever, they usually work identically, and have at response. A once com-
mon device that did not use line levels was the phonograph cartridge,
which output a few millivolts and required riaa equalization.
Line Level Line level refers to the average audio voltage level of a sig-
nal at a particular point in an audio system. According to the ihf (now
the eia), line level is any level above 25 mV rms, but most consumer line
level signals are in the neighborhood of 0.5 V rms. How the averaging is
done is important to the denition of the level. See also vu.
The output level of a preamplifier is typically line level, and the input
level of a power amplifier is line level. In commercial audio systems, line
level is metered with a VU meter, where 0 VU corresponds to .775 volts
rms of signal. The actual line level in these systems may be +4 dbm (1.23
volts rms) or +8 dBm (1.95 volts rms) even though the VU meter reads 0
VU. The 0 VU reference is simply a convenient meter reading to which to
adjust the signal when controlling the gain. It is especially important to
control levels accurately in tape recording and broadcasting to avoid over-
loads and consequent distortion. Note that nominal line level does not
imply headroom gures. The term dates from the early days of telephone
engineering usage.
Line Microphone A highly directional microphone that has an acoustical
transmission line in front of its active element, which may be of several types,
including condenser or dynamic.
Typically, a line microphone will have a fairly long tube (12 to 24 inches
long) protruding from its front. The tube will have a series of holes along
the side. Sound coming from directly in front of the unit will enter the holes
in succession, and each wave coming into a hole will add in phase to pro-
duce the composite sound reaching the transducer. Sounds coming from
the rear, however, reach the holes in reverse order, and the phasing is such
that they cancel each other out on the way to the transducer, resulting in
very little or no output. The longer the tube, the more directional the micro-
phone will be.
In the past, some quite bizarre line microphones have been built. Electro-
Voice once made such a microphone that was over 6 feet long. It had a sight,
similar to a gun sight, to aid in aiming it, and it was mounted on a tripod
with a swivel arrangement. Harry Olson, of RCA Research Laboratories,
built some line microphones with a cluster of tubes coming out the front.
They were of many lengths and were arranged in a spiral pattern, looking
something like a Gatling machine gun with different-length barrels.
Line microphones are sometimes called shotgun microphones.

216
 Lip Synch

Line Radiator Another name for column loudspeaker.

Line Source Loudspeaker See column loudspeaker.
Line Up The procedure carried out to ensure that recording, editing, play-
back, amplication, etc. equipment works to the highest possible standard.
It consists of systematic adjustment of the equipment according to a sched-
ule and may involve specialized calibration and test apparatus such as a
multimeter, tone generator, oscilloscope, etc. See also alignment.
Linked In a multichannel dynamic effects device such as a compressor/
limiter, channels may be set to all respond even if only one channel deter-
mines the activation of the effect. For example, a stereo compressor can com-
press both stereo channels the same even if only one channel determines
that the compressor should be activated. This maintains a balanced sound
overall rather than hearing an effect in one channel.
Linkwitz-Riley Crossover An active crossover network that has steep
slopes in the stopband, typically 24 dB per octave, and linear phase
response over the entire audible frequency range. Moreover, it can be
designed to introduce a constant time delay in one or more of the passbands
in order to achieve time alignment of the loudspeaker system. It was
developed in 1976 by Siegfried Linkwitz and Russ Riley.
The fourth-order 24 dB per octave Linkwitz-Riley is probably the most
common type of active crossover used in commercial sound systems. Its
very steep slopes prevent the high-frequency drivers from being damaged
by low-frequency energy, an important consideration with high-power
sound systems using high-frequency horn loudspeakers. The phase
response is such that there is no relative phase shift between the channels,
and this ensures well-controlled directional behavior of the loudspeaker
system at the crossover frequency.
Linkwitz-Riley crossovers are made by cascading butterworth fil-
ters, and always have an even number of orders, or sections. Although
most such units are fourth order, there has been some experimentation with
eighth-order Linkwitz-Riley crossovers, but they may have an audible
amount of group delay distortion and their transient response is not
perfect.
Linkwitz Transform A relatively simple equalization circuit designed
by Siegfried Linkwitz that alters the q and the low-frequency resonance
of a closed box loudspeaker. The circuit uses one opamp and some resis-
tors and capacitors, and Mr. Linkwitz programmed an Excel spread-
sheet that calculates the resistor and capacitor values based on the measured
Q and resonance frequency of the speaker in question.
Lip Synch When the dialog accompanying a motion picture or a television
program is synchronized with the movements of the actors lips, the pic-
ture is said to be in lip synch. The process of achieving lip synch in movies
can be somewhat complex since the sound is recorded independently of
the lming, albeit at the same time. See also smpte time code.
The situation can also be complex with television broadcasting, espe-
cially when the video and audio are sent across the country via satellite

217
Lissajous

repeaters. The audio and video signals are usually sent over separate chan-
nels, and because the distances between the satellites are quite large, the
video and audio sometimes arrive at the destination at different times, caus-
ing loss of lip synch. The serious viewer is often able to detect this timing
problem, much to his discomfort.
Lissajous Jules Antoine Lissajous was a French mathematician of the nine-
teenth century who described what happens when a graph is created with
sinusoidal waveforms acting at 90 degrees from each other. See also
sinusoid.
If two signals are sent to the horizontal and vertical inputs of an oscil-
loscope, the resulting pattern will be a Lissajous pattern. If the two sig-
nals are periodic and the periods are an integral multiple of one another,
the pattern will have certain symmetries that can be visually recognized.
One application of Lissajous patterns, and probably the most common,
is in determining relative phase between two signals. Aligning a multi-
track analog tape head is almost always done this way. Some home-type
amplifiers and fm tuners have small oscilloscopes built in for checking
phasing. In general, the Lissajous pattern caused by the two channels of
a stereo signal will look like a pulsating round ball of string, while a mon-
aural signal will look like a straight line at a 45-degree angle from ver-
tical. Stereo signals with crosstalk between the channels will be some-
where between these two extremes, looking like an ellipse tilted at about
45 degrees. The eccentricity of the ellipse determines the stereo separa-
tion. This separation is measured in the electrical sense, and may not pre-
dict the audible separation because high-frequency signals are most impor-
tant in subjective stereo separation, but the oscilloscope is more sensitive
to low frequencies. Most digital oscilloscopes are usually not very good
at displaying a simple Lissajous gure.
Litz Wire A special type of copper wire that consists of many small strands,
each insulated from the others. It has a large surface area and presents rel-
atively low resistance to very high frequency signals because of the
skin effect.
Live-to-Two-Track A method of recording in which the instruments and
vocals are mixed and recorded directly onto a two-channel stereo recorder.
No remixing is possible; however, the fidelity can be excellent since no
overdubbing or duplication of recorded tracks is needed. The relatively low
cost of direct-to-two-track digital recording has revived the popularity of
this medium for making master tapes, especially those intended for release
on CDs. Also called direct-to-two-track. See also direct-to-disc.
Load, Loading An impedance connected to the output of an audio device
is said to load the source. The load is that which accepts the power from
the source. Without a load, there can be no power transferred. Thus, it fol-
lows that the amount of power transferred depends on the impedance of
the load. The power is actually absorbed by the real, or resistive, part
of the impedance, and not by the reactive part.
This is true of mechanical and acoustical systems as well as electrical

218
 Logarithmic

systems. The efciency of a loudspeaker, for instance, depends on the

acoustic load placed on it. A nondirectional loudspeaker placed in a cor-
ner of a room has a higher acoustic impedance connected to it than if it were
out in the open, and because it is a velocity source, it will radiate more
sound power. Likewise, a loudspeaker placed in the throat of a horn will
see a higher acoustic impedance and will radiate more power than if it were
radiating directly into the air.
Load Impedance The load impedance is the impedance connected to the
output of a device. In the case of a power amplifier, the load impedance
is a loudspeaker, and is usually 4 or 8 ohms. In most audio devices, the
load impedance should be ve or more times the source impedance of the
device. In some rare cases, the device may be designed so the load imped-
ance should match the source impedance, and this is where the term imped-
ance matching comes from. An example of this is in broadcast equipment
such as ampliers for driving telephone lines. The impedances are matched
in these cases so signals are not reected from the amplier inputs or out-
puts, causing an echo. Sometimes such an echo could be heard on an ana-
log long-distance telephone connection. This audio echo is analogous to a
ghost on a television screen, which is caused by reections of the signal
reaching the antenna later than the direct signal.
In audio equipment that does not involve long transmission lines,
matching the impedances is not only unnecessary, it is not desirable
because signal levels are reduced 6 decibels by the loading effect of the
added impedance. This would be a disaster in power ampliers driving
loudspeakers, wasting half of the amplier power as heat in the output
stages of the amplier.
Local Feedback negative feedback around a single active circuit com-
ponent, as opposed to a feedback loop around several such components.
Localization See binaural.
Location Cart The sound recordist on a movie or television shoot uses
a unique mobile cart to hold his equipment. These can be elaborate rigs,
holding several tape recorders, mixers, radio receivers, and so on, all some-
times powered by a large battery system.
Location Recording Any recording not in a studio; for example, an orches-
tra recorded in their performance hall would be a location recording. In
lm/TV sound, the recording on the set, usually referring to the dialog.
Sometimes called location sound.
Logarithmic Having to do with the logarithms (abbr. log) of numbers rather
than the numbers themselves. In graphs of audio phenomena, frequently
the log of amplitude is plotted versus the log of frequency.
The common log of a number is the power to which the number 10 must
be raised to obtain the number. For instance, 10 to the second power (or 10
squared) equals 100, and this means the log of 100 is 2. Similarly, the log of
1,000 is 3, etc. A log scale is a scale where distances are proportional to the
logs of the represented numbers, while a linear scale (like an ordinary ruler)
has distances proportional to the numbers themselves.

219
Logarithmic

Linear and Logarithmic Scales

The reason we use log scales in audio is because the response of our hear-
ing mechanism is proportional to the logarithm of frequency and approx-
imately proportional to the logarithm of amplitude. As frequency is varied,
we perceive the pitch interval of a sound as being proportional to the fre-
quency ratio. For instance, if the frequency is doubled, the pitch rises one
octave. If the frequency is doubled again, the pitch rises another octave,
and so on. A frequency ratio of 2 is always a pitch interval of one octave.
In the case of amplitude versus loudness, we perceive a doubling of loud-
ness for each amplitude ratio of about 3. This is a ratio of 10 in power, or
10 decibels. The decibel scale is a logarithmic scale of power ratios. Thus
it takes a tenfold increase in power for us to perceive a doubling of loud-
ness, but another tenfold power increase doubles the perceived loudness
again.

220
 Looping

The primary reason we use decibel scales in audio is that a relatively

small range of numbers represents a very great range of amplitudes of
sounds. The dynamic range of the ear is greater than 10 to the twelfth
power, or a factor of 1,000,000,000,000. This is easily expressed as 120 dB.
An example of a log relationship outside audio is the behavior of pho-
tographic lm. To achieve linear increases in density of the negative image
requires successive doublings of exposure. This is the reason the shutter
speeds on your camera are adjusted so each one is double the preceding
one. Likewise, the lens openings are marked so that each f-stop admits twice
the light of the adjacent one.
Long Line An analog audio transmission line consisting of copper wire
whose electrical length is longer than the shortest wavelength to be trans-
mitted. The wavelength referred to has nothing to do with the wavelength
of the equivalent sound in air, for the speed of transmission in a line is very
much faster than the speed of sound. The propagation speed depends on
the electrical characteristics of the line and can vary from about 180,000 miles
per second down to 50,000 miles per second or so; thus the audio wave-
lengths in long lines are many miles long.
In long lines, care must be taken that the impedances are properly
matched to prevent the signal from being reected back and forth from the
ends of the line, causing audible echoes or distortions. These echoes are
analogous to ghost images in television reception.
A long line must also be designed so that the propagation speed of sig-
nals is the same for all audio frequencies, and this is done by applying spe-
cial loading coils, or series inductances, along the line. A line with variable
transmission speeds at different frequencies is said to be dispersive and
causes gross distortion of transmitted signals. Nowadays, long copper lines
are seldom used for audio transmissionthe task has been taken over by
microwave transmission and ber optics.
Long-Throw A type of dynamic loudspeaker, often a woofer, that is
designed to allow the cone to travel a long distance without encounter-
ing nonlinearities in its response.
Looping The process of post-production dialog replacement using identical-
length loops of picture, guide track, and record track. The line to be replaced
would thus repeat over and over, and the actor would go for a take when
they were ready. Also referred to as virgin looping, when recording onto
a blank piece of mag lm the same size as the picture and guide loops. When
optical sound was used, the recordings were made sequentially on a roll
and later manually synched to picture.
Although this process is not used these daysADR, or Automated Dia-
log Recording is the standard in the U.S.the act of replacing dialog is still
often referred to as looping. 19 See also adr.

19. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

221
Lo-Pass Filter

Lo-Pass Filter Short for low-pass filter.

Lo-Ro Shorthand for Left only-Right only. Among the many sound formats
lm sound people must deal with, Lo-Ro is an ordinary stereo left/right
audio pair, usually derived from mixing down a multichannel mix or even
originally mixed only in two channel stereo. There is no matrixed surround
information, thus no derived center or surround is possible, only left and
right. Television shows and industrial productions often utilize it. See also
lt-rt.
Loss The opposite of gain. When a signal passes through a circuit or
audio device, if the output power is less than the input power, the circuit
or device is said to have loss. It is usually expressed in decibels. See also
insertion loss.
Lossless See codec.
Loss Pad Another term for pad.
Lossy See codec.
Loudness Loudness is a subjective attribute of a sound and cannot be
quantied, except in a statistical sense. If a large sample of listeners is asked
to adjust the strength of two signals so that one is twice as loud as the
other one, the average power difference will be about 10 decibels, and
this will be almost independent of the absolute levels of the two sounds.
The unit of loudness is the sone, and a sound pressure level of 40 dB
is dened as having a loudness of 1 sone. 2 sones is twice as loud and cor-
responds to about 50 dB, etc. The use of sones to measure loudness is sel-
dom encountered except in some psychoacoustic research.
The loudness of a sound, especially a complex sound containing many
frequencies, has no simple relation to its sound pressure level, and it is hope-
less to try to measure relative loudnesses of different sounds by using a
sound level meter. Various attempts to build a meter that measures loud-
ness have been made over the years, but with very little success, so com-
plicated is our hearing mechanism. See also loudness level.
Loudness Control The loudness control is an addition to some amplifiers
or preamplifiers that attempts to correct for the reduced sensitivity of
our ears for low-frequency, low-level sounds. See fletcher-munson
effect.
The loudness control is simply a bass-boost circuit that has a relatively
greater effect as the volume is turned down. In some such circuits, the tre-
ble is also boosted a little. In other words, as the volume control is turned
down, the low-frequency content of the signal is reduced less than the high-
frequency content. Early loudness controls tended to overcompensate with
the result of too much bass at low levels, but more recent designs are more
convincing.
Loudness Level The sound pressure level (SPL) of a 1-kilohertz pure
tone that has the same loudness as the sound in question is the loud-
ness level of that sound. The loudness level scale is a decibel scale, and
the unit is the phon. Thus, a sound which is as loud as a 1-kHz pure tone
of 60-dB SPL has a loudness level of 60 phons.

222
 Loudspeaker

Phons are seldom encountered in audio work but are used by

researchers in psychoacoustics. See also fletcher-munson effect.
Loudspeaker The loudspeaker is the transducer that converts electri-
cal energy into acoustic energy. The rst such transducers were earphones,
and the rst such unit to make enough sound to be heard in a room was,
naturally enough, called a loudspeaker.
The most common type of loudspeaker in use today is called the
dynamic loudspeaker, which was invented by a German engineer named
Siemens. The dynamic loudspeaker is a moving-coil device. The motion
of the cone is caused by the force created by an audio signal current in
the coil, which reacts with a stationary magnetic eld provided by a per-
manent magnet. See also voice coil.
The most common type of dynamic loudspeaker in use is known as the
direct radiator, so called because the sound is radiated directly from
the loudspeaker cone, without having a horn coupled to it.
One of the most important characteristics of a dynamic loudspeaker is
its resonant frequency, which is the frequency at which it will vibrate nat-
urally if perturbed. The resonant frequency will be near the lowest frequency
that the speaker will produce well, and is that frequency at which it is eas-
iest to move the cone. The output from the speaker will be at a maximum
at this frequency. Therefore, damping must be added to a speaker system
in order to reduce this peak in response. The damping may be in the form
of acoustic absorption inside the cabinet.
Another important characteristic of a loudspeaker is its efficiency,
which is a measure of how much of the electrical energy it receives is radi-
ated as sound, related to how much is dissipated as heat.
In general, the response from a bafed direct radiator dynamic loud-
speaker will be uniform for several octaves above the resonant frequency
and will fall off at the rate of about 12 decibels per octave below the res-
onance. See infinite baffle and bass reflex.
There are several other types of loudspeaker mechanisms, that are in
relatively limited use but are still important. One of these is the electrostatic
loudspeaker, which is a very thin sheet of plastic lm suspended between
two wire grids or screens. The plastic lm is conductive and is charged with
a high direct voltage, usually about 600 volts or so. The grids are con-
nected to the amplifier output through a step-up transformer, which
applies up to 300 alternating volts of signal to them. The lm is alternately
attracted to one grid and then the other by electrostatic forces, and the result-
ing motion of the lm radiates the sound.
Electrostatic speakers generally are physically quite large, some stand-
ing 6 feet high by 2 or 3 feet wide. They are always direct radiators, and
they must be large to attain reasonable efciency at low frequencies. This
is partly because they radiate equal amounts of sound to the front and to
the rear, but the two sound waves are of opposite polarity with each
other and tend to cancel. The radiation pattern of an electrostatic speaker
in a free field is a cosine pattern, similar to the sensitivity pattern of a

223
Loudspeaker

Cross Section of a Dynamic Loudspeaker

figure-8 microphone. Because of their large size, electrostatics become

very directional in the high-frequency range, beaming the sound almost
like a ashlight. They also have very low electrical impedance, which is
primarily capacitive, and this presents a problem for some amplifiers.
Electrostatic loudspeakers, in spite of their disadvantages, can have very
low distortion, and careful design can partially overcome the directional
problems. They are not capable of extremely high sound levels because the
voltages required to permit this would be so large that an arc would occur
between the grids and the lm.
Another type of loudspeaker is the so-called planar type, which com-
bines some aspects of the dynamic and the electrostatic designs. The planar
speaker consists of a large plastic sheet with conducting wires imbedded in
it. The wires are the voice coil, although they are not coiled up but rather

224
 LSB

traverse the sheet in a grid-like pattern. Many small magnets in front and
behind set up a magnetic eld around the sheet, so current in the wires causes
a force on it and it moves as a unit, similar to the motion of the electrostatic.
Planar speakers suffer from the same directional problems as electro-
statics, but their impedance is more like that of dynamic speakers. See also
horn.
Low Bit Rate Coding Digital audio with any data reduction scheme such
as MP3. The denition of low is not given.
Low-Pass Filter A filter that uniformly passes frequencies below a cer-
tain frequency called the cutoff frequency. Usually this is dened as
the frequency where the amplitude response of the lter is 3 decibels
below its nominal value.
Many early tone controls were variable low-pass lters.

Low-Pass Filter

LP Originally a trademark of the Columbia Broadcasting System desig-

nating the then-new (1948) microgroove long-playing phonograph
record. The term was rapidly taken as a generic to distinguish the micro-
groove 3313-rpm record from the 78-rpm standard records.
L-Pad A type of attenuator typically consisting of two resistors sche-
matically arranged in the form of either an inverted, backward letter L or
a capital Greek gamma. The physical layout of the components in an L-pad
need not be in the form of a letter L! It can have a xed amount of attenu-
ation or can be variable, in which case, the two resistors are replaced with
potentiometers that are ganged together. The variable L-pad presents
either a constant input or a constant output impedance to its associated cir-
cuitry regardless of its attenuation setting and is commonly used as a gain
control in audio consoles. See gure at attenuator for schematic.
LSB Least Signicant Bit refers to the smallest signal voltage level which
an analog-to-digital converter can encode. The term was thought
to be nonrigorous mathematically, so it was replaced by the more accurate

225
L-Section

quantizing step. Anything below the quantizing step is not recognized

by the ADC. The value of the LSB is also equal to the amplitude resolu-
tion of a digital system. In other words, the minimum difference in level
between two successive samples is 1 quantizing step.
The value of the quantizing step is dependent on the number of bits
available for encoding the signal. For example, in a 16-bit encoding system
such as is used in the compact disc, the instantaneous signal amplitude
is encoded into one of 65,536 voltage values. The difference in level
between two adjacent steps is 1 LSB.
L-Section A filter or part of a lter that consists of two reactive elements,
one in series with the signal and one in shunt with the signal. The name
comes from the schematic diagrams physical resemblance to an inverted
letter L. An L-section lter has essentially at response in its passband
and 12 dB per octave slope in its stopband.
LSI Large Scale Integration refers to the technology that can produce thou-
sands of transistors on an integrated circuit. VLSI (Very Large Scale
Integration) is common now.
LT Acronym for Low Tension, which is British usage for low voltage.

L-Section

226
 Magnetic Distortion

LTC Longitudinal Time Code, as opposed to vitc. smpte time code that
is recorded linearly like an audio track, although the data is digital. Some-
times spoken as lit-see.
Lt-Rt Left totalRight total. A 2-track stereo signal that has been matrix
encoded with surround channels. See also Lo-Ro.

M
Machine Room The room in a motion picture sound mixing facility that
held many audio recorders and players using sprocketed magstock. Multi-
element sound tracks would be produced in a mixing or dubbing theater
with sounds delivered from the machine room.
Macintosh, McIntosh McIntosh is the name of a respected hi- manufac-
turer and the name of an apple variety. Apple Computer calls their com-
puter the Macintosh, a name that originally required permission to use from
McIntosh Laboratory, as noted on the Macintosh Plus computer product
label of 1986.
MADI Multichannel Audio Digital Interface. A digital audio transmission
standard proposed by the aes and standardized by the American National
Standards Institute (ANSI S4.43-1991) specifying and controlling the
requirements for digital interconnection between multitrack recorders and
mixing consoles. The standard provides for 56 simultaneous digital audio
channels that are conveyed point-to-point on a single coaxial cable tted
with BNC connectors along with a separate synchronization signal. Basi-
cally, the technique takes the standard aes3 (formerly called AES/EBU)
interface and multiplexes 56 of these into one sample period rather than
the original two. See also aes3.
MAF See minimum audible field .
Magic Eye A small cathode ray tube in which the illuminated area of
a uorescent screen varies with the level of an applied voltage. Magic eyes
were introduced in the 1930s by RCA for use as tuning indicators in some
of their radios. Some inexpensive tape recorders have also used them as
volume indicators.
Magnet Something acted upon by magnetism. Magnetism Something
acting upon a magnet. The two denitions immediately foregoing are con-
densed from the works of one thousand eminent scientists, who have illu-
minated the subject with a great white light, to the inexpressible advance-
ment of human knowledge.A. Bierce
Magnetic Distortion A type of distortion in dynamic loudspeakers
caused by nonlinearities in the interaction between the magnetic eld in
the gap and the voice coil. Magnetic distortion arises because the voice
coil moves in the gap and, in so doing, does not always encounter the same
magnetic eld strength, causing the force it experiences to vary with its
excursion. See also overhung coil and underhung coil.

227
Magnetic Tape

Magnetic Tape Magnetic tape generally consists of a plastic base of one

type or another that has been coated with a mixture consisting of a exi-
ble binder mixed with very ne particles of iron oxide.
Several oxides of iron are used, sometimes along with some small quan-
tities of other elements such as cobalt, and the density of the magnetic
particles in the binder has been greatly increased, thus increasing the sen-
sitivity of the recording process and reducing the noise.
The rst plastic magnetic tapes, made in Germany in the 1940s, were
primitive compared to the modern product. The original German record-
ing tapes were made of homogenous magnetic material mixed with the plas-
tic. It worked equally well on either side.
Magnetic tape has been the quality standard for analog audio record-
ing for many years, but now digital recording and the digital audio com-
pact disc have proven superior in some ways. Even though tape is still
often used as a storage medium in such systems as the dat, the recording
quality is determined more by other factors involving the digital encoding
and decoding of the signal than by the tape itself.
Magnetism The study of magnets and magnetic phenomena. A magnet is
a body that attracts iron. It has two ends called poles. The earth is a giant
magnet, with poles roughly corresponding to the north and south geo-
graphic poles. A bar magnet, if suspended, will orient itself in a north-south
direction. One pole is north-seeking, one is south-seeking, and they are com-
monly called north and south poles. (It follows that the north pole of the
earth is actually a south magnetic pole!)
The area of inuence around a magnet is called its magnetic eld, and
its strength is measured in teslas, after Nikola tesla.
Magnetism relates to audio via the well-known phenomenon of mag-
netic recording. Magnetic recording is usually thought of as analog
recording, but many digital recordings are also made as magnetic record-
ings, the bits of the digitally encoded signal being stored as tiny magnet-
ized areas on either magnetic tape or magnetic hard discs. However,
optical means of recording are increasing in popularity, such as recording
directly onto a blank cd recordable. The original CD-R cannot be edited
physically, so the audio signal must be transferred into a computer con-
taining software for digital editing.
Magnetomotive Force The force experienced by a wire in a magnetic eld
when the wire is conducting an electric current. The magnitude of the
force is proportional to the product of the current, the length of the wire,
and the magnetic eld strength.
The magnetomotive force is the underlying mechanism for the opera-
tion of all loudspeakers that use magnets in their construction. See also
bl product.
Magnitude The portion of the frequency response or impedance of a
device that represents the amplitude is called the magnitude, as distin-
guished from the phase, which is the other, equally important, part. In most
published response and impedance curves, only the magnitude is included,

228
 Masking

but the phase curve also contains much information about the performance
of the device.
Strictly speaking, the term only applies to complex quantities, i.e., quan-
tities characterized by a magnitude and a phase. For noncomplex quanti-
ties, the term amplitude is used. See also Appendix 11.
Mag-Optical Print A motion picture lm that has both optical and mag-
netic sound tracks so it can be reproduced in conventional theaters with
optical sound equipment and also in houses equipped with stereo mag-
netic sound. It is now obsolete, with the advent of digital motion picture
sound.
Magstock See fullcoat.
Mains In Britain, the power line has traditionally been called the mains,
and the usage is gradually migrating to the United States. In the U.S., mains
refers to the 60-hertz 115-volt power available at wall sockets.
Sometimes the primary output channels of a sound reinforcement sys-
tem control console are called the mains, as opposed to the monitor out-
put channels.
Make Before Break A type of selector switch that momentarily connects
the moving contact, or wiper, to two of the xed contacts as it is moved
from one position to the next. This is also called a shorting type of switch.
Most multiposition switches used in audio for step attenuators or for switch-
ing between various signal sources, etc., are shorting switches because the
output circuit is never left completely disconnected from an input, and no
clicks or other noise are induced in the output circuit as the switch is moved.
Marry Unlike videotaping, motion picture sound and picture editing are
handled separately. Synchronized viewing during editing is handled by
devices that synchronize picture and sound elements. When both elements
are done, they are both copied onto lm prints that combine the picture
and sound into the single lm format playable in theaters. It is said that
the elements have been married into the release print. See also double sys-
tem sound; moviola.
Masking Masking is a subjective phenomenon wherein the presence of one
sound will inhibit our ability to hear another sound. We all know that we
usually dont hear softer sounds in the presence of louder ones, but the pre-
cise description of how our ears perform masking is very complex.
At mid-frequencies, a loud pure tone of a particular pitch is much
more effective in masking softer high-pitched tones than it is in masking
lower-pitched tones. However, at very low frequencies (30 Hz to 60 Hz
approximately), such as those produced by certain pipe organ pedal stops,
very little masking occurs even though the level of the fundamental may
be very high. These stops must be very strong to be heard at all because
our hearing threshold is so high at these low frequencies, and this seems
to affect the masking phenomenon.
Almost all experimental studies of masking have used pure tones as test
signals, and the masking of complex tones by other complex tones is a
much more difcult and complicated study. It is not necessarily true that

229
Master

one can extrapolate, or sum up, the pure tone masking data and come up
with meaningful complex tone masking information. The study of the mask-
ing of musical instruments by one another is a wide open eld that should
be addressed more fully.
It is also possible for a sound to be masked by a sound that occurs slightly
earlier than the masked sound, and this is called forward masking, or for-
ward auditory inhibition. The existence of auditory backward inhibition,
where an earlier sound is masked by a later-arriving sound, has been
reported by Georg von Bksy, but others have had difculty in reproducing
his results.
Master A gain control on a sound reinforcement or recording console that
controls the level of a mixture of signals whose levels have been controlled
by the mixing pots (See potentiometer). A console will have a master
gain control for each output signal. The control itself is actually an atten-
uator, so it really can only decrease the level of a signal fed into it. The
gain, or amplication, occurs before it goes through the control.
Mastering During the time when the vinyl phonograph disc was king, mas-
tering was the common term used for the process of transferring the musi-
cal signal from a magnetic tape (usually called a master tape) to an
acetate master disc. It is the rst step in the manufacture of phonograph
records from tapes. Mastering is fraught with pitfalls and complications,
and is thought by many to be as much an art as it is a science. The edited
tape from which the acetate is cut is called the master tape. It could be an
original recording, but more often it is a copy of original tapes. The person
performing the task is called a Mastering Engineer.
The advent of the cd has changed the technology and techniques used
to record, edit, and make recordings. The work of the modern mastering
engineer consists of completing the nal changes to the elements in a record-
ing to make a nished product. This can include anything from taking raw
separate tracks and mixing, equalizing, limiting, adding reverberation, com-
pressing, etc., into the nal product, usually a CD. Many artists rely on
famous mastering engineers to create their nished recordings, consider-
ing them the nal golden ears. See also editing.
Mastering Lathe A precision turntable and cutting head assembly for the
cutting of audio signals onto an acetate disc, even though the disc is
not made of acetate. The movement of the cutting stylus over the acetate
determines the groove spacing or pitch of the disc. Most records are cut
with variable pitch in order to squeeze more time on the record. The pitch
is larger when loud bass-heavy music is being recorded, and is smaller in
softer passages. The depth of the groove on the disc is also adjustable. These
functions are performed by the pitch and depth control computer that con-
trols the depth, width, and spacing of grooves being cut. The nished acetate
is also sometimes called a lacquer master, even though it is not made of
lacquer either.
Master Tape In the audio business, the master tape is the original medium

230
 Matching

the signal is recorded on, while video editors call the master tape the com-
posite edited video, made by intercutting the video from multiple source
tapes. It has been said that this is not the only mysterious behavior prac-
ticed by video editors.
Matching When one audio device, such as a preamplifier or equal-
izer, is connected to another, they are said to be matched if the resistive
part of the output impedance of the rst is equal to that of the other. More
correctly, the impedances, not the devices, are matched. (Strictly speaking,
it is only the real, or resistive, parts of the impedances that are matched,
but in common parlance, the impedances are said to be matched.)
Much confusion exists about impedance matching. While it is true that
from a theoretical standpoint the power transferred in a circuit is maxi-
mized if the impedances are matched, there are many reasons why this is
not desired in practice. The main reason is that efciency is more impor-
tant than maximum power transfer.
Consider the ashlight battery, which is not a battery at all, but a cell.
It can be thought of as a voltage source in series with a resistor; in other
words, it has resistance, just as any electrical conductor has. If a resistor
equal in value to the batterys internal resistance (probably less than 1 ohm)
is connected across the battery, the power transfer will be maximum. This
means the energy per unit time will be maximized. But one-half of the power
will be wasted in heating up the battery because of its internal resistance.
This means that only one-half of the energy in the battery can be extracted.
If, on the other hand, a much larger resistor is placed across the battery, most
of the power will be dissipated in the resistor and only a little in the battery
because power equals the square of current times the resistance. Thus, if
the resistance ratio is 100, there will be 100 times more energy extracted from
the battery than will be wasted in heating the battery. This is the reason bat-
teries are made with the lowest possible internal resistance and they are not
designed to have matched loads connected to them. The same argument
applies to most audio equipment, especially power amplifiers.
There are instances where impedances in audio circuits should be
matched, but they are relatively uncommon. One example is the trans-
mission of audio over very long lines, such as long-distance telephone lines.
The reason is not to maximize the power, but rather to prevent reections
from the ends of the line. If a signal traveling down a line many hundreds
of miles long meets a sudden change in impedance, part of it will be
reected, just as light is partially reected when it encounters a pane of glass.
The reected signal is then heard at the other end of the line as an echo.
The signal travels down the line at a speed less than the speed of light, but
is still very fast. At audio frequencies, if the line is short, the reection occurs
so near to the time the signal left the source that it cannot be heard or mea-
sured at all.
The situation is different with very high frequencies such as television
signals. Here, the wavelength is short compared to the distance traveled,

231
Matrix

and reections cause the visual ghosts familiar to television viewers. The
wavelengths of audio signals on transmission lines are many miles long,
so no trouble exists until the lines get comparably long.
In almost all audio equipment, the output impedances are made much
lower than the corresponding input impedances, and efciencies are high.
There is no valid reason for matching impedances when the distances
between devices are short, but there exist engineers (usually from the broad-
cast or telephone industry) who insist otherwise.
The term matching is also used as a descriptor for the pairing of micro-
phones or loudspeakers with nearly identical characteristics. The pairs are
then called matched pairs, and in many instances are available from man-
ufacturers already matched. Monitoring loudspeakers used in mastering
studios should be matched and microphones used for stereo recording
should also be matched.
Matrix, Matrixing The linear mixing of two or more signal channels at
specic amplitudes and phases to form two or more new signals is called
matrixing. These new signals can be combined in similar ways to recover
the original signals. The circuit topology used for matrixing is called a
matrix.
It is important to note that matrixing is a linear addition, or mixing, of
signals, and is not the same as modulation. There are many ways in which
audio signals are matrixed. One of the most common is the way stereo sig-
nals are broadcast over fm radio: the two stereo channels, L and R (for left
and right), are added together to form a sum signal L + R. Then they are
subtracted to form a difference signal L R. The sum signal is the mon-
aural or in-phase part of the stereo signal, while the difference is the
stereo or out-of-phase portion of the signal. All the directional information
is in the difference signal.
In the stereo FM transmitter, the sum signal is used to modulate the
carrier directly. The difference signal is used to modulate a 38-kilohertz
subcarrier, which is then mixed with the sum signal before it modulates
the carrier. Thus the carrier is actually modulated with a combination of
the sum signal and the 38-kHz subcarrier. If this FM transmission is
received by a monaural FM set, the 38-kHz carrier is ignored, while the sum
signal (L + R) is recovered. The listener then hears the summation of the
two stereo channels. The FM stereo receiver, on the other hand, also recov-
ers the 38-kHz subcarrier and demodulates it to recover the difference sig-
nal (L R), which contains the stereo information. Then, to recover the
original stereo signals, the sum and difference signals are matrixed:

(L + R) + (L R) = 2L, and
(L + R) - (L R) = 2R

The purpose of this entire exercise is simply to enable monaural FM

receivers to pick up both channels of the stereo signal.
Another example of matrixing is the combination of two rear-channel

232
 Maximum Output Level

signals with two front channels for quadraphonic recording. The two rear
channels are phase shifted, respectively, plus 90 degrees and minus 90
degrees, and are then mixed with the left and right stereo signals. This com-
posite signal is then recorded onto the stereo disc. On playback, the signals
are matrixed again with proper phase shifting to recover the rear channels.
This technique was used in the so-called QS and SQ quadraphonic systems.
Incidentally, the demise of quadraphonic sound had little if anything
to do with the matrixing of the signals and consequent lack of discrete-
ness of the four signals. See also ambisonics.
Matrixing is also used to obtain left and right stereo signals from the
mid and side channels of the m-s stereo recording system. A common mis-
conception is that the M-S technique is the only one in which the width of
the stereo image can be varied. (This is done by controlling the level of the
M (mid) signal compared to the S (side) channel.) Any stereo signal, regard-
less of which microphone technique is used, can be manipulated the same
way. The sum and difference signals are obtained as in FM broadcasting,
and the level of the difference signal is varied before being rematrixed to
obtain the left and right signals. As the difference is reduced, the stereo sep-
aration, or width of the stereo image, will decrease, and vice versa. Inci-
dentally, a circuit that does this double matrixing and level controlling
is sometimes called a shufer circuit, and it was rst described by Alan
blumlein.
In three-step processing in the manufacture of records, the mother
is sometimes called the matrix.
Maximum-Length Sequence, MLS An electronically generated test signal
that has a at energy-versus-frequency curve over a wide frequency
range. The signal resembles white noise in this respect, but it is actually
periodic, with a relatively long period, or a very slow repetition rate. The
period may be as much as several seconds, and in general, the longer the
period, the more uniformly the energy in the signal will be distributed in
frequency. Because the signal looks and sounds like random noise, it is some-
times called pseudorandom noise. True random noise has a random dis-
tribution of amplitudes as well as frequency, but pseudorandom noise does
not; that is, it has a low crest factor. The maximum-length sequence is
generated by a digital process, and the signal looks like a square wave
with a random placement of the zero crossings. The name maximum-length
sequence is a bit of a misnomer, because the actual length of the sequence
can be just about anything.
The maximum-length sequence, in conjunction with a two-channel fft
analyzer, is often used as a test signal for measuring the frequency
response of various devices such as loudspeakers and other transducers.
See also deconvolution.
Maximum Output Level, MOL For an audio device such as an analog
tape recorder or cassette recorder, the MOL is generally taken to mean the
output signal level that results in 3% harmonic distortion at low fre-
quencies and usually 3% intermodulation distortion at high fre-

233
MC Phono Input

quencies. Any higher signal output than the MOL will result in rapidly
increasing distortion. The MOL is not an absolute level in the sense of
a known voltage or power level but is a function of the device itself and
also a function of frequency. The MOL of a recording device is dependent
on the type of tape used, and frequently different tapes are evaluated by
comparing their MOLs when used on a particular machine. The shape of
the curve of the MOL plotted against frequency is a meaningful characteristic
of a device such as a tape recorder, and it depends on many factors, includ-
ing pre-emphasis and de-emphasis. To achieve maximum signal-to-
noise ratio, a device or system should have an MOL curve with the same
shape as the curve of the maximum peak voltage level versus frequency of
the signal it passes. The maximum peak voltage level of a music signal
depends on the music but generally is lower at high frequencies than at
mid and low frequencies. See also nab equalization.
MC Phono Input A set of input terminals of a preamplifier designed to
accept the signal from a moving coil phono cartridge is called an MC
phono input. This distinguishes it from the previously established ordi-
nary phono input designed for a moving magnet cartridge, which
takes a much higher input voltage and has a different input impedance
and capacitance.
MDM See modular digital multitrack.
Mel The mel is the psychological unit of pitch. One thousand mels is
dened as the pitch of a 1,000-hertz pure tone whose loudness level
is 40 phons. A pitch judged twice as high would be 2,000 mels, and a
pitch three times as high would be 3,000 mels, etc.
This unit is completely foreign to musicians, to whom the concepts of
twice as high, etc., are meaningless, and it is seldom used, even by the
psychologists!
Mercury Cell A type of dry cell using zinc, mercury, and potassium hydrox-
ide as the active elements. It has a fairly long shelf life and very good volt-
age regulation, that is, its voltage remains high until it is exhausted.
It is likely to explode if recharged.
Metadata Parameters sent along with a particular data transmission, as
opposed to the actual data carried in the transmission. These data param-
eters usually include type of codec being used, number of channels, chan-
nel format, type of data encryption, etc. In dolby digital, metadata
specically refers to the parameters that travel alongside the audio in the
Dolby Digital stream as auxiliary data. The metadata here provide scala-
ble decoding information about the audio that can be interpreted in dif-
ferent ways by different receivers, allowing a producer to tailor a programs
mix to the playback environment without requiring the medium to store
multiple versions, e.g., a 5.1 mix and a stereo mix. Not to be confused with
Metadata, a trademark owned by the Metadata Corp.
Metal Particle Tape A type of magnetic tape, mostly used in the better qual-
ity cassettes, that uses microscopic particles of iron rather than iron oxide
as the magnetic medium. Metal tape is capable of much better performance

234
 Microphone

than oxide tape, especially at high frequencies, but it requires different

equalization and bias. The improvement in quality is most pronounced
at low tape speeds, and this is the reason it is seen most often in cassettes.
Metal particle tape or metal evaporated tape is also used in video record-
ing, such as Hi-8 video cassettes and most digital audio recording.
Meter An audio level indicating device; see vu meter and ppm. Also the
si unit of length, equal to the length of the path traveled by light in a vac-
uum during 1/299,792,458 of a second. In music, meter also means the num-
ber of beats contained in each measure.
Meter Bridge The area of a mixing console that holds the audio level indi-
cating devices, usually at the top of the mixer. Meters are sometimes an
option; hence, the entire assembly holding the meters may bridge the width
of the mixer. Even if the meters are integral, the area is usually called a meter
bridge anyway.
Metrology The science of precision measurement. Not to be confused with
meteorology, the science of weather.
Mho Archaic term for the unit of conductance. It represents the recip-
rocal of resistance, and this is the reason for the spellingohm back-
wards. The modern unit for conductance is the siemens.
Mic Abbreviation for microphone. Mike is also seen.
Mic Cable See microphone cable.
Mic Flag A little sign containing a network or station identication logo
clamped to a microphone used in TV broadcasting and press conferences.
Used as a bit of subtle advertising.
Microbar Originally, the microbar was dened as a pressure of one mil-
lionth of 1 atmosphere, or 1.013 dynes per square centimeter. The standard
atmosphere, or bar, is dened as the average air pressure at sea level at 45
degrees latitude, and is equal to 1.033 kilograms force per square centime-
ter. The new si denition sets the bar at exactly 10 newtons per square meter,
and thus the microbar became 1 dyne per square centimeter, or .1 pascal.
Microgroove In 1948, the Columbia Records division of the Columbia
Broadcasting System introduced the 3313-rpm phonograph record, which
had grooves designed to be traced with a stylus of .001-inch radius. The
existing standard at the time was the 78.26-rpm record with grooves of
.003-inch tip radius. Columbia established the trademark LP micro-
groove for the system, which quickly set a new standard for the future
of commercial recordings. It was not a new idea, however, for the 3313-
rpm speed had been used for years for radio transcriptions, and Thomas
Edison had introduced a long-playing record system with extremely ne
grooves before 1929. Edisons LP record played for 20 minutes per side at
a speed of 80 rpm!
Microphone An electroacoustic device that delivers an electric signal
when actuated by a sound. A microphone consists of an acoustic system
that supplies mechanical (acoustic) energy to a transducer, which con-
verts the energy into electrical energy.
Microphones are classied by their acoustical parameters, by their

235
Microphone Cable

method of transduction, and by their directional characteristics, or polar

patterns, forming a fairly complex hierarchy. Acoustically, microphones
are classed as pressure or pressure gradient or a combination of the
two. Within these classes, microphones can have various types of trans-
ducing elements, including carbon, condenser, crystal, dynamic (or
moving coil), and ribbon.
Microphones of all these classes can have a variety of directional char-
acteristics, or polar patterns. These include omnidirectional, car-
dioid, figure-8, super-cardioid, hyper-cardioid, and shotgun.
For more detail, see each of these headings and also pzm; electret
microphone; sound field microphone; parabolic microphone;
and line microphone.
Microphone Cable Cable has more than one conductor, whereas wire
has only one conductor. The common professional microphone cable will
have two, three, or four signal wires inside of a shield and then an outer
jacket. Good microphone cable should be durable but exible and have
low resistance and capacitance. The wires used should have many ne
strands for exibility and to resist breakage. The insulating materials
should resist age, chemicals, and soldering heat. The shield should be
tightly braided wires, or one or two spiral layers of wires. Such cable is
meant to be exed and as such, should not have foil shields employed
except in xed cable installations.
Microphone Preamplier The signal from a microphone is relatively low
in level, and the amplifier that rst amplies it must be carefully designed.
One consideration with such small signals is low noise level. The preamp
usually has a differential input able to handle a balanced signal to pre-
vent the addition of externally induced noise. This used to be attained by
placing an input transformer at the preamp input. The transformer allows
the line to be balanced, but also allows a step-up in voltage, giving volt-
age gain without adding signicant noise. Most input transformers have
a turns ratio of about 10, giving a gain of 20 decibels. This transformer
requires careful design also. It should have a large primary inductance
to prevent loading the microphone unduly at low frequencies, and it must
have a wide frequency response. It also must be well shielded against
stray magnetic elds to prevent pickup of 60-hertz hum.
Today most high-quality preamps use balanced circuits without trans-
formers and are designed for extremely low noise levels and distortion.
Microphonics Some components in audio devices are sensitive to vibra-
tion in the sense that they act like microphones. An example is some vac-
uum tubes, which, when thumped, cause a tinkling sound to be added to
the signal. solid-state devices are much less prone to microphonics,
but transducers, such as phono cartridges, can be very microphonic.
Some audio cables, such as microphone cables, can also add noise to
the signal when they are vibrated. Audio cable noise is correctly called tri-
boelectric noise, which is noise induced in a capacitor if the dielectric (insu-
lator) is deformed. Audio cables consist of insulated and shielded wires,

236
 MIDI

and are thus miniature capacitors, making them susceptible to triboelec-

tric noise.
Microprocessor A digital integrated circuit, or chip, that is the heart
of a small computer or calculator. Working under the control of a program
stored in the digital memory, the microprocessor does all the mathemati-
cal operations needed in the computer. Microprocessors are so fast and exi-
ble in operation and so inexpensive that they are used in many audio
applications, from the control circuits in cd players to electronic music
synthesizers.
The rst true microprocessor was the type 4004 made by Intel and intro-
duced in 1972.
MIDI The Musical Instrument Digital Interface is a digital protocol origi-
nally designed in 1983 for use between electronic music synthesizers of
various manufacture. With MIDI, musical performance information is
transferred as computer data commands, rather than transferring actual
sound les. There is a long list of commands meant to direct devices to con-
vey nuances of performed music. Since 1983, MIDI has evolved into a
diverse tool for creating, editing, and performing music and sounds, and
even sound generation for PC games, show control for theatrical devices
such as lights, recording studio equipment controllers, and even ring tones
for mobile phones.
In the basic music setup, several instruments can be connected on the
MIDI, and this can allow the keyboard on one to control the generation of
sounds on the others. This was the original purpose for the MIDI. A MIDI-
equipped instrument has a MIDI-in and a MIDI-out, each a 5-pin DIN con-
nector. The standard allows for the control of sixteen different devices and
functions at the same time. (Actually, the control is not simultaneous, for
the MIDI is a serial port, controlling the functions in rapid sequence. But
it is fast enough that it sounds like simultaneous control.) The MIDI can
also be used to interface certain computers to electronic musical instruments,
and can thus be used to transmit voices and musical sequences stored in
digital form. It also can be used to transcribe musical notation via the com-
puter, and as such, editing a performance is much like word processing.
The use of a MIDI-controlled synthesizer rather than sampled audio les
from a computer has the advantage that far less data need be processed, trans-
mitted, and stored. High-quality audio les, such as .WAV les, require large
amounts of disc storage space, however the availability of hard drives with
many gibibytes of space instead of mebibytes has partly nullied this
advantage. A MIDI-controlled synthesizer usually needs less than 10 kibi-
bytes of code to generate each minute of resulting sound. The MIDI code
only needs to tell the synthesizer how to produce the wanted sounds.
The gure below shows a simple three-voice MIDI system. The MIDI
Sequencer is a device that stores information about a series of keystrokes
and saves it in memory. It then can store similar information about another
series of keystrokes, etc. Then, on command, the sequencer will send the
information to the synthesizers at the same time resulting in three separate

237
Mid-Range

A Three-Voice MIDI System

parts being played at once. More synthesizers and other MIDI-controlled

devices can be added to increase the number and complexity of the result-
ing sounds.
The greater MIDI world is large, diverse, and beyond the scope of this
book.
The MIDI Manufacturers Association, or MMA, is an organization
of companies that produce MIDI-controlled devices and is in charge of
maintaining and updating the MIDI specication. It has done an impres-
sive job of it over the last 20 years of MIDI existence. The MMA web site
<www.midi.org> is a gold mine of information, both technical and practical.
Mid-Range The frequency span of the middle of the audio range. The mid-
range is usually considered to be from about 200 hertz to 2,000 Hz or so.
Most of any music signal is in the mid-range, and because this range is the
easiest to reproduce with audio equipment, it is sometimes neglected from
the standpoints of distortion and smoothness of frequency response,
especially in loudspeaker systems.
Mid-Range Smear A type of sonic degradation in an audio system that is
the result of nonlinearities in the equipment interacting with the broad band
of frequencies present in music to produce a veil of distortion products
throughout the mid-range. It is difcult to test a device for mid-range smear,
but the innovative spectral contamination measurement invented by
the late Deane Jensen and W. Gary Sokolich seems to correlate well with
listening tests.
Mike See mic.
Mil Thousandth of an inch. Recording tape thickness is measured in mils,
the most common size being 112 mils. One-mil tape allows 50 percent more
recording time per reel, but suffers from higher levels of print-through
and is not as stable mechanically. Some tapes, most notably those found in
cassettes, are thinner than 1 mil.
Mil Spec Short for Military Specication, an item that meets ofcial U.S.
military contract terms. This is supposed to indicate the highest quality,
and is sometimes seen referring to electronic components. Such parts are

238
 Miniplug

expensive, and usually not worth it, so commercial grade parts are more
commonly used. There are Mil Specs for everything the military uses, from
toothbrushes to cheese.
MiniDisc, MD A digital format developed by Sony that uses a 64-mm
diameter, magneto-optical rewritable disc in a thin rectangular plastic car-
rier. Their so-called ATRAC codec (Adaptive Transform Acoustic Coding)
samples at 44.1 kHz but compresses the data to t up to 80 minutes of stereo
audio onto the small disc. The codec has continued to evolve and is now
at ATRAC3-plus, which also includes reduced fidelity LP2 (2X long play,
or 160 minutes) and LP4 modes (320 minutes). The format was apparently
intended to replace the analog cassette tape. U.S. consumers do not seem
to have completely embraced the format, although it is popular in the legit-
imate theater for sound effects and radio stations. Commercial prerecorded
MD releases (rare in the U.S.) use an optical-only playback format and are
not recordable. See also codec.
Minimalist Techniques Term applied to selection of microphones for
classical stereo recording. A minimalist, or purist, will always use the
smallest possible number of microphones to make a recording; this means
only two, if at all possible.
The use of multiple microphones for stereo recording has become stan-
dard for popular music when there is no desire to reproduce in the listen-
ing room the sonic event that happened in the recording studio. The
recording is the musical event, and the mixing of microphones and signal
processing techniques is justied.
But for classical music, it is usually desirable to reproduce the sound of
a concert hall in the listening room, and this is best done without complex
mixing and signal processing. The most natural sounding recordings are
usually the ones made with the simplest microphone conguration.
This is not to imply that minimalists agree on what the simplest setup
consists of, however. Some recording engineers insist on coincident
microphones and some swear by spaced-apart microphones. Some insist
on directional microphones, whether or not they are coincident or spaced,
and some will only use omnidirectional microphones.
Minimum Audible Field, MAF The same as the threshold of hear-
ing when measured under free-field conditions. The threshold of hear-
ing is traditionally measured by presenting the subject with the test sounds
via earphones, but it has been found that the differences in the physiology
of different subjects result in differing sound pressure levels at the
eardrum due to unpredictable reections within the ear canal, pinna, and
earphone. Thresholds measured under free-eld conditions are considered
more accurate, although much more difcult to determine because of the
extremely quiet surroundings required.
Miniplug A common audio plug 3.5 mm in diameter, used for portable
stereo headphones and other consumer audio connections. It is a smaller
version of the common 14-inch phone plug. The trs form is the most com-
mon, used for stereo, but TS versions exist. It mates with the minijack.

239
Mix

Mix In the production of a commercial recording, the result of the mixdown

is often called the mix. Also a verb meaning to perform a mixdown.
Mixdown In an analog recording studio, mixdown is the recording of a
two-track master tape by mixing the reproduce outputs of several tracks
from one or more multitrack tape recorders.
Most studio recordings are made with multitrack recorders to allow the
performers and the record producer exibility. Often each instrument in
an ensemble will be recorded on its own track. Then, the relative balance
and timbre of the instrument can be manipulated by signal processing
devices while being mixed into the composite, or master, recording.
Frequently as many as twenty-four or more separate tracks are used in
a mix, and it becomes impossible for one person to handle all these vol-
ume controls simultaneously while the master is being recorded. In order
to facilitate this step, some recording consoles provide automation, or auto-
mated mixdown, as it is sometimes called. In this type of automation, the
volume settings versus time of all the controls are memorized by the con-
sole and are controlled as the tape is being replayed. The tape may be repeat-
edly played with different settings to rene the overall effect, after which
the master is recorded.
Today (2004) most mixdown is done with daws using software such as
ProTools, but many engineers still use large format automated consoles (or
digital ones).
Mixer Either a device that allows several signals to be combined into a single
signal or a person who operates the console controls during a mixdown.
Portable mixers for combining the outputs of several microphones are
common and are used extensively in remote recordings. They will con-
tain a preamplifier for each microphone input, individual gain con-
trols for each input channel, a master gain control, and a low-impedance
output circuit to feed the signal into a control console or recording device.
More complex mixing units that also contain equalization are usually
called recording consoles or boards.
MLP, Meridian Lossless Packing A lossless audio coding scheme devel-
oped by Meridian Audio Ltd. MLP has been selected as the optional cod-
ing scheme for use on dvd-Audio, as well as other transmission, storage
and archiving applications. It is a true lossless coding technology, in that
the recovered audio is bit-for bit identical to the original. Unlike percep-
tual or lossy data reduction, MLP does not alter the nal decoded signal in
any way, but merely packs the audio data more efciently into a smaller
data rate for transmission or storage. It is simple to decode and requires
relatively low computational power for playback.20
MM Phono Input A set of input terminals of a preamplifier designed
to accept the signal from a moving magnet phono cartridge. For many
years this was the most common type of phono input on a stereo system.

20. Thanks to Dennis Bohn, Rane Corp.

240
 Modulation Wheel

The input impedance was standardized at 47 kohms, and riaa phono

equalization was applied to the signal. gain was much higher than on
a similar line level input such as aux. See also mc phono input.
Modular Digital Multitrack, MDM A multitrack digital audio recording
system that records digital audio data onto a videocasette using a rotating
drum in the same manner as a dat or video recorder. MDMs are expand-
able by locking together up to 16 MDM modules for 128 tracks. There are
two popular MDM standards: the Alesis adat and compatibles which
record on S-VHS tape, and the Tascam DA-88 and compatibles, which record
on Hi-8 video tape. Other examples of MDMs include the Akai A-DAM,
and Yamaha DMR8/DRU8. Also called a modular recorder.
Modular Synthesizer See synthesizer.
Modulation See amplitude modulation and frequency modulation.
Modulation Index In amplitude modulation, the modulation index is
the ratio (expressed by a percentage) of the peak amplitude of the mod-
ulating signal to the peak amplitude of the carrier. If the modulation index
is 100%, the carrier will be turned completely off at the negative peaks of
the modulating waveform.
In the case of frequency modulation, the modulation index is the ratio
(again expressed in a percentage) of the frequency deviation to the actual
frequency of the modulating signal. The frequency deviation is the difference
between the frequency of the unmodulated carrier and the maximum higher
or lower frequency it will reach when being modulated.
In each case, the modulation index is proportional to the amplitude of
the modulating signal.
Amplitude modulation greater than 100% produces asymmetrical clip-
ping, and large amounts of harmonic distortion. This is sometimes called
splatter in am broadcasting.
The frequency modulation index can be much greater than 100%, how-
ever, and some electronic music synthesizers use this deep modulation
to attain complex musical waveforms. One of the rst commercial appli-
cations of this technique was in the Synclavier, made by New England Dig-
ital Corporation.
Modulation Noise Noise that is present only in company with a signal
is called modulation noise. In analog tape recorders, the recording process
has a certain granularity due to the fact that the magnetic characteristics
of the tape are not completely uniform. The magnetic domains are of nite
size. A recorded signal has an irregularity that sounds like the addition of
noise.
In digital audio systems, there is also an uncertainty in the level of the
signal because of quantization in the analog-to-digital converter.
This uncertainty also sounds like added noise and is not present if the sig-
nal is not present.
Modulation Wheel A modulation wheel is a hand-operated wheel usually
at the left side of a keyboard. When rotated, it causes a change in an effect
such as vibrato. Its effect is programmable by the user.

241
Modulometer

Modulation Noise

Modulometer A meter, similar to the vu meter in appearance, that responds

to the peak signal level rather than the average level.
The modulometer is a better choice for a tape recorder than a VU meter
because short peaks in the signal can easily saturate the tape but still not
be detected by the VU meter, which has a much slower response time. This
is especially true in digital tape recorders, where any overload, even
momentarily, results in hard clipping and distortion.
Many modern metering systems for tape recorders do not have any mov-
ing parts at all, substituting instead a series of small lights that provide an
animated display of the signal level. Most of these units show the peak level
as well as the average level.
MOL See maximum output level.
Mole A large number of small components such as atoms, molecules, elec-
trons, etc. The mole is quantitatively equal to 6.02214 x 1023, a very large
number indeed, which is also known as Avogadros number. It is dened
as the number of atoms in 0.012 kilograms of carbon, which is the number
of atoms in a mass of carbon in grams equal to its atomic number. The mole
unit is most useful in chemistry, where it simplies calculations in chemi-
cal reactions, but it is sometimes used by computer equipment designers
and engineers who talk about moles of bits or bytes.
Monaural Literally, one hearing. Monaural refers to a sound system with
only one channel, regardless of the number of loudspeakers used, as
opposed to stereophonic, which must employ more than one indepen-
dent channel.

242
 Mono Compatibility

Monitor Originally, monitor referred to a loudspeaker in the control booth

of a recording studio or sound reinforcement system that allowed the oper-
ators to hear the programs they were controlling.
In more recent times, the term has been used by stage performers to mean
the loudspeaker or loudspeakers placed on the stage to allow them to bet-
ter hear themselves. A better term for this is foldback.
Monitor Mix A mixed signal in a sound reinforcement console that is sent
to the monitor speakers on the stage. Because the monitors are near the
microphones, this signal is usually highly equalized to reduce the ten-
dency for acoustic feedback. A modern mixing console can produce sev-
eral monitor mixes for use by different performers on stage. Sometimes, a
mixing console is dedicated only to monitor mixes. See also foh.
Mono See monaural.
Mono Bus A mixing bus in a recording console that generates a single-
channel, or monaural, signal for use with a single-channel tape recorder
or a broadcast feed. Most modern consoles have a mono output indepen-
dent from the regular stereo outputs.
Mono Compatibility The ability of a stereo musical signal to be mixed
into a single monaural channel without violence to the timbre of the
signal. The result of the mix of two stereo channels would ideally sound
like a signal that was recorded with a single microphone.
Mono compatibility is a matter of degree and never can be perfect, for
any two signals when mixed will contain frequency components that will
sometimes partially cancel due to the relative phase angles between them.
Stereo recorded with widely spaced microphones (a-b stereo) has rela-
tively poor mono compatibility because of the large time differences in the
two channels. At low frequencies, these time differences can easily amount
to 180 degrees of phase shift causing almost complete cancellation when
mixed. It might seem that at high frequencies the problem would be worse,
but actually the phase differences are so very great and so complex that
they are nearly random, and when mixed, the result is very close to the
original spectrum.
One way to increase mono compatibility is to arrange things so the low-
est frequencies of the two channels are in phase with each other. This can
be done by placing the bass instruments in the center of the stereo spread;
in other words, making them monaural rather than stereo. Another way is
to use coincident microphone recording techniques, which automat-
ically prevent large phase differences at low frequencies.
There are several reasons why mono compatibility is desirable. The most
important is that the signals are easier to cut onto a stereo phonograph
record. This is because out-of-phase signals are represented by vertical
motion of the stylus, which is limited by geometric considerations.
Another reason is for fm broadcasting of stereo in case the listeners have
only monaural radios.
The need for mono compatibility is much less today that it was 20 years
ago. The compact disc has no requirement for uniform phase of the sig-

243
Monotic

nals at low frequencies, and stereo broadcasts are seldom heard over mon-
aural radios, especially where high quality of low-frequency reproduction
is attained.
Monotic Literally, with one ear. Generally refers to a sound presentation
where only one ear hears the sound (e.g., sound through a headphone).
MOS Film or video being shot without sound. Legend has it that it stands
for Mit Out Sound, which was supposedly spoken by early German
moviemakers in Hollywood. Pronounce each letter separately, EM-O-ESS.
MOSFET See field effect transistor.
Mother In the three-step processing technique for making phonograph
records, the mother (a metal replica of the acetate) is the second step, from
which the stampers are made. See processing.
Motional Feedback A type of mechanical negative feedback where the
actual motion of the cone of a low-frequency loudspeaker is used to
generate a signal that is fed back to the amplifier. The motion of the cone
itself is then inside the feedback loop, and distortion can be signif-
icantly reduced. The technique was rst described and patented by the Eng-
lishman P. G. A. H. Voight in 1924. This actually predates the invention of
electrical negative feedback in ampliers by H. Black of Bell Telephone Lab-
oratories. Motional feedback is suited for use only at low frequencies, usu-
ally below 500 hertz.
For motional feedback, some method of sensing the cone motion must
be used, and this motion must be converted or transduced into a voltage
fed to the amplier. There are many possibilities here, and many different
systems have been tried over the years. Suitable transducers can sense
the displacement, the velocity, or the acceleration of the cone. One of the
simplest is simply another coil next to or on top of the voice coil. This
coil will have a signal induced in it proportional to the velocity of the cone,
but it will also have another signal induced in it by transformer action from
the voice coil itself. Various schemes have been devised to separate these
two components electrically.
Another possibility is the use of a piezoelectric accelerometer
attached to the cone. This provides a signal proportional to the accelera-
tion of the cone, and this can be converted to a velocity signal by electri-
cal integration. This method is relatively costly and complex, but it has
been used commercially.
In all motional feedback schemes, great care must be exercised to ensure
that the system is stable. This means that phase shift between the drive
signal and the feedback signal must be accurately controlled.
Motional Impedance The impedance of a dynamic loudspeakers voice
coil is made up of two parts: the electrical impedance of the coil of wire
itself (the blocked impedance) and the impedance caused by the motion
of the moving parts of the speaker, or motional impedance. As the coil moves
in response to an applied voltage, it will have induced in it a back emf,
which opposes the applied voltage. This back EMF effectively increases

244
 Moviola

the electrical impedance, especially when the coil is free to move easily,
such as at resonances. It is thus possible to learn a good deal about the
moving parts of a loudspeaker by studying the curve of impedance versus
frequency.
Motor Boating A low-frequency oscillation caused by certain types of insta-
bility, usually in a power amplifier. The name comes from the charac-
teristic plop-plop sound it produces when heard on a loudspeaker.
Moving Coil Cartridge A phonograph cartridge that uses two tiny coils
of wire connected to the stylus assembly as signal generating elements
is a moving coil cartridge. The coils are in a magnetic eld, and are caused
to move by the motion of the stylus in the groove, this motion inducing the
signal voltage in them.
Some of the earliest types of phono pickups were of the moving coil
design, but the moving magnet types almost completely replaced them
when stereo records were introduced. Moving coil cartridges are charac-
terized by very high price, very low output voltage, low distortion, and
extended high-frequency response resulting from a very high reso-
nant frequency due to light weight of the moving parts. Sometimes this
high-frequency response is a negative quality factor because the resonant
peak in the response, usually at from 30 to 50 kilohertz, effectively
amplies the ultrasonic noise present in the record groove.
Although the noise is above audibility in frequency, intermodulation
distortion in the preamplier will result in added noise in the audible
range. For this reason, moving coils generally sound worse than other types
when used with less than excellent-quality preamps. The old adage that
the most money should be spent on the transducers in a sound system
is contradicted in this case.
The sensitivity of MC cartridges is usually less than 200 microvolts
rms per centimeter per second of recorded velocity. They should be ter-
minated in less than 10,000 ohms, sometimes much less.
Moving Magnet Cartridge The most common type of phonograph car-
tridge is the so-called moving magnet type. It has a small magnet con-
nected to the stylus assembly, and motion of this magnet causes a
voltage to be induced in coils of wire that are nearby. This voltage con-
stitutes the signal output of the cartridge.
Moving magnet cartridges are relatively inexpensive, and have accept-
able frequency response and output signal level, and they have been
very popular for many years. Their high-frequency response is somewhat
sensitive to the electrical input impedance of the preamplifier they are
connected to. Long cables between the cartridge and the preamp should
be avoided also because cable capacitance will affect high-frequency
response as well. MM cartridges usually have a sensitivity of from 0.5 to 2
millivolts per centimeter per second of recorded velocity, and they are meant
to be terminated with more than 10,000 ohms impedance.
Moviola The trademarked name of a company that manufactured lm

245
MP3

editing equipment, and by common usage, their products. In 1924 Iwan

Serrurier turned his home movie projector into a personal lm viewer for
Hollywood editors. The classic upright Moviola that evolved became an
industry standard and allowed a lm editor to manipulate lm and syn-
chronized soundtracks with relative ease. It could hold reels of lm and/or
soundtracks on vertical reels and the operator could watch a small optical
screen while running the reels forwards or backwards. The nal evolution
of the Moviola was the atbed editor, where the lm sat on horizontal plat-
ters on a large table-like mechanism. Today, Moviola deals with digital pro-
duction products and education, however you can still rent a lm Moviola.
MP3, MPEG-1 Audio Layer III, MPEG-2 Audio Layer III MP3 is a general
name for a digital audio le encoded to conform to any of several lossy com-
pression standards set forth by the MPEG-1 or MPEG-2 Audio Layer III. The
lename extension .mp3 was originally created for software handling
MPEG-1 Audio Layer III, and the general name has stuck. Any of several
sampling rates and bit stream rates can be used. MPEG-2 Audio Layer III
adds functionality to MPEG-1 Audio Layer III and most of it is backward
compatible. Lower sampling rate les encoded in MPEG-2 Audio Layer III
are also called MP3 les. MP3 les should not be called MPEG-3there is
no such thing. Audio Layers I and II are less complex and not generally suit-
able for higher quality audio. MPEG-1 Audio Layer III uses sampling rates
of 32, 44.1 or 48 kHz and bit rates from 32 to 320 kbits/s, while MPEG-2 Audio
Layer III adds 16, 22.05 and 24 kHz sampling rates and bit rates from 8 to
160 kbits/s. Thankfully, the software gures this all out. As a rule, the lower
the sampling and bitstream rates, the worse the delity. File size savings
can be dramatic from normal CD.
MPEG Stands for Moving Picture Experts Group. It is a working group of
iso, and its full name is ISO/IEC JTC 1/SC 29/WG 11 - Coding of Mov-
ing Pictures and Audio. Its mission is the development of international
standards for compression, decompression, processing, and coded repre-
sentation of moving pictures, audio, and their combination. MPEG held its
rst meeting in Ottawa, Canada, in 1988. An MPEG Audio Subgroup con-
centrates on the audio aspects. The audio layers of the MPEG standards
specify decoding standards but leave the perceptual encoding open. Thus,
improvements are constantly being made that are completely compatible.
MPEG-2 A term referring to a video and audio digital compression stan-
dard created by the mpeg, generally intended for uses such as digital TV
and dvds. The MPEG-2 video codec uses variable bit rates to allocate
higher bit rates for complex scenes with a lot of motion and lower bit rates
for more static scenes. The audio layers include multichannel extensions
to the MPEG-1 standards, and the addition of the high-quality aac
(Advanced Audio Coding) system.
MPEG-4 MPEG-4 is a multimedia streaming codec aimed primarily at
game applications. In addition to streaming video and digital audio, it allows
for the transmission of midi, facilitates the transfer and playback of DLS-
2 les, and incorporates a user-congurable synthesis language. This lat-

246
 M-S Stereo

ter feature is called the Structured Audio Orchestra Language (SAOL),

allowing the programmer to specify almost any existing synthesis method
and create algorithms at one end, and have the sound played back identi-
cally at the end-users player.
MPSE Motion Picture Sound Editors. An organization of persons who work
in the lm and TV sound and music editing industry. Members are allowed
to use the initials MPSE after their names in credits. See also cas.
MPX Short for multiplex. The letters MPX when found on a button on a
cassette deck mean the button activates a 19-kilohertz notch lter to elim-
inate the fm stereo pilot so it wont cause gain mistracking in Dolby noise
reduction systems. The Dolby playback system would sense the pilot and
mistake it for high-frequency content of the music. The pilot can also cause
audible birdies by beating with the AC bias used in tape recorders.
MSB Most Signicant Bit; the rst (left-hand side) digit in a binary num-
ber. See also lsb.
M-S Stereo M-S is an abbreviation for Mid-Side, or some say Mono-Stereo,
and refers to a special type of intensity stereo invented by Lauridsen
in Denmark in the mid-1950s.
The coincident microphones used for M-S recording are usually a
cardioid facing directly to the sound source and a figure-8 microphone
facing sideways. The gure 8 picks up the left half of the source with one
phase and the right half with inverted phase. When the signal is added
to the signal from the cardioid, the signals from the left side add together,
while the signals from the right side subtract due to the phase inversion.
The combined pattern of the two microphones is like a cardioid facing 45
degrees to the left. This is one of the stereo output signals from the M-S sys-
tem. Then, the gure 8 signal is subtracted from the cardioid signal (actu-
ally, it is phase-inverted and added), and left sides cancel and right sides
add, simulating a cardioid facing 45 degrees to the right. This is the other
output signal. This addition of the signals in two ways is sometimes called
matrixing.
One might well ask why this is done, when two cardioids 90 degrees
apart will do the same thing. The only reason is that with the M-S system,
the relative level of the cardioid signal can be varied with respect to the
gure 8 before matrixing, and this will vary the apparent width of the stereo
image. (This is assuming one would rationally want to vary the apparent
width in the rst place.)
Additionally, it has perfect mono compatibility, and assuming one
records the individual mic outputs without matrixing (as opposed to
recording the decoded stereo), the apparent width can be changed at any
time such as in postproduction.
One technical problem with the M-S system is that each output signal
is the result of the addition of two microphone signals, and these micro-
phones cannot occupy the same point in space. Therefore, there will be slight
differences in phase and amplitude of the signals being added, causing
an added roughness in the frequency response curve of the combined

247
MTS

output. The M-S system also assumes that the cardioid and gure-8 micro-
phones have identical frequency responses and polar patterns that are
constant at all frequencies. Neither assumption is true.
M-S recording has always been more popular in Europe than in the
United States, but it seems to be declining in popularity compared to other
intensity stereo techniques in both places. See also blumlein, alan; in-
tensity stereo.
MTS, Multichannel Television Sound The method used for broadcasting
stereo sound and alternate language channels with ordinary analog NTSC
television broadcasting. Added on to the format in 1984, MTS is built into
almost all analog ntsc television sets and receiving devices such as VCRs
(Video Cassette Recorders).
Muddy A subjective term that describes a type of distortion that reduces
the clarity or transparency of the sound of a musical instrument, particu-
larly transients. For instance, the reproduced sound of a collection of
sleigh bells should allow the listener to detect the individual bells, each with
its sharp attack intact. intermodulation distortion will add new fre-
quencies among the rich collection of natural harmonics of the bells and
will also add low-frequency components not originally present. The bells
sound thick.
The sound of massed string instruments is extremely complex, for the
many harmonics of each instrument are greatly enriched by the fact that
all the instruments are not in perfect tune with each other, giving rise to an
overall spectrum that is smeared in frequency rather than consisting of
only discrete harmonics of the fundamental frequency. We perceive this
smearing as a chorus effect, and we say it adds richness to the tone. Inter-
modulation greatly complicates this spectrum by adding new frequencies
that bear no harmonic relation to any of the frequencies originally present.
This confuses our ears pitch-sensing mechanism, and we say the sound is
muddy.
Mu-law A nonlinear codec, often used for telephone-grade digital audio.
It is logarithmic instead of linear in its quantizing, putting more accu-
racy into the lower level signals and reducing the word size of samples.
Previously supported on many computer platforms, it seems to have been
superceded by current popular digital audio compression schemes such as
AAC, mp3, RealAudio, and Windows Media. Sometimes seen as u-law.
Mult A connection, usually in the form of an auxiliary box, that shorts two
or more signals together. If one or the other or both of the signals are at a
low impedance, the mult will cause distortion and a reduction in level.
The mult must be used with care and should not be used as a substitute
for a mixer. Likely derived from the term, multiple.
Multi-Effects Processor An audio signal processing unit that combines sev-
eral different audio effects. Originally, effects devices would only have one
or two functions such as reverb, phasing, anging, de-essing, limiting, etc.
Modern dsp meant that many effects could be economically done in one

248
 Multipattern Microphone

device. Some detractors argue that the effects are seldom as good as the
best separate units.
Multimeter A piece of test equipment that measures several electrical
parameters such as resistance and AC and DC voltage and current.
The original multimeter is thought to be obsolete by many technicians due
to the introduction of digital multimeters, but there are many die-hards
who still like analog multimeters better because when measuring a volt-
age or current that is varying with time, you can see the variations in real
time on the meter. Digital meters present you with a blinking display of
different numbers that are not as convenient to interpret. See also vom;
dmm; vtvm.
Multipath Distortion Multipath distortion is a type of distortion afict-
ing fm and television broadcasting. It is the receipt of the transmitted sig-
nal over more than one path due to reections from hills, buildings, etc.
Because the path lengths are different, there is a delay between the various
signal arrival times. In analog television, this causes the familiar ghosts, or
multiple images, on the screen.
In FM radio, multipath manifests itself as an undesirable distortion of
the high frequencies. The effect is most noticeable in a car FM receiver, where
the signal will fade in and out very rapidly as the car moves among the
reected signals. This is sometimes called picket fencing. The effect does
not occur in am broadcasting because of the relatively much longer wave-
lengths of the transmitted carrier signals.
Multipattern Microphone A microphone that can be adjusted to have sev-
eral different polar patterns is sometimes called a multipattern micro-
phone. Some of the rst microphones of this type were the RCA ribbons
designed by Harry Olson in the late 1930s. While a ribbon microphone has
inherently a gure 8 pattern, Olson was able to modify the design to effec-
tively change the pattern from figure 8 to cardioid, and even to omni-
directional. These microphones enjoyed a long life in radio broadcasting,
recording, and motion picture sound applications, and many are still in use
today.
The most common multipattern microphone found today is a type of
condenser microphone based on a German design by von Braunmuhl
and Weber. The active element in this microphone is the cardioid condenser
capsule, which is a condenser microphone with a perforated back plate
and a sound entrance to the back plate. Sound reaching the microphone
from the rear is delayed by the back plate on its way to the back of the dia-
phragm. The same sound is delayed by diffraction on its way to the front
of the diaphragm, and therefore the pressure on the diaphragm is the same
on each side, and no output results. This means that the microphone is dead
for sounds coming from behind. Sounds from the front reach the front of
the diaphragm unimpeded, but are delayed rst by diffraction and then
by the delay in the back plate so that they arrive at the rear of the diaphragm
out of phase, causing a greater differential pressure on the diaphragm. This

249
Multipattern Microphone

Multipattern Microphone

results in a relatively large output signal for sounds coming from the front.
(This is a renement of the design of the dynamic unidyne microphone
invented by Benjamin Bauer much earlier.)
A double microphone can be made by mounting a diaphragm on each
side of the back plate, resulting in two cardioids back-to-back. When the
output signals are added, the result is an omnidirectional. If the outputs
are added out of phase (i.e., if they are subtracted), the response will be a
gure 8. If only one output is used, it is a cardioid. Thus, three patterns are
available from a single unit, or, if the responses of the two microphone ele-
ments are adjustable in sensitivity, the pattern can be continuously var-
ied from one shape to another.
This type of microphone was introduced commercially by Neumann in
1953 as the model M-49. Other manufacturers have taken up the idea, and
the microphone is very popular in recording studios. Its primary disad-

250
 Multitrack

vantage is that its active elements are somewhat large, causing problems
in uniformity of output due to diffraction effects. The ideal microphone
would be small compared to the shortest wavelength of sound it is to
reproduce. The wavelength of a sound of 20 kilohertz is about 0.5 inch,
and it is difcult to make a microphone this small with sufcient sensitiv-
ity to ensure a good signal-to-noise ratio.
Multiple Pass In error correction in digital audio devices, the mul-
tiple pass scheme uses the same hardware many times.
Multiplex Adapter An external device used with a monaural fm tuner
to allow it to demodulate fm stereo broadcasts, common in the early 1960s
when stereo broadcasting was getting started. The device that decodes an
sca broadcast is also sometimes called a multiplex adapter.
Multiplex Graphophone Grand A special type of cylinder record player
built by the Columbia Phonograph Company around 1898. It had three sep-
arate styli and three separate reproducing horns. It therefore was undoubt-
edly the rst pseudo-stereophonic record player, although it was not
advertised as such. Its advertising brochure extolled its intensity of vol-
ume and sweetness and richness of tone which seem almost beyond belief.
This is probably also the rst use of the term multiplex in reference to audio,
although with a different meaning than it has today.
Multiplexing When signals are combined in such a way that they can later
be separated, they are said to be multiplexed together. A multiplexing device
is called a multiplexer, abbreviated mux. There are many schemes for mul-
tiplexing. Probably the most common is the frequency multiplexing used
for broadcasting stereo signals over fm. In this technique, the signals are
added together to form a sum signal, and are subtracted to form a dif-
ference signal. The sum signal is used to modulate the FM carrier, and
this is what is heard if a monaural receiver is used. The difference signal
is used to modulate a 38-kilohertz subcarrier, and this modulated car-
rier is mixed with the sum signal so it also modulates the FM carrier. A mon-
aural receiver is not sensitive to the difference signal because it is translated
in frequency above the audible range by the 38-kHz subcarrier.
Multitrack A recorder, usually a tape recorder (either analog or digital) hav-
ing more than two independent recording tracks. Two-track recorders are
merely considered stereo recorders. Multitracks generally must have the
capability to enable and disable recording on any track, and to perform selec-
tive synchronization. The rst multitrack tape machines were analog
recorders, and had three tracks using 12-inch-wide tape, and it was not long
before 4- and then 8-track machines using 1-inch-wide tape were available.
The introduction of the Dolby A noise reduction system allowed the track
widths to be narrower without too much penalty in signal-to-noise
ratio. This brought forth machines with 24 and 32 tracks on 2-inch-wide
tape and 4-tracks on 14-inch tape.
It is interesting to note that the rst multitrack audio recorders were
optical recorders on motion picture lm. The Bell Laboratories produced
a 4-track optical sound system using 35-mm lm in the mid-1930s, and

251
Mumetal

the Disney studio introduced fantasound with the movie Fantasia in

1941.
Mumetal A metal with extremely high magnetic permeability that
is, in which it is easy to establish a magnetic eld. Mumetal is used as a
magnetic shield for such low-signal devices as magnetic tape heads and
microphone input transformers. It was invented in the 1920s in England
as a loading material for submarine telegraphic cables. By wrapping the
insulated copper wire with Mumetal, the inductance of the cable was
increased, reducing the signal attenuation and increasing the word-handling
capacity. The name is derived from the Greek letter (mu), which is the
symbol for magnetic permeability.
The formulation of Mumetal was patented in 1930, and the name is still
a trademark registered by Telcon Metals Ltd., of Crawley, Sussex, England.
Munchkin Effect A sound effect produced by pitch-shifting speech
upwards to create a comical, high-pitched voice, presumably named after
the effect used in the motion picture The Wizard of Oz.
MUSICAM An acronym for Masking pattern adapted Universal Sub-
band Integrated Coding And Multiplexing, a perceptual audio codec
invented by CCETT, Philips, and the irt. It was a precursor to the MPEG-
1 Audio Layer I and II. Due to the oddities of international trademark law,
in the United States it refers to a company that markets licensed Layer II
products. The original MUSICAM algorithm is not used anymore, and the
term should not be used when referring to MPEG-1 Audio Layers.
Musicasting One type of sca broadcasting by fm multiplex of a music sig-
nal designed as background music for subscribers such as supermarkets.
It uses a separate 67-kHz subcarrier to frequency modulate the main
carrier of the FM station. Musicasting was authorized by the fcc in 1955
and was quite popular for several years.
MUT, Make-Up Table A motor-driven device designed to load and rewind
motion picture lm to and from platters. 35-mm lm is shipped on 2000-
foot reels, yet theaters often use platters where all reels have been spliced
together into one 18,000-foot roll on a moving horizontal turntable. The pro-
jectionist uses the MUT to assemble the whole lm, and to take it apart again
to t onto 2000-foot reels for shipping it back.
Mute A button found on some recording consoles that, when pressed,
silences the input or other signal it is associated with. Some audio recorders
also have mute buttons.
Muting A circuit in an fm tuner that disconnects the signal output (or
shorts it to ground) when the tuner is tuned between stations is called
muting.
When an FM receiver is not receiving a carrier, its output will be very
noisy. This noise level can be greater than the signal level of most broad-
casts, and the result is a loud burst of noise when the dial is moved from
one station to another. Interstation muting removes this noise, reconnect-
ing the signal in the presence of a carrier. Walkie-talkies and communica-
tions radios refer to this as squelch. See also noise gate.

252
 NAB Equalization

Muting of the signal also occurs in some types of error correction

in digital audio systems.
Mutual Inductance The property that exists between two coils of wire, as
in a transformer, when the magnetic lines of force generated by current
in one coil induce a current in the other coil. Mutual inductance governs
the amount of current in one coil caused by a changing current in the
other one.
Mylar A plastic material used for recording tapes. Mylar is the registered
trademark of the DuPont company, which rst made the plastic. Later, oth-
ers started to make it, and the generic term became polyester.
Mylar is a tough plastic that is much stronger than cellulose acetate,
which is the traditional tape material. However, Mylar will stretch a long
way before breaking, and a stretched tape cannot be repaired. The better-
quality polyester tapes are prestretched to reduce this unfortunate ten-
dency. (Acetate tape will break with the force required to stretch Mylar, but
the break can be spliced back together.)

N
NAB The National Association of Broadcasters is an American trade asso-
ciation representing over-the-air radio and TV broadcasters. Formed in 1922,
TV interests were included in 1951 with a name change to NARTB (National
Association of Radio and Television Broadcasters). In 1958, they changed
their name back to NAB. They established various standards for radio and
television broadcasting, analogous to the bbc in England. The rst stan-
dards for magnetic tape recorders were established by NAB/NARTB, and
they included tape speed standards (15 ips for commercial and broadcast
use and 7.5 ips for less demanding requirements), and the equalization
curves for the recording and playback circuits of the recorders. See nab
equalization.
NAB Equalization In 1954, the National Association of Radio and Television
Broadcasters (NARTB) established a standard de-emphasis curve for use
in professional tape recorders operating at 15 inches per second, and it is still
in use today. A few years after the 15 ips standard, tape recorders had
improved enough to allow the 7.5 ips tape speed to achieve high-frequency
response to 15 kHz, and the same de-emphasis was adopted as the stan-
dard reproduce curve. This equalization curve was designed to take advan-
tage of the tapes existing at the time, but tape has undergone much
improvement, especially in high-frequency signal handling capacity, since
1954. The standard, therefore, does not exploit the capabilities of modern tape;
on the other hand, if the standard were changed, a great many existing tape
recordings would be obsolete. If a new standard were adopted, there could
be at least a 10-decibel improvement in high-frequency signal-to-noise
ratio, meaning a 10-dB reduction in tape hiss. See also nagramaster.

253
NAB Hub

In 1958, NARTB simplied its name to National Association of Broad-

casters (NAB), and so followed the name of the curve. More recently, the
iec has recognized the curve as a standard along with others such as ccir.
CCIR turned over their tape standards to the IEC in 1970 and CCIR became
ITU-R in 1992. Original IEC eqs are now called IEC1 (ne CCIR), and NAB
EQs are called IEC2.
NAB Hub The center hub of a 1012-inch diameter professional reel of record-
ing tape is of a standardized size, and is called the NAB hub (after the
National Association of Broadcasters, which established the standard). It
is a 3-inch diameter hole as opposed to the eia-style hole of 516 inch. A larger
hub aids tape handling.
Tape can be bought wound on NAB hubs without reel flanges, and
is called a pancake. Care must be taken in using pancakes on a tape
recorder to avoid spilling the tape off the hub. Generally, the tape is wound
off the hub once and onto a reel, and is not usually rewound onto the hub.
This is the reason many NAB hubs could be found in the trash behind
recording studios. See also din hub.
NAB Reel A 1012-inch diameter aluminum or plastic tape reel incorporat-
ing an nab hub used on professional analog tape recorders.
Nagra A Swiss line of professional portable tape recorders originally
designed to record on-location synchronous sound for motion pictures. They
were the industry standard for thirty years, owing to the quality and rugged-
ness of the recorders. They used 14-inch-wide tape and were generally
equipped with a crystal synch generator, or later, smpte time code.
The Nagra model III (mono) and IV-S (stereo) were portable, designed for
use while being carried, and were superior in performance in the areas of
flutter and wow, signal-to-noise ratio, and distortion com-
pared to high-end studio recorders. These analog models have been dis-
continued, and the digital Nagra model V (stereo onto hard disk) and model
D (multitrack onto 14-inch-tape reels) are available as portable recorders.
Nagramaster An equalization curve developed by nagra for use in
analog tape recorders that uses a larger high-frequency pre-emphasis dur-
ing recording and de-emphasis during playback to increase the audible
signal-to-noise ratio at 15 ips by about 10 dB. It is not compatible with
any other tape equalization such as nab or iec.
Nano Nano is a prex meaning one billionth.
Nanoweber The weber is the unit of magnetic flux, but is much too large
to be useful in measuring the recorded ux on recording tape. The nanowe-
ber (nWb) is equal to 10 to the minus ninth power webers, or one billionth
of a weber. The recorded level is expressed in nanowebers per meter of tape
track width to normalize the measurement as a ux density.
The standard reference level, which results when the tape recorder meter
is reading 0 VU with a mid-frequency signal, has changed over the years
as tapes have become better. The standard originated by Ampex in the 1950s,
known as Ampex Operating Level, was 185 nWb per meter; 250 nWb per
meter is more common today. Some European recorders use as much as

254
 NC Curve

NC Curves

520 nWb per meter as a reference for 0 VU. The higher the reference level,
the greater the signal-to-noise ratio will be, but the distortion will
also be greater and the headroom will be reduced, so a compromise is
required. The exact measurement of tape uxivity is somewhat complicated,
so many variations of these numbers are seen for various reasons.
Narrowband A relatively short frequency span, dening a signal or fil-
ter that encompasses a small bandwidth. Filters sharper than one-third
octave are generally considered narrowband lters.
NARTB Short for National Association of Radio and Television Broad-
casters, and was the name for the nab from 1951 to 1958.
NBS Short for the U.S. National Bureau of Standards, whose name was
changed in 1988 to the National Institute of Standards and Technology,
abbreviated nist.
NC Curve, NC Contour NC stands for Noise Criterion, and refers to the
ambient or background noise in an auditorium or room.
Because our ears are relatively insensitive to low levels of low-
frequency noise due to the fletcher-munson effect, relatively greater
amounts of low-frequency noise are allowed in auditoriums and record-

255
N Curve

ing studios, etc. In an attempt to relate the apparent noisiness of such a room
to an objective measure, the NC curves were developed. These curves are
essentially contours of equal loudness; they are spaced apart by about 5
decibels (they are not quite parallel due to the complex response curves
of our hearing mechanism); and they are numbered. NC-15 represents a
very quiet environmentso quiet that the average person would not be
aware of any background noise at all. NC-20 is noticeably noisier, and NC-
25 is considered about the limit for a good auditorium for music listening,
although many think NC-25 is too noisy.
To determine the NC value of a room, it is only necessary to measure
the background noise with a sensitive sound level meter that has an
octave band filter, and plot the measured levels on a graph of NC
curves. If the measured points are all below the NC-20 curve, for example,
the room is said to meet the NC-20 requirement.
Most of the noise in an auditorium, assuming adequate isolation from
external noise sources such as airplanes and freeways, etc., is caused by the
ventilation system, and, unfortunately, the cost of such systems rises dra-
matically with decreases in noisiness. Low-frequency noise is the most
difcult to control, so it is somewhat lucky that our ears are not very sen-
sitive to it at low levels.
N Curve See academy curve.
Near-Coincident Pair A spaced-pair microphone set-up that uses direc-
tional microphones placed approximately 7 inches apart, or the aver-
age spacing between ears on a human head. This allows for some amount
of phase difference in the two signals, but not enough to lose mono com-
patibility. ortf is the most common near-coincident arrangement, but oth-
ers include nos, and the faulkner microphone array.
Near Field The sound eld very close to a source of sound is called the near
eld. By very close is meant less than one wavelength at the frequency
of interest. It is difcult if not impossible to make meaningful sound pres-
sure level measurements, such as with a sound level meter, in the
near eld because the nature of the eld itself is very complex. Frequently
the acoustic energy is moving across the surface of the source, or maybe
there is a large air velocity near the source. standing waves are also
present in many cases if the source is deeply convoluted. In any case, it is
not possible to predict the sound level in the far eld from measurements
in the near eld, so when measuring sound pressure level, a sound level
meter must always be at least one wavelength of the lowest frequency of
interest from the source.
Near-Field Monitor A loudspeaker designed to be heard relatively
close, such as a person sitting at a mixing console with monitors on the other
side of the console.
Necessory An indispensable, or needed accessory, such as an audio cable
between a preamp and an amplifier.
Needle-Drop A recording of a music passage usually purchased from a
music archive that will be used in a movie or video production soundtrack.

256
 Neopilot

The name comes from the prior use of phonograph records to archive musi-
cal selections.
Needle Scratch Synonymous with needle talk.
Needle Talk The direct acoustic output of the stylus assembly of a phono-
graph cartridge is called needle talk, from the days when the stylus was
called the needle.
Modern cartridges, which use very low tracking forces, produce very
little needle talk, but in the days of the 78-rpm record, the needle talk could
often be heard in the room above the sound of the loudspeaker.
Negative Feedback If a portion of the output signal of an amplifier is
mixed with the input signal in an out-of-phase condition, it will partially
cancel the input signal and the gain of the amplier will be reduced. This
is called negative feedback, and its amount is measured by the amount of
gain reduction. If the gain is reduced by 10 decibels, then it is said to have
10 dB of feedback.
It so happens that distortion (both harmonic and intermodula-
tion) caused by the amplier will be reduced more than the signal will
because they are not in the input signal in the rst place. The feedback has
the effect of reducing anything that the amplier adds to the signal, which
is a reduction of distortion, or in other words an improvement in linear-
ity. Feedback also reduces the ampliers output impedance, increas-
ing its damping factor. The frequency response of an amplier can
be made atter with proper feedback.
If care is not taken in the design of a feedback amplier, serious prob-
lems can occur. If the amplier has large amounts of phase shift at some
frequency, the feedback becomes positive rather than negative, and the
amplier will become unstable and will oscillate. Also, large amounts of
feedback can cause transient intermodulation distortion (TIM).
Neodymium A rare-earth metal element, atomic number 60, used in mak-
ing strong magnets. Such magnets are often used in dynamic micro-
phones and loudspeakers.
Neopilot A system for the synchronization of a motion picture camera with
the tape recorder recording the sound. The Neopilot system uses a clever
way to superimpose a 60-hertz signal generated by the camera or by a
crystal-controlled oscillator on the full-track tape in such a way that it
is not sensed by the normal full-track playback head, and so is not heard
with the recorded sound. This is done by using a special record head with
two narrow tracks near the center of the tape. One track records the 60-Hz
signal with one polarity, and the other track records the same signal with
inverted polarity. These two low-frequency signals cancel each other out
in the full-track playback head. The 60-Hz pilot signal is played back by a
narrow-track head that encompasses just the width of one of the recorded
60-Hz tracks. The recovered signal is used to control the speed of the tape
recorder in playback so the sound remains in synch with the picture. This
process is called resolving.
The Neopilot system was developed by Stefan Kudelski for his famous

257
Neutrodyne

Nagra portable tape recorders, and it revolutionized the recording of motion

picture sound. Before then, there were many other systems for synchro-
nization, most of them using large heavy studio-type recording equipment.
The advent of stereo recorders with room for a synch track on the tape obvi-
ated the need for Neopilot. See also crystal synch; smpte time code.
Neutrodyne A special circuit developed and marketed by the Hazeltine
Corporation starting about 1923 for decreasing the tendency for trf radio
receivers to oscillate. The Neutrodyne used a special trimmer capacitor
between the grid and plate of each tube in the rf section to neutralize
the interelectrode capacitance. It is said that the Neutrodyne was the
rst receiver that could be logged; that is, the same station would usu-
ally come in on different days at the same place on the dials! It made a great
sensation among radio buffs of the day. Radio Craft magazine, in its March
1938 issue, stated: If you didnt have a Neutrodyne, you were such a social
outcast that your very dog hid under the bed at your approach and looked
at you reprovingly, because the dogs of Neutrodyne owners wouldnt asso-
ciate with him.
NFB Short for negative feedback.
Nibble See nybble.
NICAM 728 Short for Near Instantaneous Companded Audio Multiplex.
This mouthful is a digital system designed by the bbc in the early 1980s
for the transmission of six different audio programs from London to the
transmitter sites in various parts of Great Britain. The sampling rate is 32
kHz with 14-bit coding, which corresponds to the European Broadcasting
Union (EBU) standard. The meaning of 728 is that 728 kilobits per second
comprise the raw data.
The NICAM system uses an interesting scheme for reducing the bit rate
of the transmission, called NICAM-3. The audio programs are converted
initially into 14-bit linearly coded samples. They are then compressed to
10 bits for transmission. At the receiving end, a digital expansion process
expands the signals in a way complementary to the way they were com-
pressed. This is thus a compander system operating entirely in the digi-
tal domain. As in all companders, the noise level varies with the signal
level, increasing as the signal level increases. Ideally, the noise is at a low
enough level that it is masked by the program, and the BBC has done very
careful listening tests to ensure that this is the case with NICAM-3. To help
in this, the signals are subjected to a pre-emphasis before being encoded
and a complementary de-emphasis after decoding at the receiving end.
Nickel-Cadmium (NiCd) Battery A common type of rechargeable bat-
tery, consisting of NiCd cells. They use cadmium oxide as an electrolyte
and nickel as the electrodes, and they are capable of many discharge-
recharge cycles if used properly. NiCd cells have a terminal voltage of 1.2
volts and retain this voltage well while they are being discharged. They
do not have as much capacity as a similar-sized alkaline cell but are
more economical in the long run if recharged many times.
Nicads are losing popularity due to better chemistry cells such as

258
 Noise Cancelling Microphone

Lithium-Ion (Li-Ion) and Nickel-Metal Hydride (NiMH) and the toxicity

of cadmium to the environment.
Nickel Metal Hydride Cell (NiMH) A common type of rechargeable cell
that has superior performance over NiCd cells. They have a much higher
capacity than NiCd, especially under high current draw, while retain-
ing a at voltage during discharge and no memory effect. They are pop-
ular for applications such as digital cameras, unless the even better but
expensive and demanding Lithium-Ion cells are used.
NIST, National Institute of Standards and Technology The successor to
the old National Bureau of Standards. The name change occurred in 1988.
NIST is responsible for the maintenance of all types of scientic standards
and units of measurement in the United States.
Noise Noise can be dened as any unwanted sound that is not related to
the wanted sound (if it is related, it is called distortion), and this de-
nition sufces for most uses; but in electronics, noise is specically dened
as a more or less wideband addition to a signal by any electronic or
mechanical component.
Additions to a signal that are integrally related frequency-wise to the
signal are generally called distortion of one type or another. The most com-
mon type of noise is sometimes called random noise because it is unpre-
dictable from moment to moment, and contains a continuous distribution
of energy spread over frequency. True random noise sounds like a hiss-
ing, and has no detectable pitch. An example of random noise is the sound
heard in fm receivers when tuned off-station.
Unfortunately, just about anything you do to a signal adds some noise
to it, and there is no such thing in nature as a noise-free signal. pream-
plifiers add noise, power amplifiers add noise, tape recorders add
noise, phono cartridges add noise, and loudspeakers add noise. In
copying an analog tape recording, for example, the copy will have at least
3 decibels more noise than the original, no matter how good the tape
recorder is.
An important measure of the quality of a signal is the signal-to-noise
ratio, measured in decibels. The signal-to-noise ratio provides informa-
tion on how much more powerful the signal is than the noise that inevitably
rides along with it. Signal-to-noise ratio is sometimes confused with
dynamic range, to which it is loosely related at best.
Ambrose Bierce denes noise as a stench in the ear. Undomesticated
music. The chief product and authenticating sign of civilization.
Noise Cancelling Headphone See active noise cancelling.
Noise Cancelling Microphone A specially designed dynamic micro-
phone that has both sides of the diaphragm exposed to the sound eld so
that sounds coming from a relatively large distance are cancelled because
the sound pressure causes no net force on the diaphragm. For sound orig-
inating very close to the microphone, such as when someone holds the micro-
phone up to the mouth when talking, the rear of the microphone is effectively
shaded from the sound eld, and the diaphragm does receive the sound

259
Noise Figure

pressure. Noise canceling microphones are used in high-noise areas where

clear communication is important, such as airplane cockpits, etc.
Noise Figure The ratio between the theoretical johnson noise, or ther-
mal noise of a device, and the actual measured noise is called the noise gure
and is expressed in decibels.
Noise Floor The noise oor of an audio device is the noise power generated
by the device in the absence of any input signal. It is generally measured
in decibels referenced to a particular power. Sometimes the noise oor is
measured in terms of rms voltage rather than power, and this makes sense
in the case of devices such as voltage ampliers or tape recorders.
In any case, the measurement is useless unless the bandwidth of the
measurement and the weighting network used (if any) are specied.
Noise Gate A noise-reduction device through which an audio signal is
passed. When the signal level is very small, the noise gate will close, elim-
inating any residual noise that may be riding on the signal. In the presence
of a signal, the noise gate will open, allowing signal and noise to pass
through. Under these conditions, the noise is masked by the signal. When
the signal is interrupted, the noise is audible (or more audible), having no
signal to serve as a masking source.
A noise gate is actually a special type of expander with an innite
expansion ratio below a preset threshold. The setting of this threshold is
adjustable, so as not to cut off too much of low-level signals.
Noise gates are sometimes used with electronically amplied musical
instruments such as guitars to reduce background noise when the instru-
ments are not being played. Their effectiveness is in large part determined
by the time constants associated with the gain reduction. Often, the
background noise can be heard switching on and off with the signal.
Noise Generator Sometimes it is desirable to use a random signal that
contains all frequencies at the same time as a test signal for audio equip-
ment. This signal is called white noise or pink noise, and is produced
with a noise generator. A good quality noise generator is not a trivial device.
It must have a very uniform output versus frequency and it must be sta-
ble in level over a long time.
One use of a noise generator is in the measurement of frequency
response of audio devices. The noise is input to the device, and the out-
put is measured with a spectrum analyzer, which averages the output
voltage for a time period. If the noise spectrum at the input is at, then
the spectrum of the output noise will be the frequency response of the device.
It is important to average the output over some time because the noise sig-
nal is random and is only at in frequency on a statistical basis.
Another use of noise as a test signal is in the measurement of the fre-
quency response of a room-loudspeaker combination. pink noise input
to the loudspeaker and a one-third octave band analyzer connected to a
measurement microphone are used to analyze the acoustic signal. The sys-
tem may be equalized to attain a specic frequency response. This is often
done in setting up sound reinforcement systems. See also equalization.

260
 Normalled Connection

Noise Reduction See compander; dynaural.

Noise Shaping In oversampling compact disc players, digital lter-
ing is used. By a special technique of inverting unneeded bits from one sam-
ple and adding to the next sample, the average quantization error for
slowly varying signals (i.e., low frequencies) will be reduced. This changes
the shape of the spectrum of the quantization noise, increasing its high-
frequency content while reducing its low-frequency content. The
increased high-frequency content is reduced by the anti-imaging filter.
Non-Fill When a record is pressed, soft vinyl plastic is pressed between
the two stampers, which are heated by steam. Sometimes, especially if
the heat is too little, the vinyl will fail to ow completely around the grooves,
and a large amount of noise is the result. Records with non-ll can be
identied visually, and are normally rejected by quality control at the press-
ing plant before being placed in jackets.
Nonlinear Literally, not on a line. In a nonlinear audio device, the output
voltage is not predictable from a knowledge of the input voltage for all
signal levels. For instance, a typical amplifier with a gain of 10 would
put out 20 volts with 2 volts input, and put out 30 volts with 3 volts input.
The implication is that 50 volts input would produce 500 volts output, but
this is impossible with most ampliers.
There is a point at which the amplier gain is reduced as the input level
is increased. This is the nonlinear region of the amplier. All devices have
local peculiarities or deviations from linearity at specic signal levels.
Nonlinearity produces certain types of distortion, such as harmonic
and intermodulation distortion. See also linear.
NoNoise The trade name of a single-pass noise reduction system made
by Sonic Solutions. NoNoise is a sophisticated digital signal processing sys-
tem that analyzes the digitized signal and senses transient noises, such as
clicks and pops, and continuous noises, such as tape hiss and mains hum.
It removes the transients and makes a substitute signal based on the actual
signal immediately before and after the transient. After analyzing the sig-
nal for a while, it can distinguish continuous noise from the varying sig-
nal and then digitally removes the noise.
Its primary use is in the restoration of old recordings such as analog tapes
and rare old records for release in CD form.
Normalize To boost the highest sample of a digital soundle to the maxi-
mum amplitude the system is capable of encoding, short of clipping
(0dBFS), and then raising all other samples by the same proportion. In effect,
raising the volume of the sound le so it is as loud as possible without going
into overload. This maximizes resolution and minimizes certain types of
noise. It is not, however, a substitute for compression, limiting, and other
dynamic range treatment to make a signal seem louder. Because of the
time involved in examining every sample, some software will use short-
cuts with varying amounts of accuracy.
Normalled Connection A connection, typically on a mixer or patch panel,
where the signal path is continuous into the signal chain with no plug

261
Norvalizing

inserted in the jack. Usually normalled connections are made via trs jacks.
The use of normalized patch panels reduces the number of patch cords one
has to have on hand if they represent the most often used connections. A
disadvantage of normalized patch bays is that it is not intuitively obvious
what the normalization connects. See also patch bay.
Norvalizing Hollywood slang for the act of playing a sound effect at a lower
level in a vain attempt to hide the fact that it is not in sync.21
NOS Acronym for Nederlands Omroep Stichting, which is the Netherlands
Broadcasting System. The NOS has developed a method for stereo record-
ing that uses two cardioid microphones placed 30 cm apart and angled
at 90 degrees from one another. The method is said to provide more ambi-
ence than coincident microphone techniques and fewer phase prob-
lems than widely spaced microphones. See also a-b stereo; ortf; m-s
stereo; and intensity stereo.
NOS also refers to New Old Stock, i.e., a replacement part that is old,
but unused and in new condition. Similar to NIB, New In Box.
Notch Filter A filter that rejects a narrow band of frequencies; also some-
times called a band-reject lter, although a notch lter usually has maxi-
mum attenuation at one frequency only.
Notch lters are used to remove specic frequencies, such as 60-hertz
line-induced hum from audio signals. If a notch lter is sufciently
sharply tuned, it will have minimal effect on the signal, other than remov-
ing the offending frequency. The maximum attenuation of a notch lter
can be quite large, typically 60 decibels or more.
One example of a notch lter is the whistle lter in some am tuners.
This lter is tuned to reject 10 kilohertz, which is the frequency that sep-
arates AM stations on adjacent channels. This puts an effective limit on the
frequency response of AM broadcasts at a little below 10 kHz.
When the fcc allocated frequencies for AM broadcasting many years
ago, they never anticipated the requirement for frequency response
above 5 kHz, and they spaced the allowed carrier frequencies 10 kHz apart.
Noy One noy is the noisiness of a noise for which the perceived noise level
is 40 PNdB (PN stands for perceived noise). The noisiness of a noise that
is judged by a subject to be n times that of a 1-noy noise is n noys.
This denition is from psychoacoustics, and practitioners of this dis-
cipline seem to be the only ones who use it, let alone understand it. The noy
is the result of an attempt, rst by psychologists and later by psycho-
acousticians, to establish a simple scale of perceived noisiness. See also mel.
NPN Short for Negative-Positive-Negative, which refers to a type of tran-
sistor that is constructed of three layers of semiconducting material. The
type of impurity doping in the semiconducting materials determines
whether the resulting semiconductor is P or N type. In a circuit, the

21. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

262
 NTSC

emitter is usually a negative voltage with respect to the collector. A small

positive base current results in a large negative collector current. See also
pnp; semiconductor.

NPN Transistor

NR Short for noise reduction.

NRC Acronym for National Research Council. This is an organization estab-
lished by the Canadian government for the purpose of establishing stan-
dards. The NRC is also engaged in wide-ranging research, which is
conducted at several well-equipped laboratories in Canada. A good deal
of basic research into psychoacoustics and sound reproduction has been
conducted at these laboratories.
NSA National Sound Archive, which contains a large collection of recorded
material, including over 750,000 records and 50,000 tapes, administered by
the British Library in London. The NSA is also actively involved in the
restoration and preservation of old recordings, and it, along with Cambridge
University, developed a computer-based system for sound restoration
called CEDAR (Computer Enhanced Digital Audio Restoration) that is
somewhat similar to the nonoise system. One of the goals of the NSA is
to nd a permanent storage medium for sound that will last at least 500
years without deterioration.
NSRC The National Stereophonic Radio Committee, which was set up by
the Electronics Industries Association (now the Electronic Industries
Alliance, eia) in the U.S. in 1958. The purpose of the NSRC was to study
the various proposed systems for fm stereo broadcasting and to make
recommendations to the fcc for standardization. It made its report in 1960,
recommending the Crosby system as capable of the highest quality results,
but the FCC chose an inferior system, in use to the present day, because it
had the capability of simulcasting.
NTSC The National Television Systems Committee (NTSC) is the body that
determined the American standards for the transmission of the original ana-
log television broadcast signals. Also known as Never The Same Color.

263
Null Test

Null Test A deceptively simple test for detecting distortion in power

amplifiers, wherein the input signal is subtracted from the attenuated out-
put signal. The difference is the distortion and would be zero in a perfect
amplier. The signal can be anything, such as noise, sine waves, or music.
The problem with this test is that it cannot distinguish relatively benign
linear errors such as very small frequency response errors and time
delays from nonlinear distortions that cause sonic degradation. Some lin-
ear errors, such as overall time delay, are inaudible, and very small frequency
response errors are also inaudible, but both show up strongly in the null
test.
Nuvistor A miniature type of metal vacuum tube introduced by RCA in
1960. Nuvistors were made in several types, including triodes and a
tetrode. The nuvistor was not much bigger than early transistors and
was capable of excellent performance. It is regarded as the highest devel-
opment of the vacuum tube.
NXT A special process developed by the Agfa company for restoration of
old analog audio tape recordings. When some magnetic tape ages, the binder
holding the magnetic oxide material in the tape coating undergoes chem-
ical changes that cause the coating to become soft and to slough off onto
the heads as the tape is played. The tape plays poorly because it sticks in
the mechanism, and playing also further damages the tape. The NXT process
is a treatment that stabilizes the coating and allows the tape to be played
without further damage. The process does not effect a permanent cure, and
the tape soon reverts to its previous state. For this reason, an NXT-treated
tape should be copied just after treatment. See also sticky shed.
Nybble One half of a byte, consisting of a group of four binary digits.
Presumably, it is spelled with a y because of the similar spelling of byte.
Nyquist Frequency One-half the sampling frequency in a digital sys-
tem. In such a system, the original signal must be sampled at a certain
rate per second before the individual levels can be converted into numer-
ical values for further processing. Obviously, the sampling rate must be high
enough to adequately encode the highest frequency of interest. It was
shown by Nyquist that a sampling rate of two times the maximum signal
frequency is just sufcient to describe that highest frequency without
ambiguity. This is called the Nyquist criterion.
Strictly speaking, the Nyquist criterion is correct, and two samples per
cycle are enough to describe a sine wave. However, in order to reconstruct
the original sine wave from so meager a quantity of samples, time aver-
aging must be used. In other words, you have to look at a sampled signal
for a certain time to decide exactly what the frequency and amplitude are;
there are ambiguities in both for short intervals of sampled waveforms.
The problem is, music is not a steady-state continuum of sound, but is
constantly changing in time, sometimes very rapidly indeed, and our ears
do not perform the time averaging required to reconstruct the unam-
biguous signal from the reconstructed output of a digital-to-analog
converter.

264
 OE, Operator Error

An analogous condition arises when a signal is present in random noise.

Time averaging can extract the signal, but the ear, having a short integra-
tion time, hears the noise as well as the signal. This phenomenon is some-
times called the Perman doctrine after Andr Perman, an engineer working
for Brel & Kjaer who studied the effect.
Nyquist Plot One format for plotting the frequency response function
and the impedance function, where the real part of the function is plot-
ted against the quadrature part. The Nyquist plot is useful for identi-
fying resonances in circuits or systems, and it is therefore sometimes used
to represent the impedance of a loudspeaker. See Appendix 11; heyser
spiral.

O
OBU Outside Broadcast Unit. A British term for a team of technicians
responsible for remote recording or broadcasting. Also known as on loca-
tion production.
Octave An octave is a frequency ratio of 2 to 1. An octave band consists
of all the frequencies within an octave. There is one octave between 100
hertz and 200 Hz, and also between 1,000 Hz and 2,000 Hz. Octaves are
perceived as equal pitch intervals, even though the true bandwidth in
hertz varies with the frequency level of the octave.
The name arises from the musical practice of dening the eight notes of
the scale within a doubling of the frequency. To our ears, two frequencies
an octave apart sound like the same note.
It is interesting that our ears obey a precisely logarithmic law when
assigning subjective pitches to frequencies. Even though an octave is
strictly speaking a subjective judgment, it is so closely equal to a frequency
doubling (no matter where one is in frequency) that it has been dened as
an objective measure. (With sine waves, or pure tones, the ear does not
have this precision of assigning subjective octaves to frequency doublings,
but pure tones are never found in nature.) filters having one octave of
bandwidth are called octave-band lters. See also Appendix 8; one-third
octave filter.
Octothorpe According to our friend Dennis Bohn of Rane Corp., the ofcial
name for the symbol # is octothorpe. The name was coined by the Bell Tele-
phone Laboratories when they added the # and the * symbols to telephone
keypads and the octo comes from the eight points the symbol has, and
Thorpe is the name of an engineer who was instrumental in getting the
# on the telephone keypad. Of course, in musical notation the # means
sharp and in common English # represents number or pound sign.
Octothorpe is only proper usage in regard to telephone keypads.
OE, Operator Error A failure in any mechanical or electronic system caused
by inappropriate action on the part of the humans setting up or operating

265
OED, Oxford English Dictionary

the system. Also known as cockpit trouble and a short between the head-
set.
OED, Oxford English Dictionary The unimpeachable source of several
denitions found herein.
Oersted The oersted is the unit of magnetic eld strength (abbreviated Oe)
and is named for Hans Christian Oersted, a Danish physicist born in 1777.
Oersted discovered the relationship between electric current and a mag-
netic eld by noting that a compass needle will be deected when near a
wire carrying a current.
Off Axis Not directly in front of a microphone or a loudspeaker.
Off-Axis Coloration A dull or colored effect on sound sources that are not
placed within the acceptance angle of the microphone. To avoid off-axis
coloration, the user places mics so that they are aimed at sound sources that
put out high frequencies, such as cymbals, when miking a large source. In
addition, one should use a microphone that has a flat frequency re-
sponse over the recording eld, i.e., has similar polar patterns at mid-range
and high frequencies. In general, large-diaphragm mics have more off-axis
coloration than smaller mics with 34-inch diaphragms or under.
Offset Angle In a record playing system, the angle that the cartridge
head shell makes with the tonearm is called the offset angle, and it is
designed to minimize the angular tracking error of the stylus in the record
groove. The optimum offset angle is dependent on the length of the ton-
earm, and the tonearm pivot must be the correct distance from the center
of the record as well.
Ohm The ohm (abbreviated , the Greek letter capital omega) is the unit
of electrical resistance; it is that which opposes an electric current in a
conductor. See also ohms law.
Ohms Law The mathematical relationship between electrical voltage,
current, and resistance was rst formulated by the German scientist
Georg Simon Ohm, and it is named after him. Ohms law says the current
in an electric conductor is directly proportional to the voltage across it and
inversely proportional to its resistance. In other words, the voltage and cur-
rent in a conductor exhibit a linear relationship.
Ohms law works for direct voltages and currents, and also is correct
for alternating currents if the resistance is a pure resistance. How-
ever, if the resistance has any reactive components (inductance or
capacitance), the current depends on the frequency as well as the volt-
age. See also impedance.
Most electrical conductors obey Ohms law very precisely for small cur-
rent levels, but some materials, most notably the semiconductors used
in transistors and diodes, have a much more complex nonlinear volt-
age versus current relationship.
An easy way to remember Ohms law is illustrated in the gure, where
V indicates voltage in volts, I indicates current in amperes, and R indi-
cates resistance in ohms. To nd the value of any quantity in terms of the
others, simply cover its symbol up with your nger. For instance voltage

266
 On Axis

is equal to current times resistance, IR, and current is equal to voltage

divided by resistance, V/R.

Ohms Law

OL Abbreviation for overload. Often an LED is used to indicate when a

circuit is close to or into overload, and will be marked OL.
OMFI, Open Media Framework Interchange A le format rst proposed
by Avid to allow for digital audio data interchange among digital dubbers,
editorial workstations, and hard disk editors.
Omnidirectional Literally, from all directions. Omnidirectional micro-
phones are equally sensitive to sounds coming from all directions. Para-
doxically, such a microphone is also called nondirectional.
For a microphone to be truly omnidirectional over the entire audible
range, it must necessarily be physically small, at most a half-inch or so in
diameter. When the wavelength of the sound being picked up becomes
small compared to the dimensions of the microphone, the microphone
becomes directional in its response, that is, more sensitive to sounds com-
ing from the front and less sensitive to sounds coming from the sides and
rear. Real microphones thus are always more directional at high frequen-
cies, which can be an advantage. When such a microphone is at some dis-
tance from a sound source, for instance while recording an orchestra in an
auditorium, the rising high-frequency characteristic helps to compensate
for the loss of high frequencies due to air absorption. Some microphones
with relatively large diaphragms use two diaphragms back-to-back to
decrease directionality at high frequencies, but they attain figure-8 direc-
tional patterns at high frequencies, which can be undesirable.
On Axis Directly in front of a loudspeaker, or in the direction of maxi-
mum sensitivity of a microphone.
By convention, frequency response curves of loudspeakers are mea-
sured on axis, usually with the measurement microphone at 1 meter from
the cabinet front. This generally gives the attest frequency response.
Microphones also usually have the most uniform frequency response when
used on axis.

267
One-over-f ( 1f ) Noise

One-over-f ( 1f ) Noise A type of electrical noise that occurs in semicon-

ductors and is inversely proportional to frequency. It has a burbling
sound and is caused by the fact that the current in semiconductors is not
perfectly smooth, as if the charges clump together.
One-third Octave Filter A bandpass filter whose bandwidth is almost
exactly 23% of its center frequency. Such a lter will have a bandwidth
that spans a musical interval of about a minor third, and this led to its name.
One-third octave lters have been used for many years for frequency analy-
sis of acoustical signals, and their shapes (i.e., their frequency response
curves) have been standardized by various standards organizations such
as the International Standards Organization (iso) and ansi. This is to ensure
that lters made by different manufacturers will give the same answer when
used to analyze the same sound. Their center frequencies have also been
standardized for the same reason.
Because the one-third octave lter has a bandwidth proportional to its
center frequency, it is a type of proportional bandpass lter. This means its
bandwidth becomes greater and greater as its center frequency becomes
higher and higher. In other words, its resolution decreases with rising fre-
quency. The reason such lters are useful for analyzing acoustical signals
is that this proportional response is similar to the response of the human
ear. We sense equal frequency ratios as equal musical intervals, so propor-
tional lters have a constant subjective, or musical, bandwidth as frequency
is varied. Moreover, it has been shown by many experiments that the so-
called critical bands within which the ear sums energy to determine
loudness are about one-third octave wide over much of the audible range.
For this reason, a frequency analysis need not have a greater resolution than
one-third octave in order to correlate quite well with our perceived loud-
ness of a sound. See also octave.
Many real-time analyzers consist of a bank of one-third octave
lters, and they may be analog or digital lters.
Opamp, or Op-Amp Short for operational amplier, a term originally
used for high-gain ampliers in analog computers. An operational
amplifier theoretically has a gain of negative innity, but practical designs
only approximate this ideal. If the amplier has innite gain, and proper
feedback is placed around it, the characteristics of the amplier will
depend only on the feedback and not the amplier.
In recent years, opamps have been produced in the form of integrated
circuits and with differential inputs. They are extensively used in
audio circuits. For an opamp to be good for audio use, it must have a wide
bandwidth, and must be very fast in its ability to change its output volt-
age. See also transient intermodulation distortion.
Opamps are convenient to use in circuits because they are available in
integrated circuit form and are relatively inexpensive. Many opamps are
available which have been optimized for audio use.
Open Abbreviation for open circuit, which denotes no electrical con-
nection between two devices or two components within a device.

268
 Operating Level

Some components, such as transistors and resistors, etc., are said

to be open if no current will pass through them. The open is a common
failure mode for circuit components and is generally less destructive to
the rest of the circuit than the short.
Open-Circuit Voltage Rating The output voltage of a microphone with
no load, i.e., with innite resistance such as in an open circuit, or when
driving a resistive load at least twenty times the microphones internal
impedance. One of the standard specications of microphones.
Open Loop An amplifier without feedback is said to be operating in
the open-loop mode, or to be an open-loop amplier. The feedback around
the amplier closes the loop.
The performance of an amplier with feedback depends a great deal on
the gain of the amplier without feedback, i.e., on the open-loop gain, and
this is something that is part of the specication of audio opamps.
Open Microphone A microphone that is connected to a broadcast or
recording system. A live microphone.
Operating Level In analog tape recorders, the operating level or refer-
ence level is a signal level near the maximum possible for the recorder but
low enough to ensure acceptably low distortion. The difference between
operating level and the maximum possible level is known as headroom,
and is measured in dB. The operating level is typically about 10 dB below
the level that causes 3% third harmonic distortion (or sometimes total har-
monic distortion or THD) at mid-frequencies (around 400 Hz to 1 kHz).
Operating level then gives approximately 10 dB of headroom before
severe distortion, but this is dependent on several factors such as tape speed
and frequency. Often, very high level signals are compressed by satura-
tion and have high distortion, but for brief signals, this may be acceptable.
In practice, it is difcult to record program material and precisely achieve
the maximum signal-to-noise ratio. For the original American tape
recorders of the 1950s, the operating level for analog tape recorders was
generally accepted by the industry as a level at which a tone of 400 hertz
would produce 1 percent third harmonic distortion (or THD), essentially
the same in this case, as the predominant distortion in a tape recorder is
3rd harmonic. Maximum level was considered 3% distortion, and the head-
room was 6 dB. Later, a uxivity of 185 nanowebers/meter on the tape
was established as the operating level, which was often referred to as
Ampex Operating Level after the originator of the American tape
recorder. This level was lower than the 1% distortion level, hence there
was more headroom, and a little less signal to noise ratio. Referencing a
uxivity was easier than measuring distortion. A test tape with a known
uxivity of 185 nWb/m could be played back and the tape recorder
adjusted so 0 dB showed on the vu meter. When recording, the operator
knew that signals near 0 dB were good for signal-to-noise and that 6 dB
of headroom was available for louder peaks.
Later improvements to analog tape led to increased operating levels, in
approximately 3 dB increments. Thus, a major improvement in tape in the

269
Operational Amplier

1960s, often called low noise/high output, could use a reference level of
250 nWb/m with the same distortion as the old tape, and enjoy a 3 dB
improvement in signal to noise ratio. This was called a +3 or elevated
operating level. More tape improvements led to +6 dB (370 nWb/m) and
even +9 dB (520 nWb/m) operating levels. If you had a 185 nWb/m test
tape and wanted to have a 250 nWb/m operating level, the recorder would
be adjusted so that the test tape would play at an indicated 3 dB on the
VU meter. Engineers were certainly free to trade signal to noise improve-
ment for less headroom and use a +6 reference level on a tape designed for
+3, or do the opposite and reduce signal to noise for more headroom. The
electronics in some tape recorders were not designed to handle elevated
levels and would produce unacceptable distortion. Some other recorders,
such as the Swiss Nagra, used special circuitry to cancel third harmonic
distortion to allow very high operating levels with good headroom and low
distortion.
VU meters are only marked 3 dB above 0, so some signals may still be
recorded with reasonable distortion even though the meters are pegged.
The skillful operator can juggle various factors such as the spectral content
of the signal, musical dynamics, tape speed, and recorder characteristics to
achieve a good balance between low distortion and low noise.
European methods for measuring uxivity differ from the American,
so the numbers will be slightly different. A standard European (IEC/DIN)
test tape at 320 nWb/m is about equal to the American (ansi) uxivity of
290 nWb/m at mid-frequencies, thus about 1.3 dB higher than an Amer-
ican 250 nWb/m tape. Research by John McKnight in 1998 indicated that
this difference was due to measurement errors and is not real.
Because there are so many variables, uxivities may vary a bit in
specications. Carefully recorded test tapes can be purchased with known
uxivities to enable adjustment of a recorders operating level. When
exchanging tapes with others, it is important to note the reference level used
along with the other tape data. With analog cassettes, operating levels are
not well standardized, although 0 dB levels of 145200 nWb/m are com-
mon. The lack of standardization is likely due to wide variations in cas-
sette tape and recorder performance and the desire to simplify operation
for consumers. See also reference level.
Operational Amplier See opamp.
Optical Sound Track The method for photographically recording sound
on lm for reproduction in the theater. Optical sound tracks were introduced
in theaters in the United States about 1930. See variable area, variable
density, and sva for details. With the introduction of digital sound in
motion pictures, Dolby developed a system of optical printing of the dig-
ital code beside the existing optical analog sound track. At about the same
time, DTS (Digital Theater Systems) was introduced and uses a relatively
simple code printed with the optical sound track to synchronize the lm
motion to the multitrack digital sound that is recorded on a CD-ROM to
be played in concert with the lm.

270
 ORTF

Optimum Source Impedance The optimum source impedance is that

source impedance connected to the input of an active device that results
in the minimum noise added to the signal. The OSI depends on the type
of device but is independent of the input impedance and independent
of the circuit conguration. field effect transistors have an OSI of
several megohms, whereas most bipolar transistors have an OSI of sev-
eral thousand ohms.
Opto-Isolator A semiconductor device that contains a light source,
usually in the form of an led, and a light-sensitive device such as a pho-
totransistor or light-dependent resistor. Current in the LED causes it to emit
light (usually infrared), which increases the conductivity of the light-sen-
sitive element. Opto-isolators are commonly used as replacements for
relays and switches and sometimes can replace mechanical attenuators.
They are also sometimes called opto-couplers. The advantage of the opto-
isolator is that its input circuit has no electrical connection to its output cir-
cuit, and such complications as ground loops can thus be easily avoided.
Orange Peel When records are pressed, the stampers are placed between
the platens of the record press, which applies a force of about 150 tons. If
the surface of the platens is not perfectly smooth, the irregularities in its
surface will be impressed in the record through the stamper, which is quite
thin metal. A regular pattern of tiny bumps is the usual effect, and it is called
orange peel because of the resemblance of the record surface to the skin of
an orange. Orange peel causes an increase in low-frequency noise.
Records with orange peel can be identied by looking at light reected from
the smooth surface between the inner grooves and the label. Any depar-
ture from a mirror-like smoothness is bad news.
The constant heating and cooling of the press platens causes the crys-
talline structure to roughen the surface, so they must be removed and
ground at periodically, which is an expensive operation.
Order In discussing filters, the number of poles a certain lter possesses
is called the order of the lter. Thus a T-section is a third-order lter and
an L-section is a second-order lter, etc. The slope, in decibels per octave,
of the lter response in its stopband is equal to 6 times the order.
ORTF Acronym for Ofce de RadiodiffusionTelevision Franaise, the
French national broadcasting system. The French have designed a micro-
phone conguration for stereophonic recording that is known as the
ORTF method.
The method calls for two cardioid microphones to be spaced 17 cen-
timeters apart and placed at an angle of 110 degrees. The microphones are
thus each 55 degrees from the centerline. The 17-cm separation is to simu-
late normal human ear spacing, and the 110 degrees is to simulate the direc-
tional pattern of the ears.
Recordings made in accordance with the ORTF method sound more open
and spacious than those produced by the X-Y method of recording, and are
quite well suited to headphone listening; but they do sound somewhat
dry and lacking in warmth due to the directional patterns of the cardioids,

271
Orthogonal

which pick up little ambient room sound. Because of the close spacing of
the microphones and the resultant similarity in phase, ORTF does provide
mono compatibility, that elusive quality broadcasters are forever seeking.
Orthogonal Two phenomena are said to be orthogonal to each other if they
can exist in the same medium at the same time and not interfere with one
another. The two motions of the stylus in a stereo record groove are an
example.
Orthophonic Victrola The Orthophonic Victrola was a new type of acoustic
phonograph introduced by the Victor Company in 1925. This was the year
that Victor introduced the rst electrically recorded records, and the new
player was sold as the ideal reproducer for them. The Orthophonic Victrola
had many design innovations, and surely was the best acoustic phonograph
ever produced. Its design was based on research conducted at Bell Tele-
phone Labs, and it contained the rst exponential horn, which was
folded in the lower part of the cabinet. The horn mouth was about 18 inches
by 30 inches, and the low-frequency reproduction from it was much better
than any player made earlier.
The stylus-diaphragm assembly was also very different from traditional
designs, having much greater exibility and introducing the spider to
couple the stylus lever to the diaphragm at several points instead of just at
the center. This also improved the bass response. The Orthophonic Victrola
used a spring-wound turntable motor, although many other phonographs
had electric drive motors by this time.
It is interesting that the rst electrical recordings sounded worse than
the current acoustic ones. In several cases, an acoustic and an electrical
recording were made of the same performance, and the listener could choose
(as is sometimes the case today, when we can choose between a stereo
record and a compact disc of the same performance). An example we
have heard is Paul Whitemans orchestra playing Gershwins Rhapsody in
Blue. The acoustic recording sounds less distorted and generally smoother
than its electrical counterpart.
The Orthophonic Victrola sold for $350 in 1925. It greatly outsold the
contemporary Brunswick Panatrope, which was the rst all-electrical
phonograph.
Oscar The name of the rst binaural microphone system, made in about
1932 by Harvey Fletcher and the other engineers at the Bell Telephone Lab-
oratories. The original Oscar was a tailors dummy head with microphones
where the ears would normally be, and it was used in experiments in bin-
aural transmission of music in the Academy of Music in Philadelphia.
Oscillator An electronic device that generates a periodic signal of a par-
ticular frequency, usually a sine wave, and sometimes a square wave
or other waveform. Oscillators are common in audio devices, and are also
extensively used as test signal generators.
Oscilloscope The oscilloscope (scope, for short) is a common instrument
that displays the instantaneous voltage of a signal versus time on the
face of a television-like screen. In other words, it displays the waveform

272
 Output Impedance

of a signal. It does so by means of a special device called a cathode ray

tube (CRT), in which a beam of electrons races toward the screen after being
deected in the vertical and horizontal directions by voltages applied to
deection plates in the tube. The electrons impinge on the phosphor coat-
ing on the inside of the screen, causing the phosphor to emit light and trace
out the path of the beam. A television picture tube is a cathode ray tube.
The oscilloscope is probably the most useful of the arsenal of test equip-
ment available today. It can be used to estimate the frequency and phase
of a signal, to see distortion in some cases, and to measure the instan-
taneous voltage of very rapidly changing signals. See crt.
Some of the newer designs of oscilloscopes use LCD screens (analogous
to screens used in most laptop computers) instead of CRTs. These scopes
use digital signal processing, and there is sometimes a noticeable time delay
between a change in the signal and the change on the screen. This delay
means they are not very good at displaying changing lissajous figures.
OSI See optimum source impedance.
Ossicles The three tiny bones (the smallest bones in the body) that trans-
mit vibration of the eardrum to the inner ear, or cochlea. They are some-
times called the hammer, the anvil, and the stirrup.
OTL Short for Output TransformerLess. OTL refers to a type of power
amplifier using vacuum tubes but having no output transformer. Usu-
ally, an OTL amplier will have several output tubes connected as cath-
ode followers and acting in parallel to attain a sufciently low
output impedance. In the case of ampliers designed for driving elec-
trostatic loudspeakers, it is relatively easy to design an amplier with
no output transformer because the loudspeaker itself has very high input
impedance, making it a natural to be driven by tubes.
transistor ampliers, of course, do not need or use output trans-
formers because of their inherent low impedance, and the term OTL is
reserved for tube ampliers.
Outboard Any audio equipment that is not contained within another
device. Separate effects devices such as reverbs and equalizers that are
apart from the mixing console would be one example, whereas the equal-
izers built into the mixer are not considered outboard.
Out of Phase A condition where two signals have a phase difference. If the
difference is 180 degrees, or one-half cycle, it should be called out of po-
larity, phase being a continuous variable rather than discrete.
Output Impedance The output impedance of a device is the actual imped-
ance at the output terminals. A preamplifier with 600 ohms output
impedance means the output signal appears to be in series with a 600-ohm
resistor. Thus, if a 600-ohm resistor is connected across the output, the
signal voltage will be cut in half; one-half appearing across the load resis-
tor, and one-half appearing across the internal impedance.
The internal resistance of a battery also can be considered to be an out-
put impedance, with the battery voltage in series with it. Under no load
(i.e., no current being drawn), the voltage will be at its maximum value,

273
Output Transformer

but it will be reduced under load because the current causes a voltage drop
across the resistance in accordance with ohms law. The same thing hap-
pens with an audio amplifierthe lower the output impedance, the less
the output voltage will vary with load.
A power amplifier may have a rated impedance of 4 or 8 ohms, but
this does not mean the true output impedance is that value. The rated imped-
ance is simply that impedance into which the amplier will deliver its great-
est power. The amplier is capable of a certain maximum voltage output,
determined by the internal power supply. It also has a certain maximum
current output, usually determined by the output transistors. This
maximum voltage at the maximum current can only be delivered to a
specic impedance, according to Ohms law, and this is the rated impedance
of the amplier. See also damping factor.
Output Transformer A type of audio transformer mostly used in tube-
type power amplifiers that couples the output tube plates to the
loudspeaker load. The output transformer in such a power amplier is
a very important and critical part of the amplier, and much effort has been
expended over the years to perfect its design.
Outro The opposite of intro, used in popular music to designate the end-
ing of the piece.
Outtake In a recording session where several takes of each selection are
recorded, the ones not used in the nal master tape are called outtakes.
Overbias The use of more bias current in an analog magnetic recorder
than is required for maximum sensitivity. Normally, the bias is adjusted
so that a test tone will be recorded at maximum level as read on the out-
put vu meter. If the bias is increased above this value, the recorded level
will go down. If the gain reduction amounts to, say, 2 dB, the recorder is
said to have 2 dB of overbias. The overbias will reduce the distortion
and the sensitivity to dropouts but will also reduce the high-frequency
response, so the record equalization control must be adjusted to com-
pensate. The frequency of the test tone depends on the tape speed and the
record head characteristics and typically is 5 kHz at 712 ips tape speed and
10 kHz at 15 ips.
The proper adjustment of the bias in a tape recorder is a compromise
because bias affects the noise, distortion, frequency response, and dropout
sensitivity in a complex way.
Overdrive To input a signal to an audio device in such magnitude that
an overload occurs is to overdrive it.
Overdub In analog tape recording, if a tape is copied from one machine to
another and if another signal is added to the copy at the same time, then
the copy is called an overdub, and the process is called overdubbing, or
sweetening.
Successive overdubbing can create the effect of a great many different
musical lines all playing simultaneously. One disadvantage of overdubbing
is that it involves repeated copying of the various tracks as new ones are
added, and this adds noise at the rate of 3 decibels per copy. A better,

274
 Oversampling

although more expensive, way to achieve the same effect is to use a multiple-
track tape recorder, and record the tracks one at a time, each time while lis-
tening to the previous tracks. This is called sel-sync in Ampex parlance,
although it is still informally referred to as overdubbing.
Overdubbing can also be performed in the digital domain in an editing
work station or DAW, and this does not add the 3 dB of noise with each
successive addition of another track. The repeated digital mixing and pro-
cessing does add some distortion, although probably not very much.
Overhung Coil In a dynamic loudspeaker, the voice coil can be made
to be longer than the magnetic gap in which it resides. It is then called an
overhung coil, because the ends of the coil overhang the gap. The raison
dtre of the overhung coil is to decrease the nonlinear distortion of the
loudspeaker for a given power output. This is accomplished because the
coil, as it moves in and out of the gap, maintains a constant number of turns
of wire in the magnetic eld, so the force on the cone is not dependent on
the position of the cone.
There is a trade-off, however, for the turns of the coil that are not in the
gap do not result in any force on the cone, and they simply add resist-
ance to the voice coil impedance, which wastes amplier power. This
reduces the efficiency of the speaker, but many designers, especially of
small bookshelf systems, believe the lower distortion is more valuable
than high efciency. See also underhung coil.
Overload An overload is said to occur when the input signal level in an
audio device is so large that it drives the device out of its linear range
and into distortion or clipping. Overload may be continuous or may
occur only on short peaks in musical waveforms. The latter condition is
common with certain waveforms, such as sharp percussive sounds that have
a peak value much greater than their average value. This peak clipping,
as it is called, must be avoided for true high-fidelity recording and repro-
duction, although a small amount of it may be quite difcult to hear in
practice.
Overmodulation A situation that occurs when the amplitude of a sig-
nal exceeds the limits of the broadcasting transmitter system. Broadcast
stations almost always have limiters in the signal path to prevent over-
modulation at the transmitter. The opposite of undermodulation.
The term only actually applies when true modulation of a carrier is
involved; sometimes it is also used to mean overloading or clipping,
which is not correct usage.
Overs The overloading of peaks in a digital audio signal, producing clip-
ping of the signal and, if severe, bursts of noise. Some software will
attempt to count overs, and some CD replicators will reject masters that
have excessive overs. In any event, an over in a digital system means dis-
tortion, whether plainly audible or not.
Oversampling In reading or copying digital audio data, the signal may be
sampled at a higher sampling frequency than was used in generating the
digital signal in the rst place. For instance, in some compact disc play-

275
Overshoot

ers, the nominal 44.1-kilohertz sampling frequency is raised to 176.4

kHz, or a factor of 4 higher. This means three articial samples are created
in between each pair of original samples. These samples are zero in level,
and they do not change the information content of the original samples.
digital ltering is then used for interpolation of the zero samples to val-
ues intermediate between the true 44.1-kHz sampled values. But because
the sampling rate is now so much higher, a very gentle anti-aliasing
filter can be used rather than the brickwall filter usually needed,
resulting in much less phase distortion of the signal. See also sampling.
Oversampling can occur at different multiples of the original sampling fre-
quency, such as 8X or more.
Overshoot If a compressor or limiter is subjected to a sudden large input
signal level, its gain reduction circuitry may not be fast enough to prevent
the output from being momentarily too high. This initial excessive level is
called overshoot, and its severity depends on the speed of the device.
Also, imperfect transient response in an audio device will result in the
waveform going past the desired value on fast signal transitions. This is
also called overshoot and is frequently seen on square waves and impulse
responses of CD players.
Overtones Overtones are tones produced by a musical instrument that are
higher in frequency than the fundamental. They may or may not coin-
cide with the frequencies of a harmonic series.
All musical instruments produce complex sound waveforms that
repeat at their fundamental frequency. Any complex periodic waveform
such as this can be expressed as a collection of sine waves of frequencies
of integral multiples of the fundamental, or lowest frequency. These sinu-
soidal components are called harmonics, or a harmonic series, the rst
harmonic being the fundamental. Generally, the harmonics excluding the
fundamental are called overtones, the rst overtone being the second har-
monic, etc. Overtones are sometimes called partials.
Oxide The magnetic material in almost all analog magnetic tapes consists
of the various oxides of iron and sometimes chromium. The business side
of the tape is called the oxide side. Debris that rubs off the tape and col-
lects on the tape heads and guides is mostly oxide and binder, and the
phenomenon is called oxide shed. Digital audio magnetic recorders usu-
ally use metal particle tapes that do not use oxides, but metallic particles
encapsulated in a material to prevent their oxidation.

P
The Greek letter phi, pronounced fee in Greek and usually fye in
English. It is often used as an abbreviation for phase. Some control con-
soles have a small switch on their input modules labeled that inverts the
polarity of the signal in that channel. This is to allow all the signals being

276
 Panpot

mixed together to have the same polarity regardless of wiring errors in

patch panels and microphone cables and in the microphones themselves.
See also phase invert.
The letter is also the physical symbol for magnetic ux.
PA Short for Public Address. An archaic term for a sound system used in
auditoriums and other venues where people gather to listen to music or a
speech.
Pack See tape pack.
Pad A short name for attenuator, usually with a xed amount of insertion
loss. Pads are used between audio devices if there is a danger of the out-
put of one device causing a signal overload in the input of the other device.
Some microphones have very high output voltages and will overdrive
some preamplifiers, so a pad of 10 to 20 decibels is inserted in the micro-
phone line.
PAL, Phase Alternating Line The terrestrial analog television broad-
casting system used in most of Europe, (except France, which uses secam)
is called PAL. The system uses 625 horizontal lines with interlaced scan-
ning at the rate of 50 elds per second. The phase of the color portion of
the video signal is reversed for each alternate line, so in one frame the alter-
nating lines have alternating polarity. This improves the color quality
because the reverse polarity compensates for unavoidable phase errors in
the adjacent line caused by the transmission chain. PAL receivers, unlike
those of the ntsc system, do not have tint controls because the colors are
always true due to the color signal phase correction.
Pan To position the stereo image of an instrument or vocal across the stereo
image. See also panpot.
Pancake A pancake is a 1012-inch reel of recording tape without the reel
flanges. In other words, it is simply the tape wound onto a hub with no
side supports. Recording studios often bought tape in this form and spooled
it onto smaller empty reels to use. Spooling the tape off onto another reel is
one thing, but it is quite another to try to spool it back onto the pancake.
This usually results in a tangled disaster. One solution is to keep a pair of
loose anges around to sandwich the tape in when spooling the pancake.
P & Q Subcodes See subcodes.
Panotrope The Panotrope was the rst phonograph that used electrical
rather than mechanical means to extract the signal from the groove of the
record. It had a magnetic cartridge, an amplifier using several tubes,
and a loudspeaker. It was made by the Brunswick company, introduced
in about 1925, and sold for something like $650.
Panpot Short for panoramic potentiometer, which is two connected vol-
ume controls with a common knob, so wired that as one is turned up,
the other is turned down. If the two stereo channels are controlled by
a panpot, the apparent position of the sound will move from left to right
as the control is turned. The balance control on most stereo amplifiers
is actually a simple panpot.
Panpots are used in recording to place the apparent position of a sound,

277
Parabolic Microphone

such as a soloist or other instrument, anywhere between the two loud-

speakers. Its operation relies on the ability of our ears to localize a sound
by level differences heard by our two ears. See also binaural. For a pan-
pot to work properly, it must follow an accurately prescribed attenua-
tion curve. In many recordings, several instruments are given separate
positions by using a panpot on each one when the nal mix is made.
Assigning instrument locations by panpot is somewhat poor at best,
because it assumes that the listener will be midway between the speakers.
Otherwise, intensity cues do not work very well, and the listener tends to
localize the sound as coming from one speaker or the other, rather than being
spread between them.
There exist multichannel panpots for use in motion picture sound tracks,
where a sound may be panned across three channels as the source moves
across the screen or from side to side using the surround speakers.
If a panpot were made that supplied small time and spectral differences
between the channels as well as simple volume differences, it would be
much more effective, although it would be complicated and expensive to
implement.
Parabolic Microphone A microphone system consisting of a reecting
surface in the shape of a paraboloid and an otherwise conventional car-
dioid microphone mounted at its focus. Such microphones are extremely
directional at high frequencies, making them well suited for picking up bird
songs outdoors. The reector is usually made of clear plastic, and may be
from 18 to 36 inches in diameter. Parabolic microphones are often used by
television crews at sporting events in order to capture dialogue (or mono-
logue) on the eld.
Parabolic microphones have somewhat limited frequency response,
and are not used where high-quality sound is required. However, they
are much more directional and sensitive on-axis than shotgun
microphones.

Parabolic Microphone

278
 Partials

Paragraphic Equalizer A form of parametric equalizer that uses

slider controls similar to graphic equalizer controls rather than rotary
knobs.
Parallel Electric circuit elements are said to be connected in parallel if
the current divides at the connection point and part of it goes to each
element. This is in contrast to a series connection, in which all the current
in each element is the same.
In a parallel circuit, the voltage across each element is always the same,
but the current in each element depends on the individual impedance of
the element. As more elements are connected in parallel, the combined
impedance is reduced. For this reason it can be difcult to drive many loud-
speakers in parallel from a single amplifier; the combined impedance
can become so low that the amplier cannot supply sufcient current to
drive them.
In most active audio devices, the input impedance is relatively high
and the output impedance relatively low so several devices can be con-
nected or paralleled to the output of a single device with no problems.
Parameter See parametric equalizer.
Parametric Equalizer A parametric equalizer has several parameters
that can be varied. Typically there are several resonant circuits in such a
unit, and they can accentuate or reduce selected frequency bands. They
can be adjusted to be quite selective in frequency, operating in a very nar-
row band, or the response can be widened to cover a much wider band.
So-called sweepable parametric equalizers are able to move the center
frequencies of the equalization bands. Also sometimes called a peaking
equalizer. See also equalizer.
Parasitic Oscillation A malfunction occurring in some audio devices, espe-
cially power amplifiers, in which the device will generate an ultra-high-
frequency signal during a part of the audio signal. The oscillation is present
only when the signal is present and occurs only during part of the audio
waveform. Parasitic oscillation can burn out tweeters. Although the
ultrasonic component is not directly audible, its existence causes audi-
ble distortion in the signal.
Partials The frequency components that make up musical sounds, also
called overtones or harmonics, although these terms are more specic.
Harmonics are dened as integral multiples of a lowest, or fundamen-
tal, frequency of a sound, and they are precisely in tune with each other
and the fundamental. Many musical instruments closely produce true har-
monics, but many others, such as the piano, do not. Harmonics are num-
bered in succession from the fundamental, which is the rst harmonic.
The term partial is dened as any component of a complex tone, whether
it is in tune with the fundamental or not. Partials may be lower in frequency
than the fundamental; then they are sometimes called subharmonics. Over-
tones are dened as the harmonics above the fundamental, but in common
usage they are taken to mean any partials above the fundamental.
Certain instruments, such as bells and some drums, produce a collection

279
Pascal

of partials with little resemblance to a harmonic series. In the case of most

bells, there is a strong partial at one-half the frequency of the perceived pitch
of the bell, and it is called the hum tone. In many other percussive instru-
ments there are strong out-of-tune partials that can be heard as distinct pitches
bearing little or no relation to the perceived pitch of the instrument.
Pascal The si unit of pressure, abbreviated Pa, equal to 1 newton per square
meter. The name comes from the French scientist Blaise Pascal. The refer-
ence pressure equivalent to 0 dB sound pressure level is 20 micropascals.
Passband The passband of a filter is the frequency span that the lter
passes, or the range of frequencies not attenuated by the lter. The pass-
band is usually measured between the points where the response is 3 deci-
bels down in amplitude relative to the maximum level.
Passive A device is called passive if it contains no amplication circuitry
and a signal suffers a loss in power or level in passing through it. This loss
is called insertion loss. Some passive devices, such as transformers, can
increase the voltage or the current of a signal, but they all result in a power
loss.
Many audio equalizers used to be passive, but most are active now.
crossover networks inside loudspeaker cabinets are passive.
Passive devices in general do not add any appreciable nonlinear dis-
tortion or noise to the signal, but they may have so much insertion loss that
additional amplication is needed, and this always contributes some noise
and distortion. Passive devices can and do cause phase distortion, how-
ever. There are many passive equalizer devotees among recording engineers
who claim they sound better (or at least different from active ones) and they
are often traded at high prices.
Passive Radiator A passive radiator is a device resembling a low-frequency
loudspeaker (woofer), but it has no voice coil or magnet assembly.
It is a cone with a exible suspension. It is placed in some loudspeaker
cabinets as a replacement for a port or ducted port. Such a system oper-
ates something like a bass reflex system, but the extra mass of the pas-
sive cone reduces the frequency of the resonance, allowing low frequency
performance with a smaller cabinet than a true bass reex would need. The
passive radiator, however, has some other problems. The mass of the cone
is relatively great and the resulting resonance is difcult to damp, and this
can lead to bass hangover and a boomy sound.
PA System See public address system.
Patch Bay Generally, a group of connectors on a panel where patch cords
are interconnected in various congurations to route audio signals. The most
common design consists of modules of two rows of twenty-four 14-inch jacks
in a 1U rack panel. This design dates from early telephone company use.
As many modules as needed are added to form the full patch bay, although
just one such module can be called a patch bay. Traditionally, outputs are
wired to the top jacks, and inputs wired to the lower jacks. In order to avoid
having to use patch cords at every jack, the input/output pairs that are nor-
mally used as-is are prewired such that the signal is already connected with-

280
 PCM

out the need for a patch cord. However, inserting a plug into the upper jack
disconnects the prewired connection through a switching mechanism,
sending the upper jack signal only to the inserted cord. Similarly, inserting
a plug into the lower jack would disconnect the prewired connection, and
allow only the inserted cord to be connected to that lower jack. This is called
a normalled pair of jacks. Another conguration of jack pairs is called half-
normalled. In this case, inserting a plug into the top jack does not discon-
nect the normalling, so that the output can be routed to 2 placesthe normal
jack below, and the inserted patch cord. However, inserting a plug into the
lower jack does disconnect the normalling. Usually there are also provi-
sions to have no normalling at all between jacks, and some designs allow
easy changes between congurations. Although most patch bays present
1 -inch phone jacks on the front, the rear connections may be 1 -inch phone
4 4
jacks, RCA jacks, punchdown terminals, or solder lugs. Many large mix-
ing consoles use a smaller jack and patch cord known as the Tini-Tel. Tele-
vision signal patch bays may look similar, but use a special high-frequency
connector. See also patch cord.
Patch Cord A relatively short audio cable with connectors on each end is
commonly called a patch cord. It is used for interconnecting various devices
where the conguration must be changed from time to time. Recording stu-
dios make good use of patch cords.
The most common type of patch cord is probably the one using quarter-
inch phone plugs on each end. This connector was rst used by the tele-
phone exchanges, hence the name. The telephone patch cord contains three
conductors: two for the signal and a shield. The connector also has three
contacts, called tip, ring, and sleeve (thus, called trs). The tip is always the
hot side of the signal, the ring is the low side, or ground, and the sleeve
is the shield. There also exist phone plugs that have only two conductors:
tip and sleeve. They are common in nonprofessional audio equipment.
microphone signals should not be patched with phone plugs because
the shield, or ground connection, is the last one to make contact, and loud
pops and 60-Hz hum can result if a live microphone preamp input is
patched with a TRS plug; rather, three-pin XLR-type plugs and jacks
should be used. Pin number 1 in the XLR plug is always connected to the
shield. The reason is that the connectors are so designed that pin 1 makes
contact rst, ensuring that the ground connection is made before the sig-
nal connection. This greatly reduces the transient thumps and pops that
occur when a circuit is patched with the power turned on. A group of
similar receptacles (or jacks) in an audio system is called a patch bay,
and the act of plugging and unplugging the cords is called patching. Record-
ing consoles used in studios make extensive use of patch bays for ease of
rerouting signals, especially microphone signals. See phone plug; trs; xlr.
Patch Panel Another name for patch bay.
PC Board See printed circuit board.
PCM Pulse Code Modulation (PCM) refers to any type of digital encod-
ing and decoding of a signal. Some types of PCM are:

281
PD

1. linear PCM, where the voltage range of the analog signal to be

digitized is divided into a certain number of equal intervals. All the quan-
tization steps are the same size in such a system. This is the most com-
mon PCM for audio applications.
2. nonlinear PCM, wherein a compander is inserted before a linear
PCM encoder. This will give the system better resolution at low signal lev-
els at the expense of resolution at high signal levels. This system is desir-
able when the number of bits for encoding is limited, and is widely used
in telecommunications work.
3. Floating-point PCM. The oating-point converter produces an out-
put digital word consisting of two parts: an exponent and a mantissa. The
exponent represents the gain in a variable-gain amplifier that precedes
a linear PCM encoder, which produces the mantissa. The gain is thus
adjusted stepwise in order to best encode the variable-level input signal,
and higher resolution is attained.
4. Differential PCM. A form of linear PCM where the difference between
the present sample value and the previous sample value is encoded, rather
than encoding the actual value of each sample. This has the advantage of
requiring fewer bits than does straight linear PCM. If the encoding is done
with only one bit, this special case is called delta-sigma modulation,
originally called Sigma-Delta by its inventors.
Probably the most used schemes are linear PCM and Delta-Sigma dif-
ferential PCM. However, when audio people speak of PCM, they invari-
ably are referring to linear PCM, and Delta-Sigma is invariably explicitly
called that.
PD See prodigi.
Peak Peak value is the maximum instantaneous excursion from zero of an
audio waveform. The peak value of a sound is also the maximum instan-
taneous pressure excursion of the sound.
Usually, when measuring peak values of signals, a relatively short time
interval is considered. For instance, the peak value of a recording of Beetho-
vens Ninth Symphony will, strictly speaking, be represented by a single
value somewhere in the piece, but the peak value of the waveform of the
symphony (determined over short time intervals of maybe a second or so)
will vary with the dynamic level of the music at that particular time. See
also rms and crest factor.

Peak Value of a Waveform

282
 Penthouse

Peak Expansion A volume expander can be adjusted so the expansion

only occurs when the input signal is above a certain loudness level. This
results in the expansion of the peaks in the program and does not affect the
rest of the signal. Peak expansion can add naturalism to music that has
been overly compressed.
Peak Hold A function of some volume indicators that indicates the peak
level of the signal and holds that level until it is either exceeded by a higher
peak or the indicator is reset by a time delay or manual reset button. See
also ppm.
Peak Level Meter Same as peak program meter (PPM) and peak program
indicator (PPI). See ppm.
Peak Limiter See limiter.
PEAQ, Perceptual Evaluation of Audio Quality PEAQ is an objective
method for determining the perceived quality of wide-bandwidth audio. Its
raison dtre is to eliminate the need for subjective evaluation of different
types of audio signal processing and reproduction systems. Subjective eval-
uation of audio systems is time consuming, difcult, and expensive, because
it uses a panel of human subjects for extended listening tests and relies on
the subjects to explain what they heard using various statistical methods. Sub-
jective evaluation tests are reliable when properly conducted, but they are
not used very often due to their difculty and expense. PEAQ was stan-
dardized by the International Telecommunications Union (itu) in the doc-
ument published as BS-1387. The standard was a joint effort by many
workers in laboratories in Canada, the Netherlands, France, and Germany.
The standard describes in detail how to objectively measure the quality of
audio processing systems such as different codecs associated with differ-
ent sampling rates. Two methods are described in the standard, a basic and
an advanced version. The advanced version is said to be more exactly
repeatable in its results.
PEC/Direct In lm re-recording, the act of switching between playback from
the recorder (either off the play or record heads) and the console bus. PEC
stands for photoelectric cell and originates from when monitoring off opti-
cal photoelectric cell was as close as you could get to playback.22
Pegged, Pegging When a strong signal connected to a meter, such as a vu
meter on a tape recorder, causes the meter to reach its maximum possible
level, the meter is said to be pegged. The moving needle in the meter
contacts a mechanical stop called the peg, hence the name. Pegging a VU
meter is not a good idea because the needle may be bent or even broken
by severe pegging. Of course, solid-state level indicators do not have
moving needles, but they are still said to be pegged when overloaded.
Penthouse A metal housing containing the magnetic reproduce head
stack for pickup of the sound from 35-mm and 70-mm motion picture lm
that has been sound-striped. It is called a penthouse because it sits atop

22. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

283
Pentode

the movie projector, with the lm going through it before entering the pro-
jector itself. Most digital readers now reside in the penthouse. There are a
few readers called basement readers that are retrots in the optical
sound head.
Pentode A vacuum tube containing ve active elements: the cathode,
control grid, screen grid, suppressor grid, and plate. The pentode has high
gain and high efciency and is used as a voltage amplier stage in audio
ampliers. It has greater distortion than the triode, however.
The name comes from the Greek hodos, meaning path, and penta mean-
ing ve. Early pentodes were called pentahodos.
%AlCons Percentage Articulation Loss of Consonants, a gure of merit for
speech intelligibility in a room. A value of 10% loss or less is typically con-
sidered acceptable. In the 1950s, V. M. A. Peutz of the Netherlands conducted
research using taped spoken words and human listeners, and determined
that articulation loss of spoken vowel sounds did not correlate to intelligi-
bility like consonants did. He determined a formula that correlated the
%Alcons to measured room characteristics. Instruments such as the TEF
analyzer can also calculate the %Alcons. %Alcons is popular in the U.S.,
but has shortcomings such as being limited to the 2-kHz band if the TEF
analyzer is used. See also rasti.
Perceptual Coding The digital audio coding scheme used on compact
discs (CDs) yields an amount of data often too immense to store or trans-
mit economically, especially when multiple channels are required. As a
result, new forms of digital audio coding, often known as perceptual cod-
ing, have been developed to allow the use of lower data rates with a min-
imum of perceived degradation of sound quality. Perceptual coding
schemes are lossy, meaning the decoded signal is not identical to the orig-
inal signal that was encoded because some of the data is discarded. But the
losses are carefully designed so that they are essentially not audible. To per-
form this trick, the coder analyzes the input music signal and decides not
to encode certain parts that would be inaudible due to masking in the ear,
where loud sounds drown out softer sounds that are also present in the
music. In a sense, a perceptual coder fools the ear into thinking it is hear-
ing the entire music signal in its original form, and some are better than
others at doing the job. Of course the acuity of the ear listening to the result
varies also, adding to the complication. Dolbys audio-coding algorithm
AC-3 is such a coder. See also codec.
Perfect Pitch See absolute pitch.
Perfs Slang for perforations in motion picture lm, also called sprocket
holes. Dolby Digital sound data on 35-mm lm is located between the perfs.
Period In a waveform that repeats a particular pattern over and over, the
time required for one repetition is called the period. The waveform is then
called periodic, and can be expressed as the summation of a series of sine
waves called harmonics. See also fourier analysis.
Periodic When the waveform of a signal repeats the same shape over
and over, the waveform is said to be periodic, and the period is the time

284
 Perspecta Sound

Periodic Waveform

required for one repetition. The repetition rate is called the fundamen-
tal frequency, and is equal to the reciprocal of the period, or F = 1 T. If
the period varies from repetition to repetition but still has the same gen-
eral form, the signal is quasi-periodic. If there is no discernable repetition
period, the signal is aperiodic. white noise and pink noise are exam-
ples of aperiodic signals.
Most musical instruments produce periodic, or nearly periodic, sounds,
but some do not. Many percussion instruments produce sounds resembling
random noise. See also fundamental.
Permalloy Permalloy is a special metallic alloy that has a very high mag-
netic permeability. In other words, it is easy to establish a magnetic eld
in it. It is used in the pole pieces of analog tape heads and for the cores
of some audio transformers.
Permeability The relative ease with which a magnetic eld can be estab-
lished in a medium is the mediums magnetic permeability. Of the com-
mon metals, iron has a high permeability; mumetal is a special alloy with
very high permeability. Magnetic tape heads are designed with high-
permeability cores to increase their sensitivity, and magnetic shields
are designed to have high permeability. The units of permeability are hen-
rys per meter (H/m), and its symbol is the Greek letter (mu).
Perspecta Sound A type of single-channel magnetic sound recording sys-
tem for motion pictures that used three subaudible tones (30, 35, and 40
hertz) mixed with the audio signal as control tones for varying the volume
and placement of the sound image reproduced through three loud-
speaker systems behind the screen. The tones were separated by band-
pass filters and fed into the control inputs of variable gain ampliers.
You might well ask why the tones were considered subaudible; the reason
was that the theater loudspeakers of the day were not able to reproduce
such low frequencies at audible levels.
Perspecta sound was introduced in 1954 by the respected recording engi-
neer Robert Fine, and was essentially an updated hi- magnetic version of
the old vitasound optical system of the 1930s. Like Vitasound, it was com-
patible in that a standard projection system with no control circuitry would

285
PFL

simply reproduce the sound in a conventional way. Even though Perspecta

sound was a monaural system, it was said to produce pleasing and con-
vincing audible effects in the theater.
PFL, Pre Fader Listen On a mixing console, the PFL button on an input allows
monitoring of that input channel regardless of the input fader position.
PFX, Production Sound Effects Sound effects recorded on the production
sound track during lming. PFX are kept separate during dialog editing
and mixed with the music and Foley effects to create a complete sound track
without the dialog. This track can then be mixed with any number of for-
eign language dialog tracks to create nished foreign language versions of
the picture.
Phantom Center Channel A third loudspeaker placed between the two
conventional loudspeakers in a stereo system to provide simulated three-
channel stereo reproduction. The signal fed to the center channel is simply
the sum of the two stereo channels. This technique was investigated at some
length by Paul Klipsch (inventor of the famous klipschorn) in the late
1950s. The phantom channel does tend to reduce the effect of the hole in
the middle but is not equivalent to true three-channel reproduction. The
phantom signal can be derived simply by summing the two stereo chan-
nels with resistors and feeding this sum to a third amplier and then to the
speaker. If the power ampliers have output transformers, as all tube-type
ampliers do, the phantom signal can be obtained by suitable connections
to the transformer secondaries, avoiding the need for the extra amplier.
Phantom Image A monophonic sound panned equally to both left and right
stereo speakers. See also phantom center channel.
Phantom Power condenser microphones require a preamplifier to
be close by due to the extremely high impedance of the microphone itself.
This preamplier is in the housing of the microphone, and it needs a power
source. Sometimes a battery is used, but more often a multiwire cable brings
the audio signal from the microphone and brings the power from an exter-
nal power supply to the preamp. This is a rather bulky and expensive
arrangement. To eliminate the multiconductor cable, frequently a scheme
called phantom powering is used, whereby the preamp power is carried
by the same two wires that carry the signal. The key to its operation is
the fact that the signal is alternating current and the power is direct
current, and they can be separated by the action of a transformer.
The direct-current path is from the power supply through two resis-
tors connected across the input transformer primary, through the micro-
phone cable, and then through the secondary windings of the microphone
output transformer and out through its center tap and to the preamp. The
direct current in the transformer secondary does not magnetize the core
because its direction in each half of the coil is such that it cancels itself out
as far as the core is concerned. There is no direct current in the primary of
the input transformer to the control console or tape recorder, even though
the power supply voltage is present on it. This voltage causes no harm
to the audio signal.

286
 Phase Linear

Microphones and inputs without transformers separate the AC audio

signal from the DC phantom power with blocking capacitors.
The voltage used for phantom powering is usually 48 volts, but it can
vary from about 12 to 52 volts. Microphones that use the lower voltages
have a regulator circuit to reduce the higher voltages so no harm is done
when they are plugged into a 48-volt phantom power supply. The techni-
cal details of phantom power are given in the iec standard 268-15.
Phase If a sine wave signal is viewed on an oscilloscope, the horizontal
axis is time, and the shape of the wave is the same as the mathematical sine
function, or sinusoid, related to the geometry of a right triangle. The sine of
an angle of a right triangle is the ratio of the length of the side opposite the
angle to the hypotenuse and is a function only of the angle itself. The sine
curve is thus a plot of this ratio versus angle, and the value of the angle is
called phase.
In the sinusoidal signal, time replaces angle in the graph, although it
is convenient to speak of angle as well as time being the independent vari-
able. A sinusoid goes through 360 degrees in one cycle, whereby it then
returns to its starting point. The sine wave signal is also said to go through
360 degrees in one cycle; another way of saying the same thing is that the
signal has gone through 360 degrees of phase angle, or phase change. The
phase is actually a measure of time; 360 degrees equaling one period of
the signal. The time represented by phase change of a certain number of
degrees is thus dependent on frequency.
Phase Compensation In some analog tape recorders, there is a special
equalizer whose purpose is to make the phase response of the machine
more linear, i.e., to reduce relative phase shift as frequency varies.
The purpose of phase compensation is to improve the reproduction of
transients, whose shape depends on the relative phase of the frequency
components.
Phase Distortion Phase distortion in an audio signal is relative phase
shift between various frequency components of the signal. Phase dis-
tortion causes transients to change in shape, but has a smaller effect on
steady-state tones.
Most transducers, such as loudspeakers and phono cartridges,
produce signicant phase distortion. See also phase.
Phase Invert A switch found on the inputs of many recording consoles that
inverts the polarity of the signal in that input. Its purpose is to correct
for cables that may be wired backward, microphones that may not be wired
according to standard, or for a special effect such as to make a mic sound
different in a mix by making the polarity inverted. The phase invert switch
is commonly labeled with a circle with a forward slash through it, some-
thing like a Greek letter Phi (). The term taken literally is meaningless,
but nevertheless has come into popular use. Correct usage would be polar-
ity invert. See also .
Phase Linear An audio device that will pass an audio signal without caus-
ing phase shift at any frequency is said to be phase linear.

287
Phase-Locked Loop

Phase-Locked Loop A phase-locked loop, or PLL, is an electronic circuit

consisting of a voltage-controlled oscillator and a frequency dis-
criminator connected in such a way that the output of the discrimina-
tor controls the oscillator. Such a system will lock on to an applied signal
and will oscillate at that frequency. The PLL has some inertia, and will
not change its frequency instantaneously if the input frequency changes.
It thus acts as a stabilizing inuence, or a sort of frequency ywheel. PLLs
are used in servo-control circuits, modulation, demodulation, phase mod-
ulation, frequency synthesis, and are frequently used as FM detectors, as
they require no tuned circuits, and so do not require alignment.
According to Agilent Labs, one can claim that PLLs are the most ubiq-
uitous form of feedback systems built by engineers.
Phaser A phaser, or phase shifter, is a device that gives an effect similar to
flanging, but with less depth. It works by shifting the phase of the sig-
nal and adding it back to the signal. This causes partial cancellation at fre-
quencies where the phase shift approaches 180 degrees. Phasing is
sometimes called skying in Britain.
Phase Shift Phase shift is a characteristic of a device and is the change in
phase impressed on a signal that passes through the device.
An audio device will always add a time delay to an applied signal. If
this time delay is constant at all frequencies, the phase shift between the
input and output of the device will be a linear function of frequency.
Such a system is called phase linear. (The term phase linear was used as a
commercial trademark for some audio devices that were not actually phase
linear in their performance.)
Very few devices are truly phase linear, and deviations from phase lin-
earity are called phase shift. equalizers in particular exhibit large amounts
of phase shift. In a complex waveform, phase shift will cause a distor-
tion of the waveform, even though the frequency response curve may
be perfectly at. There is considerable controversy over whether the ear
can detect this type of distortion, with many audio luminaries shedding
more heat than light on the subject. It is denitely true that phase shift can
be heard if it is large enough; and in agreement with most hearing phe-
nomena, it is probably true that there is a threshold value below which it
is inaudible and above which it is audible. The subject needs further study.
digital audio systems in particular have large amounts of phase shift at
very high frequencies.
Phasey, Phasiness Phasiness is a subjective term used to describe certain
anomalies in the perception of stereophonic imaging, and it is difcult
to dene. The term is used to describe the subjective impression when the
two ears hear similar sounds that differ signicantly in phase.
If one hears exactly the same sound in each ear, either via headphones
or loudspeakers, the resulting image location will be centered. If the rel-
ative phase of one of the channels is shifted, the image will move one way
or the other. If the phase is shifted more and more, a point is reached where

288
 Phone Plug

the image is diffused and smeared, and nally the sound becomes unpleas-
ant to listen to.
It is said that certain intensity stereo recordings and the ambison-
ics system sound less phasey than other types of stereo recording.
Phasing See phaser.
Phasing Plug See compression driver.
Phlogiston A hypothetical substance formerly supposed to exist in com-
bination in all combustible materials and to be released in the process of
combustion; the element re, conceived as xed in ammable substances.
(OED) See also smoke.
Phon The phon is the psychological unit of loudness level and is
dened as the sound pressure level (SPL) of a 1,000-hertz pure tone
that is judged to be the same loudness as the sound in question.
Because of the ears complicated response versus frequency charac-
teristics, it is no good to try to relate the perceived loudness of sounds
directly to their sound pressure level as read on a sound level meter.
The phon is therefore an attempt to relate perceived loudness to objective
measurements of the sound. See also fletcher-munson effect;
loudness.
Phonautograph The phonautograph was a device invented by Leon Scott
in 1857 that recorded the waveform of sounds with a stylus on a rotat-
ing glass disc coated with soot by holding it over a smoking ame. It used
a short horn coupled to a diaphragm which in turn moved the stylus,
and it is surely the most important precursor to the phonograph. The sur-
prising thing is that it took Edison another twenty years to take the step
from just recording the sound to being able to reproduce it.
Phone Patch A direct connection from an audio device to a telephone line
is called a phone patch if the audio signal is sent directly to the line. An
example of the use of the phone patch is the sending of program from the
studio to a remote transmitter in radio broadcasting. In order to prevent
extraneous noise and to avoid system damage due to component failure,
the analog signal is sent through a telephone hybrid rather than being
directly connected to the telephone line. If the signal is digital data, a
modem is used to interconnect with the telephone line.
Phone Plug One of the most used of audio connectors is the quarter-inch
phone plug, which gets its name from its origin in the old telephone switch-
boards. Phone plugs come in two main varieties, the three-conductor type
known as tip, ring, and sleeve (TRS), and the two-conductor tip and
sleeve (TS). The sleeve is always the shield or ground connection. When
used on stereo headphones, the TRS plug may be called a stereo plug. There
are several other smaller phone plugs common in audio. A .173-inch diam-
eter version is often used in patch bays and is typically called by any of
several trademarked names such as Bantam, TT, Tini-Telephone, Tini-Tel
or miniature. Most references to a miniplug refer to a smaller plug used for
audio circuits and in portable stereo headsets in the TRS version. The Amer-

289
Phonograph

ican version is traditionally .141 inches in diameter; however, some retail-

ers refer to it as a 18 inch (.125 inch), or 3.5 mm (about .137795 inch). The
international standard is 3.5 mm. Smaller still is the subminiplug, commonly
seen on cellular phone earsets and some other audio and control devices
where space is at a premium. The American version is traditionally .097
inches in diameter, however, some retailers refer to it as a 332 inch (.09375),
or 2.5 mm (about .0984252 inch). The international standard is 2.5 mm. See
also patch cord; trs.
Phonograph An irritating toy that restores life to dead noises, according
to Ambrose Bierce.
The term phonograph was used by Edison in his 1877 patent application
for the rst device capable of recording and reproducing sound, but the
rst known commercial use of the term was by the Englishman F. B. Fenby,
who in 1863 patented the Electro Magnetic Phonograph. His device, how-
ever, was for recording the keystrokes of a piano or other keyboard instru-
ment, as the player piano. Later, others used different terms:
gramophone (Berliners trademark for the rst disc recorder/repro-
ducer) and Graphophone (rst used by Alexander G. Bell for his inven-
tion of a machine that he claimed was an improvement on the phonograph).
Edisons phonograph recorded the signal as a variable-depth groove,
embossed, rather than cut, in a sheet of tin foil wrapped around a cylinder.
It was a hill-and-dale modulation rather than the side-to-side, or lateral,
modulation of the gramophone. There was a great controversy in the early
years of the phonograph industry over the relative merits of the vertical
and the lateral modulation. Edison continued to insist on the superiority
of vertical, even in his diamond disc phonographs introduced after
Berliners Gramophone began to cut into the cylinder machine market. The
Emerson Record Company introduced a disc system in 1916 that used mod-
ulation at 45 degrees to the surface, halfway between lateral and vertical!
This was an early attempt at compatibility between the competing systems.
It is strange that Edison thought the greatest commercial value of the
phonograph to be in dictating machines rather than in the recording of
music. As far as the exploitation of the capabilities of his invention was con-
cerned, it might be said that Edison barely scratched the surface.
Phonograph Cartridge See cartridge.
Phono Plug A small inexpensive coaxial connector used for interconnec-
tion of many audio devices, especially consumer devices. It was rst used
by RCA to connect the phonograph tone arm to the amplifier and is some-
times called an RCA plug or a pin plug. The mating jack is called a phono
jack, RCA jack, or pin jack. Phono plugs and jacks are notorious for becom-
ing noisy and unreliable after a period of use. Some of the newer ones are
gold plated, and this reduces the contact noise caused by corrosion or oxi-
dation of the metal. The same plug is used for most consumer composite
video connections and S/PDIF connections, although the cables for such
should have 75-ohm characteristic impedance. See also rca plug.
Picket Fencing The irregularly occurring distortion of fm broadcasts as

290
 Pilot Tone

picked up by a car radio in the presence of multipath distortion. Mul-

tipath distortion comes about because of the relatively short wavelengths
transmitted by FM and television transmitters. Because these wavelengths
of FM signals are so short, sometimes moving the receiving antenna only
a few feet will make the difference between a listenable and a completely
garbled signal.
Pickup Alternative term for phonograph cartridge. See cartridge.
contact microphones or other transducers used in electric gui-
tars are sometimes called guitar pickups.
Piezoelectricity, Piezoelectric Piezoelectricity is an electric charge that
occurs in some substances when they are squeezed or otherwise subjected
to a mechanical stress. Quartz is one of the better known piezoelectric mate-
rials, and is commonly fabricated into small pieces, called crystals,
which are used for frequency standards. The crystal vibrates, similar to
a miniature tuning fork, in response to an electric signal, and the frequency
of vibration is very stable. Virtually all electronic watches use quartz crys-
tals as frequency standards, and they can be very accurate.
A piezoelectric element was also sometimes used as the active part of a
phonograph cartridge, although this use was restricted to inexpensive
players. The stylus moving in the groove of the record causes the element
to bend, and it generates an alternating voltage, which constitutes the sig-
nal. It is also possible to build microphones on the same principle, and
many such were made in the past. They are characterized by low price and
high impedance, which limits their use because of the short cables required.
Pi-Filter A three-pole ripple lter used in power supplies so called because
of the resemblance of its schematic drawing to the Greek letter (Pi). It
consists of a shunt capacitor, a series inductor, and another shunt capac-
itor. A similar conguration in a multipole lter is called a pi-section.
Pigtail The end of an audio cable that simply has bare wires rather than
any type of connector. Pigtails are used to connect cables to binding posts
or screw terminals. Some fuses with bare leads are called with pigtails.
Pilot A 19-kilohertz tone transmitted along with stereo FM broadcasts
in order to synchronize the local oscillator in the receiver to 38 kHz for
the detection of the stereo subcarrier.
This 19-kHz tone can be troublesome even though it is above audibil-
ity for most people. If not ltered out of the receiver output, it can interact
with the bias oscillator of a tape recorder, causing spurious tones to be
recorded on the tape. It also can cause mistracking of compander systems
such as Dolby, interfering with their ability to restore the proper dynamic
level of high-frequency signals. See also fm stereo.
Pilot Tone The pilot tone is a 60-hertz signal recorded on one track of an
analog tape that is used for motion picture sound recording. The pilot tone
signal is generated by the camera when the lm is being shot, thus the
frequency is an accurate measure of the camera speed. The tone, or pilot,
is then used later to synchronize the tape playback to the picture action.
This technique, which has many variations, allows movie sound to be

291
PIM

recorded independently of the lm, rather than being recorded directly on

the lm as it is shot. This is called double system sound. Pilot tone sys-
tems have essentially been replaced by smpte time code, which records
a digitally coded timing signal on the analog tape. The time code, when
played back and decoded controls the speed of the playback machine to
synchronize the picture and sound. Digital sound recorders are also used
with SMPTE time code.
PIM Phase Intermodulation Distortion. PIM arises in amplifiers that have
a nonlinearity such that one signal will cause phase modulation of another
signal. Phase modulation is the same as frequency modulation but to
a lesser degree. This is in contrast to ordinary intermodulation dis-
tortion, which is the amplitude modulation of a signal by another signal
and is sometimes called aim, for amplitude intermodulation distortion. For
PIM to occur, a strong signal must modulate the bandwidth of the
amplier, and this varying bandwidth varies the phase of another signal
that is also being amplied. The measurement of PIM in audio devices is
relatively new, but it has been known and measured in video ampliers
for some time, where it distorts the color rendition of the television picture.
The measurement of PIM is somewhat similar to the smpte method of
measuring AIM. A high-frequency tone (7,000 hertz) is mixed with a low-
frequency tone (60 Hz) of 4 times the amplitude and fed into the dut. The
DUT output is fed into a lter to remove the low-frequency tone, and the
phase of the resulting high-frequency tone is compared with the phase of
the input high-frequency tone. The unit of measurement is time, with the
results usually expressed in nanoseconds as a function of signal level. PIM
measurement techniques are not standardized, and there is no general
agreement on its degree of audibility.
Pin 1 Problem An equipment condition coined by Neil Muncy in 1995. It
refers to analog audio equipment with balanced connectors (e.g., XLR-3)
where pin 1 (normally connected to the interconnecting cable shield) is con-
nected to the audio circuitry ground inside the equipment, instead of being
immediately tied to the chassis and power ground. When the device is
connected to other equipment in a system, any shield currents on pin 1
can then modulate the audio ground, causing hum and buzz problems that
can be difcult to troubleshoot and eliminate. Such a problem does not show
up on standard bench tests, although a simple modied AC transformer
can be used to reveal a piece of equipment with the problem.
Pinch Effect Phonograph records are cut with a pointed, chisel-shaped cut-
ting stylus and are played back with a circular, elliptical, or other suitably
shaped stylus. Because the playback stylus has a nite width compared to
the cutters knife-edge, it will be squeezed upward during those times when
the groove is narrowest. This is the pinch effect, and it causes harmonic
distortion in the reproduced signal. In the case of monaural record-
ings, the pinch effect can be canceled by using a cartridge that is not sen-
sitive to vertical motion of the stylus; but in the case of stereo records, the

292
 Pink Noise

Pinch Effect

vertical component of stylus motion is where all the directional informa-

tion is.
Pinch effect is minimized by using a playback stylus with very small
radii on the sides. The elliptical stylus was the rst of these, but there are
several other designs.
Pinch Roller See puck.
Ping-Ponging Recording studio slang for overdubbing. Also, the rapid
panning of an audio track back and forth between extreme left and extreme
right.
Pink Noise Pink noise is a type of random noise that has a constant amount
of energy in each octave band, as opposed to white noise, which has
equal energy per hertz. Pink noise can be made from white noise by pass-
ing it through a filter with a 3 decibel per octave rolloff.
Because our ears tend to hear pitches on a logarithmic frequency scale,
our perception of loudness correlates with the energy in constant per-
centages of bandwidth. In experiments conducted many years ago, Har-
vey Fletcher, of Bell Telephone Labs, found that sounds within one critical
band of frequency (about one-third of an octave) had a perceived loud-
ness depending on the energy in that band of frequencies. Sounds sepa-
rated in frequency by more than a critical band each had a loudness
depending on its own energy content, and none inuenced the loudness
of another.
Because the critical bandwidth is a constant percentage of the center fre-

293
Pinna

quency of the band, as frequency rises, the critical bands become wider and
wider in frequency. A third octave centered at 1,000 Hz is 230 Hz wide, but
a third octave centered at 100 Hz is only 23 Hz wide. Our ears integrate the
energy in these bands to determine perceived loudness. Therefore, pink
noise, which has constant energy per percentage bandwidth, sounds as if
it has relatively uniform loudness at all pitches. (It would sound equally
loud at all pitches if it were not for the fletcher-munson effect.) This
also means that white noise sounds very bright and lacking in low-frequency
content.
Pink noise is a useful test signal for measuring frequency responses of
audio equipment if the detecting instrument is a real-time analyzer
(RTA) with octave band or one-third octave band response. The noise sig-
nal is input to the device under test, and the output is plugged into the real-
time analyzer. Because the RTA has octave or one-third octave lters, it will
show flat response for pure pink noise. Any irregularities caused by the
dut will show up as deviation from at response.
Pinna The outer ear, the visible part of our hearing mechanism. The shape
of the pinna has a great deal to do with our ability to determine where a
sound is coming from. See also binaural.
Pirate Recording An unauthorized recording of a musical performance.
Pirate recordings were often made with home-type equipment under poor
conditions and the resulting quality was bad enough that it posed no threat
to commercial recordings. On the other hand, with the advent of very small
battery-operated DAT, MiniDisc, and MP3 recorders, pirate recordings can
be made with great care and attention to quality. Even though unautho-
rized, sometimes these recordings can be of commercial value. There are
reportedly some people who simply enjoy the challenge of planning, set-
ting up, and doing pirate recordings of famous musicians without being
detected, and would not sell the recordings or otherwise prot from the
activity.
Pi-Section, -Section A three-pole filter consisting of two shunt reac-
tances and one series reactance. It is so called because of its schematics phys-
ical resemblance to the Greek letter . The section has a at frequency
response in its passband and an 18 dB per octave slope in its stopband.
Pitch The pitch of a musical sound is a subjective quality relating to where
on a musical scale the sound is perceived to be. Pitch is closely related to
timbre, another subjective attribute of a musical sound.
Pitch is related to the frequency of a sound, but not in a simple way.
In general, perceived pitch increases as the frequency rises, and the pitch
will increase by one octave if the frequency is doubled. This is true for
complex sounds that are not too close to the upper or lower frequency lim-
its of our hearing. At frequencies above 4,000 hertz or so, the perceived
pitch changes very little with frequency; and with sine wave tones rather
than complex tones, pitch perception is very inaccurate. It has been
shown that at certain frequency ranges, two tones of different timbre can
be two octaves apart in frequency and still have the same pitch. This effect

294
 Pitch

Pi-Section

was observed earlythe French scientist and musician F. M. Mersenne

noted in his Harmonie Universelle of 1636 that the bass recorder, a large ute-
like instrument, sounds an octave lower than its actual frequency output
would indicate. This has not been adequately explained to this day.
It is interesting that a persons two ears do not generally hear a sound
as having the same pitch, especially in the case of sine waves. It is common
to nd differences of several musical half-tones in pitches of sine tones pre-
sented to each ear individually. For most people, the pitches fuse into a sin-
gle pitch sensation if the two ears hear the tone simultaneously, but in some
rare cases, a single note from a musical instrument will be perceived to be
sounding two pitches at once. Such a hearing problem is of course a dis-
aster for a musician.
Pitch also refers to the number of grooves per inch on the surface of a
phonograph record, cd, and dvd. See also variable pitch. Incidentally,
there is normally only one groove on each side of a phonograph record,
and only one digital path on each layer of a CD or DVD. However, there
exist some LP records with more than one groove on a side. An example is

295
Pitch Control

Monty Pythons Matching Tie and Handkerchief album. One side has two inter-
laced grooves, so you hear one of two programs on that side depending on
how you hit the lead-in grooves.
Pitch Control A control that can vary the speed of a turntable or analog
tape machine. Some digital devices, such as CD players, also have pitch
control. See also varispeed.
Pitch Tracking A slightly misleading term meaning frequency-to-voltage
conversion. pitch is a subjective term, whereas frequency is a purely
objective measure of a signal.
A pitch tracker will accept a complex periodic signal and extract from
this the fundamental frequency. It will then convert this frequency into
a direct voltage output that can be used as a control voltage in an elec-
tronic music synthesizer. Pitch tracking becomes more difcult as the input
waveform becomes more complex, but practical devices are made that
can handle most musical instruments, including the voice.
PIV Acronym for Peak Inverse Voltage, which is the maximum voltage
a semiconductor diode can withstand in the reverse direction without
damage. Also called PRV, or Peak Reverse Voltage.
Planar See loudspeaker.
Plate In an electron tube, the anode, or positive electrode, is called the
plate.
Plate Dissipation The maximum power that the plate of a vacuum tube
can dissipate without permanent damage to the tube. It is equal to the plate
voltage times the plate current and is measured in watts. Some high-power
tubes are so designed that the plate may be operated continuously in a red-
hot condition without damage.
Plate Resistance The internal resistance of a vacuum tube, measured
by the change in the plate voltage divided by the change in the plate cur-
rent. The plate resistance depends on the potential of the control grid and
is usually expressed as a family of curves taken at various values of grid
bias. These are called characteristic curves.
Plate Reverberation One of the rst synthetic reverberation devices uses
a steel plate that is under tension supplied by springs at the corners. The
plate is vibrated in accordance with a signal from a loudspeaker-like
device, and the vibration is sensed at another place on the plate with a con-
tact microphone of one type or another.
Plate reverberation has been used a great deal in the recording industry
over the last forty years, even though it sounds very little like natural rever-
beration. The advantage of the plate is its small size and low cost compared
to a reverberation room.
The advent of digital reverberation simulators has doomed the plate,
even though they dont sound like real reverberation either.
PLL See phase-locked loop.
Plosive A sudden puff of air emitted by the voice when pronouncing words
containing the letter p and, to a lesser extent, t. See also pop filter.
Plug-in A piece of software designed for a specic audio task and designed

296
 Point Source

to work from within another piece of software. For example, many com-
panies produce plug-ins for Digidesigns ProTools editing program.
PM Permanent Magnet, normally an adjective, as in PM speaker for a
loudspeaker with a permanent magnet.
PMCD Pre Master Compact Disc. Originally, master recordings for com-
pact discs were submitted on specially formatted 34-inch U-Matic
videocassettes made on Sony 1610 or 1630 digital recorders. With the
advent of CD-R, the Sonic Solutions and Sony companies came up with
an equivalent master format on a CD-R, which not only contained the
audio but also special data tracks and special cues for the laser beam
recorder (lbr). A special hardware/software combination was required
for creating the CD-R, but an ordinary player could reproduce the audio
for testing purposes. One could submit either a 1610/1630-style tape or
a PMCD to be replicated into CDs.
The system has fallen by the wayside and today PMCD generally de-
notes an ordinary red book compliant audio CD-R that is ready for dupli-
cation exactly as is.
P-Mount A P-mount is a universal standard mounting arrangement for
installing a phono cartridge in a tonearm. In the P-mount system, the
headshell is not neededactually it is a part of the cartridge. P-mount
cartridges are simply plugged into the end of the arm and are easily inter-
changed.
The P-mount ensures the correct placement of the stylus in relation
to the record groove by using a standardized distance from the end of the
arm to the stylus tip, and the angle of the stylus cantilever is accurately
controlled.
PNdB Perceived Noise decibel, which is the unit of perceived noise level.
It is equal to the spl of the 1,000-hertz octave band of pink noise, which
is subjectively judged to be equally noisy as the noise under consideration.
The PNdB is seldom used outside the psychology laboratory.
PNP Short for Positive-Negative-Positive. PNP refers to a type of tran-
sistor that is constructed of three layers of semiconductor material that
are positive-negative-positive. The manner of atomic doping of the semi-
conductor material determines whether it is P or N type. The emitter is
a lead attached to a P layer, and in a circuit, its voltage is positive with respect
to the collector lead, which is attached to other P layer. Signals to the base
lead, attached to the N layer, control current ow. See also npn.
Point Source A hypothetical sound source that is very small compared to
the wavelengths of sound it is radiating, and that is radiating into free
three-dimensional space.
True point sources do not exist, but some sound sources approximate
point sources, especially if the listener is some distance away. A point source
would radiate sound equally in all directions; it would be completely omni-
directional. One school of thought says that a loudspeaker for music
reproduction should behave like a point source, so it would radiate sound
equally in all directions in the listening room. Such a loudspeaker would

297
Polarity

excite all the reverberation in the room and would give the listener a
maximum sense of envelopment in the sound eld. This assumes good
acoustics in the listening room.
Musical instruments, especially large ones like pianos, are far from point
sources, and they radiate sound very nonuniformly into space, sending high
frequencies in some directions and mid and low frequencies in other direc-
tions. This makes life difcult for the poor recording engineer, especially
if he chooses to use close miking, for there is no one place that will pick
up a representative sound of the entire range of the instrument. This is one
reason that very close miked recordings do not sound the way the instru-
ments actually sound in a room. For a more natural sound, the microphone
must be at least one wavelength away from the instrument at the lowest
frequency of interest. A room with good acoustics is required, however,
because at the increased distance, the reverberation of the room becomes
a more signicant part of the sound that the microphone hears.
Polarity If the two wires to a loudspeaker are reversed, the signal is
turned upside down, that is, points on the waveform that moved the
loudspeaker cone outwards now move it inward. This is called a polarity
reversal. It also is equal to a 180-degree phase shift.
Many times, the term out of phase is used when a polarity reversal is
meant. Out of phase is a matter of degree, whereas polarity denes the
condition.
Recently there has been renewed interest in the absolute polarity
of an audio signal. It is thought that there should not be a polarity rever-
sal anywhere between the microphone and the loudspeaker. In other
words, sound pressure in a positive direction at the microphone should
result in a positive-going sound pressure at the listeners ears. Some exper-
iments have been conducted to determine the ears sensitivity to this polar-
ity reversal, and indeed it seems that it is possible to tell the difference
between a signal and its inverted brother. It is a different matter, however,
to tell which one has the correct absolute polarity, and the matter has not
been conclusively proven.
Polarize To place a constant (static) voltage across a device or circuit
element is said to polarize it. An example is the condenser microphone,
which requires a polarization voltage to charge the capacitive element. Some
electronic circuit elements, such as electrolytic capacitors, require a
polarization voltage to allow them to operate in a linear fashion.
Polar Pattern See polar response curve.
Polar Response Curve The polar response curve, sometimes called the polar
pattern, is a plot of the sensitivity of an audio device as a function of
angle around the device. Thus, the polar response of a loudspeaker tells
one the relative strength of signals radiated in various directions. To be
meaningful, polar response plots must be measured at many frequencies,
for no device will have the same polar curve over a wide frequency band.
Polar response curves of microphones are especially important because
they tell the user how to point the microphone to optimize the frequency

298
 Polyester

Polar Pattern

response. Ideally, polar curves would all be circles, indicating no discrim-

ination due to direction, or at least they would all be parallel, indicating
the same response at all frequencies.
Pole The part of the surface of a magnet where the magnetic eld is most
concentrated is called a pole. Any magnet has two poles; one is the north-
seeking pole, or simply the north pole, and the other is the south-seeking,
or south, pole. These names are given because a suspended magnet will
orient itself with the north pole pointing north. The earth is a giant mag-
net and thus also has north and south magnetic poles, but it follows that
the magnetic pole in the northern hemisphere is actually a south pole and
vice versa. See also pole piece.
Also a mathematical condition where a polynomial has solutions that
approach innity, used in lter design and analysis.
Pole Cup A cylindrical hole in a speaker cabinet into which a pole is inserted
to support or elevate the speaker.
Pole Piece A shaped piece of high-permeability metal, usually soft iron,
which serves to concentrate and direct the magnetic eld from a perma-
nent magnet to maximize the efciency of such devices as dynamic loud-
speakers, magnetic cartridges, and cutterheads. Pole pieces are
needed because magnets are difcult to make in the complex shapes
needed in such applications.
Pollywogs Tiny irregularities in the surface of the tape used in cassettes,
caused by errors in the manufacturing process. Pollywogs cause momen-
tary loss of contact between the tape and the head, producing dropouts.
The term is primarily British.
Polyester A type of strong, exible plastic used as the backing for certain
magnetic recording tapes. It was developed by DuPont, who called it Mylar.
Mylar tape will last much longer than acetate tape but can be stretched
if placed under too much tension, whereas acetate will break. It is possible
to repair a broken tape but not a stretched tape! Some tapes are made from

299
Pop, Popping

Mylar that has been prestretched, or tensilized, and they are much less
likely to cause trouble.
Pop, Popping When a microphone encounters air currents, its diaphragm
is jostled and the result is either a rustling noise or, if the wind is strong
enough, explosive sounds called pops. Needless to say, this is not good for
the microphone and thus must be avoided. Some microphones, especially
ribbon types, are very sensitive to damage by air currents.
Most directional microphones will pop when a singer or speaker is at close
range and utters plosive sounds, especially the letter p. See also pop filter.
Popcorn Noise A type of intrinsic noise produced by a defective piece of
solid-state electronics. It resembles the sound of popping corn. Also a
slang term in the motion picture sound world that refers to the ambient
background noise in a movie theater. The audience eating popcorn is only
one contributor to this kind of popcorn noise.
Pop Filter An acoustically transparent (transondent!) plastic foam mate-
rial placed over a microphone that will reduce the effect of air currents
and plosives (pops) from the voice at close range. Such foam must be
open-cell or reticulated, that is, all the foam bubbles are open to each other.
Closed-cell foam would block much of the sound. Some forms of pop lters
are screens made of a metal ring several inches in diameter with a very ne
nylon gauze (commonly made of panty hose material) stretched over it and
placed in front of the microphone. Pop lters are sometimes built into the
mesh ball in the business end of a vocal microphone. The pop lter also
acts as one type of windscreen.
Port The opening below the loudspeaker in a bass reflex system is
sometimes called a port. See also ducted port.
Portamento A musical technique of sliding between two pitches. Sometimes
called glissando. Synthesizers can make use of this under midi control.
Portastudio A trade name by the Tascam company for a small tape recorder
that recorded four tracks on compact cassettes. The unit would also
provide basic studio functions such as overdubbing and mixing. The track
format and speed was not compatible with regular stereo compact cassettes.
Port Noise At very low frequencies, the air movement in a bass ducted port
of a loudspeaker cabinet makes a distinct sound or distortion. Of course,
if you speak negatively about it, this is known as a port noise complaint.
Post Simply means after as opposed to pre. In recording studio parlance it
is used to indicate that the signal has already had the designated effect
added, such as postequalization, posteffects, or postfader. Also used as an
abbreviation for postproduction.
Post Echo See print through.
Postfader In a mixing console, sends are either prefader or postfader. Post-
fader sends are affected by the position of the input fader, whereas prefader
sends are not. An effects send is often postfader, so the level of the send to
the effects device reects the relative position of the input fader. In some
cases, a send can be switched to be either pre- or postfader.
Postproduction Any work on a recording, lm, or video that is done after

300
 Power Amplier

the main recording session or lming is completed. Typical postproduc-

tion work can include editing, sweetening, and mixdown.
Pot See potentiometer.
Potentiometer A variable resistor, usually controlled by a rotary knob, used
extensively as a volume control, tone control, etc. The potentiometer has
three terminals: one at each end of the resistor and the third connected to
the movable contact, or wiper. The term is also used for a step-type
attenuator, although this is actually a multiple contact switch wired to
a series of xed resistors. The common slang term for potentiometer is pot.
The slide pot is a linear motion potentiometer used for faders.
POTS An acronym for Plain Old Telephone Service, meaning a traditional
analog telephone line.
Poulsen, Valdemar Valdemar Poulsen invented the rst working model
of the magnetic recorder. He called his invention, made in 1899, the
Telegraphone, and it used a steel wire running at seven feet per second
as a recording medium. He rst thought of it as a device for recording Morse
code telegraph transmissions for later transcription, and he also seems to
have thought it might be used to record telephone conversations. However,
he didnt hit on the idea of using it to record music.
Poulsens model was preceded by eleven years by a description of a
remarkably advanced concept for magnetic recording by Oberlin Smith,
who wrote a letter to the editor of Electrical World. See smith, oberlin.
Power Power is dened as the rate of doing work and is measured in watts,
1 watt being 1 joule per second, which is equal to the power absorbed by
1 ohm of resistance when 1 ampere of current is in it. Power in an elec-
trical circuit is calculated by the voltage squared divided by the resist-
ance, the current squared times the resistance, or the voltage times the
current.
The power rating of an amplifier is the maximum power it will
deliver to a specic load and is rated in watts of continuous power. The
term rms power is a misnomer and should be avoided (see rms). An
amplier will deliver its maximum power into only one value of load
impedance. Connecting a load of higher impedance will reduce the power
output but will not cause any other problems; but connecting a lower than
rated impedance to an amplier will usually cause distortion as well as
reduced output.
Power-handling capacity of a loudspeaker is somewhat difcult to
dene. Any speaker will handle much more power when reproducing music
than when reproducing a sine wave test signal because of the large ratio
of peak power in music compared to the average power. One of the limi-
tations in speaker power handling is due to heating of the voice coils,
and another is mechanical damage due to large excursions of the moving
parts. See also decibel.
Power Amplier An amplifier that accepts a low-level audio signal and
strengthens, or amplies, it to a suitable voltage and current level ade-
quate to drive a loudspeaker or other load.

301
Power Amplier

The four key parts of a power amplier are the input stage, driver stage,
output stage, and power supply. The power supply provides the electrical
power to drive the loudspeaker. Most modern ampliers use solid-state
electronics, and have a dual, or bipolar, power supply, one side providing
positive voltage at fairly high current and the other side providing nega-
tive voltage at fairly high current.
The output stage, which is also usually bipolar, or push-pull, acts as
a controller and connects the positive and negative power supply outputs
to the load in response to the audio signal. In a sense, the output stage acts
as a pair of variable valves, gradually turning on and off the currents as
required to duplicate the waveform of the signal. The British term valve
for our vacuum tube is quite descriptive here. In actuality, it is the effec-
tive resistance of the output devices that is varied. One undesirable con-
sequence of this state of affairs is that the output devices, which must pass
all the current that enters the load, will dissipate heat because of their inter-
nal resistance. This heat is wasted energy and can cause problems such as
premature failure of components.
The function of the driver stage is to control the output devices and usu-
ally to split the signal into two parts with opposite polarity, which is nec-
essary to drive the positive-going and negative-going output devices. This
function is also called phase splitting.
The input stage of the amplier provides the needed voltage gain for
the amplier.
The design of the output stage has traditionally attracted the most atten-
tion, and many different designs have been developed to increase its
efciency and reduce its distortion. Ampliers can be categorized into
several different classes as follows:
Class A: This was the rst type of amplier, and it is in theory the most
linear of the many types. Class A operation means the output devices are
always conducting current, even when the signal level is zero. The instan-
taneous signal level modulates the current up and down, but the current
never reaches zero. In fact, the average current is almost constant in a class
A amplier, regardless of the signal level. This means it is always dissipating
a good deal of heat in the output stages, and it is therefore very inefcient.
Efciency is dened as the amount of power consumed by the amplier in
relation to the amount of power it delivers to the load.
The class A amplier is capable of low distortion, even with small
amounts of negative feedback, but its low efciency relegates it to fairly
low output power capabilities. There are some people, however, who insist
that the only amplier t for music is a class A amplier, and some high-
power class A ampliers have been built that are so inefcient that they
could be used as electric heaters.
Class B: The class B amplier was developed to improve the efciency
of the class A designs. In it, each output device conducts current for only
half of the waveform. One device handles the positive-going parts of the
waveform while the other device handles the negative parts. In effect, the

302
 Power Amplier

Class A, B, and C Ampliers

signal is handed off to each output device in turn. This can work well
from an efciency standpoint, but there is a problem with the switching
from one device to the other at very low signal levels. If switching is not
done absolutely smoothly, a type of distortion called crossover distor-
tion occurs.

303
Power Amplier

Class AB: Problems with crossover distortion in class B ampliers led to

a design in which both output devices conduct current when the signal level
is very low. In effect, the amplier operates in class A at very low levels
and gradually becomes class B at higher signal levels. The result is much
reduced distortion, but at the expense of lower efciency than class B.
Class C: To further increase efciency, although at the expense of a great
deal of distortion, the output device of the class C amplier conducts cur-
rent only over a portion of the input waveform. The lower level parts of
the waveform are not amplied at all, resulting in the very high distortion.
Class C ampliers are used in rf circuits to amplify narrow frequency bands.
if ampliers in radio and television receivers are class C. The ampliers
are connected to sharply tuned resonant circuits tuned to the frequency of
interest, effectively ltering out the distortion. The class C amplier is not
used as an audio power amplier.
Class D: An interesting type of power amplier in which the audio sig-
nal varies the width of a series of constant-amplitude pulses. The pulse rep-
etition rate is much higher than the highest audio frequency. The average
area under the pulses is equal to the instantaneous power level of the sig-
nal. The pulses are then passed through a low-pass filter, which
smooths, or integrates, them into an amplied replica of the input wave-
form. The phenomenon is called pulse width modulation (pwm); although
this may appear to be a digital technique, it is not, for the formation and
modulation of the pulses are done by purely analog means. The class D
amplier is very efcient because the output devices are operating either
completely on or completely off, neither condition dissipating any
appreciable power in the devices themselves.
Some so-called switching power supplies operate very much like a class
D amplier. The width of the pulses is modulated to vary the output volt-
age or to maintain a constant output voltage under conditions of varying
load current.
Class G: A type of power amplier circuit somewhat similar to but more
efcient than class B. In a class G amplier, there are two bipolar power
supply voltages that are selectively connected to the output stage in accor-
dance with the signal level. When the signal is low, the supply voltages are
low, providing good efciency. When the signal levels are high, a higher
voltage supply is switched in to allow the output devices to pass the extra
power. Compared to class B, class G operation has several advantages:
higher peak-power capability, reduced thermal problems, increased power
output, and relatively low cost.
Class H: This type of power amplier circuit is an extension of the class
G design. In class H, the power supply voltage for the output devices is
modulated so that it is always just above the signal level. This increases
the efciency of the amplier because it is the excess power supply volt-
age that causes unnecessary power dissipation and heating in the output
devices. The class H amplier, like the class G, has two bipolar power sup-
plies, but instead of switching back and forth between them as the signal

304
 Preamplier

level changes, the high-voltage supply is actually amplitude-modulated

to keep it just above the instantaneous peak signal level. The class H
amplier has the same advantages as the class G but is even more efcient.
See also tube.
Power Bandwidth The bandwidth of an audio device such as a power
amplifier measured while delivering its full rated power output. This is
limited by the slew rate of the device and is always narrower than the
small signal bandwidth.
Powered Mixer A mixer that also has a built-in power amplifier for driv-
ing passive loudspeakers directly. This means you lug around one
device instead of two.
Powered Speaker See self-powered speaker.
Power Response A type of frequency response where sound power ver-
sus frequency is plotted instead of the usual sound pressure versus
frequency. Power response is applicable to loudspeakers because a sim-
ple spl response curve is dependent on the angle between the speaker and
the microphone. Most spl measurements are made on axis of the loud-
speaker and at 1 meter distance. This does not provide any information on
how the speakers polar response curve varies with frequency. The
power response integrates all the energy produced by the speaker, regard-
less of angle. The rationale for measuring it is that in a real room, the lis-
tener hears sounds from all angles as the sounds reect from the walls, etc.,
and therefore the perceived frequency response will be more like the
power response than the spl response.
However, there is a complication, as usual in acoustics. A corollary
of the haas effect says that the ear will judge the timbre of a sound by
the earliest sound it hears. The reected sounds are later than the direct
sound in a real room, so the spl response of the speaker is what the ear
hears rst. Therefore, the spl frequency response at the angle where the
listener is located should be flat, as should the power response.
Power Supply See b-plus and supply rails.
PPI Peak Program Indicator. See ppm. Many radar screens are also called
PPIs meaning Plan Position Indicators
PPM Short for Peak Program Meter, which is similar to a vu meter in appear-
ance, but which responds to the peak, or maximum, level of the signal
rather than the average level. The ppm is useful in applications where instan-
taneous overloads create large amounts of distortion, such as digital
audio devices. Any signal level above what a digital system can encode is
clipped at that level. A meter that reads the average level of a signal does
not respond quickly enough to register the short transient peaks that
abound in music.
The PPM is often found in control desks/consoles, especially for broad-
cast, and in some analog tape recorders and also in digital machines. It
may take the form of a series of leds rather than a mechanical meter, but
the function is the same. See also modulometer.
Preamplier, Preamp In an audio system, the rst amplifier to accept the

305
Precedence Effect

signal from the transducer is generally called a preamplier. Exam-

ples are phono preamps, which amplify the output of the phono cartridge,
and mike preamps, which amplify the signal from a microphone.
Preampliers usually must accept very low-level signals and amplify
them without adding appreciable noise. Many preamps also contain
appropriate equalization (see riaa), such as phono preamps and tape
playback preamps. An exception is the pre-preamplifier, which is a spe-
cially designed low-noise preamplier to be used with moving-coil car-
tridges. It does not contain any equalization because its output is fed into
an RIAA preamp.
The design of good quality preampliers is not easy, for they must be
made to be immune to the introduction of many types of noise. Microphone
preamps have balanced inputs for this reason. Also, preamps contain-
ing equalization, such as RIAA preamps, must handle a very wide dynamic
range of input signals without overload distortion. Often the noise input
to preamps will be of very high frequency, such as in the case of radio or
television interference. If the preamp is nonlinear at these high
frequencies, the signal may be detected, and the interference enters the
audible frequency range.
Precedence Effect See haas effect.
Pre-Echo Sometimes when a record is being pressed and if the grooves are
very close together, one groove may inuence the shape of the adjacent one,
causing the sound from the rst one to be faintly heard in the second one.
This is called pre-echo because it is usually heard at the very beginning of
a record or record band as a soft precursor of the music to come.
In making records, the stamper has negative grooves, or hills, which
form the grooves during pressing. When the soft vinyl ows out from the
center as the press closes, it must ow over the interior hills on its way to
the outer ones. Because of its molecular structure, it tends to remember its
recent history, and it retains some of the shape of the previous hill as it set-
tles into its nal form. See also print-through.
There is another, little-known type of pre-echo which sometimes
occurs in digital audio systems. It arises from ripple in the passband
of the anti-aliasing and anti-imaging filters. In some low-pass
lters, the amplitude response varies up and down at a regular rate along
the frequency axis. This is called ripple and is usually held to a low value
such as plus or minus 12 decibel. This ripple, however, results in an
impulse response of the filter, which has skirts both before and after
the main response peak. (The impulse response of a system completely
describes its frequency and phase response.) This means that if a short
impulse is presented to the lter, there will be some output signal before
the main burst of the impulse, and in reality it can be several millisec-
onds before the main burst. This results in a general muddying of musi-
cal transients.
There is presently no agreed-upon method of measuring this type of pre-

306
 Pre-Emphasis

echo, but it has been shown to be clearly audible. It seems that the dual-
channel fft analyzer is the best instrument to measure it, for it can pro-
duce a picture of the magnitude of the impulse response of a system.
Pre-Emphasis Pre-emphasis is a type of high-frequency boost applied to
signals about to be broadcast on fm stations, or which are about to be
recorded on tape or a phonograph record. The reason for the boost is to
increase the level of the higher frequencies so they are well above the high-
frequency noise generated by the transmission medium (i.e., radio broad-
casting) or the recording medium (i.e., tape recorder or vinyl disc) to increase
the apparent signal to noise ratio.
The long-term averaged spectrum of most music has a maximum at
about 500 hertz, and falls off above and below this range. In other words,
there is much more energy in a music signal in the mid-range than at the
high and low frequencies. This means that the full dynamic range of the
recording medium is not used at the extremes of the frequency range, and
the signal-to-noise ratio is poorest there. The ear is much more
sensitive to high-frequency noise than low-frequency noise due to the
fletcher-munson effect. Pre-emphasis brings the high-frequency con-
tent of the music up to a level further above the ambient noise level of the
recording medium.
In order to restore the proper balance of high and low frequencies to
the reproduced music signal, the boost added by pre-emphasis must be
removed by a complementary cut. This is called de-emphasis, and is
applied to the signal when reproduced from the recording medium or
received with an FM receiver. The de-emphasis reduces the strength of the
high-frequency noise, just as if the treble tone control is turned down. In
other words, the noise reduction takes place in the de-emphasis, and the
pre-emphasis is added to make the de-emphasized signal have flat fre-
quency response.
Pre-emphasis combined with de-emphasis is a very effective noise reduc-
tion technique, and has been in use for many years. Its effectiveness
depends on the high-frequency signal handling capability of the medium,
i.e., the greater the capacity of the medium, the greater the amount of pre-
emphasis that can be used without adding too much distortion.
In the case of early monaural phonograph records, various manu-
facturers did not agree on the optimum amount of pre-emphasis, and pre-
ampliers had several different settings of high-frequency rolloff to
accommodate different records. The advent of the stereo record in 1958
heralded the establishment of the riaa standard pre-emphasis curve, and
all records made after that date had the same pre-emphasis. At last, any
phono preamp with the RIAA playback curve could play any record with
correct equalization. In a way, it is unfortunate that the RIAA standard
calls for so much pre-emphasis, for the result is a very greatly exaggerated
high-frequency signal on the record. This signal is difcult to track with
the stylus, causing undue distortion and placing heavy demands on the

307
Prefader

rst stage of the preamp, which must linearly amplify this signal. It is not
feasible now to change the standard because of the great number of records
and players in existence.
The opposite effect exists in the case of tape recorders. When the nab
set the standard pre-emphasis curve for tape recorders in 1954, the best tapes
available then had relatively poor ability to accept strong high-frequency
signals, and the amount of pre-emphasis was quite conservative. In the
meantime, tape quality has increased markedly and much greater pre-
emphasis could be used with no signicant added distortion and with con-
comitant further reduced noise on playback.
An attempt was made by the Ampex company in the 1960s to introduce
a more suitable curve, called the AME curve (for Ampex Master Equaliza-
tion), but it was not universally adopted and fell into disuse. Stefan Kudel-
ski introduced a different curve (the nagramaster curve) in his Nagra
recorders in the 1970s, and it has enjoyed relative success in recording mas-
ter tapes.
All this points up the complex problems entailed in standardization in
disciplines where improvements in quality are constantly being made.
DAT and CD also have an optional pre-emphasis, with a signal bit (ag)
to activate it or not. Unfortunately, not all devices have the deemphasis (par-
ticularly computer CD drives), so it is possible to have a DAT or CD with
preemphasis and hear it undecoded on your system. However, it seems the
vast majority of these tapes/CDs appear to be non-preemphasized.
Prefader In a mixing console, sends are either prefader or postfader. Pre-
fader sends are unaffected by the position of the input fader, whereas post-
fader sends are. A monitor send is usually prefader, so the monitor
continues to operate when the input fader is down.
Prepolarized Condenser Microphone See electret microphone.
Pre-Preamplier A pre-preamplier, in addition to being linguistically ques-
tionable, is a very low-noise amplifier designed to operate at extremely
low signal levels.
Some modern moving coil cartridges have such low output voltages
that standard phono preamplifiers have too little gain to achieve
enough playback volume. moving magnet cartridges have several mil-
livolts of output while moving coils may have only a few tenths of a mil-
livolt. A step-up transformer can be used between the cartridge and
preamp to increase the voltage gain, but transformers are difcult to shield
against hum, can cause distortion, and may not have perfect transient
response.
The pre-preamplier is usually preferred in this application. It is a lin-
ear amplier (i.e., it has at response with no equalization), and is
designed to operate into the input of a standard riaa preamplier.
Presence Quite a few years ago, many home sound systems had a pres-
ence switch as a part of the preamplifier or amplifier. It caused a rise
in level in the frequency range between about 1 and 3 kilohertz, which
purportedly increased the illusion that the music was actually present

308
 Printed Circuit Board

in the room. The peak introduced in the frequency response was,

however, a distortion; it was eventually recognized as such and fell into
disuse.
The promotion and sale of the presence switch is further proof of the
theory that in the audio business, you can sell anything. See also loudness
control.
Pressing The common stereo record is frequently called a pressing in audio
circles, no doubt because it is manufactured by being pressed from a bis-
cuit of soft vinyl in a large hydraulic press. CDs and all other optical discs
are also pressed similarly, but no one seems to call them pressings.
Pressure Gradient Microphone A microphone that is sensitive to the vari-
ation of pressure over a distance rather than to the pressure itself. This pres-
sure variation with distance, called pressure gradient, is proportional to the
particle velocity. The particle velocity is the instantaneous velocity of the
air molecules, and is not to be confused with the speed of sound, which is
the velocity of propagation of a sound wave.
Pressure gradient microphones are of many different types, depending
on the electroacoustic transducer used. See gradient microphone and
velocity microphone.
Pressure Microphone Any microphone that is sensitive to variations in
the instantaneous air pressure surrounding it is called a pressure micro-
phone. dynamic and condenser microphones in their simplest forms
are pressure microphones. A perfect pressure microphone would be com-
pletely omnidirectional because the sound pressure accompanying a
sound wave is a scalar, or nondirectional, quantity, but practical designs
depart from the ideal because of sound diffraction around them. The
smaller the microphone body, the more ideal its response will be.
Pressure Pad On some analog tape recorders, usually of the inexpensive
type, there is a small felt pad that presses the tape against the head during
record or playback. This is to ensure good contact between the tape and
the head; however, it causes more problems than it solves. Excessive head
wear is one problem, and another is the introduction of scrape flutter
to the tape motion. This problem is worse with higher tape speeds. The pres-
sure pad is accepted as a necessary evil in the conventional cassette recorder,
however Nakamichi 3-head cassette machines do not use them. They push
the pad away and keep the tape against the heads by controlling the tape
tension by using two capstans and pinch rollers.
Preview Head A preview head is a supplementary playback head on a tape
recorder. It is designed for reproducing the master tapes for producing
phonograph records. See also variable pitch.
Primary The winding of a transformer to which the input signal or cur-
rent is connected, as contrasted to the secondary, which is the winding
where the output signal or current is extracted.
Printed Circuit Board, or PC Board Most electronic circuits made today
in mass production are constructed on epoxy berglass sheets, or boards,
onto which the electric conductors have been applied. Originally, the elec-

309
Print Through

tronic components were mounted to the board through small holes for the
lead wires, which were then soldered to the conductors. Today most cir-
cuit components are surface mounted, and holes in the board are only used
to establish electrical connections from one side of the board to the other.
The term printed circuit is a misnomer because the boards are not
printed, but rather etched. Sometimes the correct term, etched circuit
board, is heard.
Print Through When analog tape recordings are wound tightly on the reel,
the adjacent layers of tape sometimes inuence one another so that the sig-
nal from one layer will magnetize or print onto the next layer. This causes
a faint echo of the signal that may be heard as a pre-echo, audible before
the main signal. Print through is sometimes also called interlayer transfer.
Print-through is worse at high recorded signal levels, and some tape
types are much more susceptible to it than others. Thin tapes are more prone
to it than thick ones.
It is surprising, but print-through is much less obvious if the tape is not
rewound after recording or playing, and is stored in a tails out condi-
tion. Print through on a tails out tape exhibits Post Echo, since the
printed signal is played after the fact instead of before. This may be the rea-
son it is less audible and therefore, commercial and master analog tapes
are stored tails out. Another reason for storage tails out is that the tape pack
on the reel is very smooth and uniform, while it is quite ragged when fast
rewound on most machines, making rewound tapes more susceptible to
damage by abrasion or bent reels.
Processing In the manufacture of phonograph records from the master
acetate, several steps are taken, and they are collectively called process-
ing. In one-step processing, the acetate is rst silvered and then electro-
plated with metallic nickel. When this plating is stripped away from the
acetate, it is called a matrix and is a negative of the record, having hills
where the record has valleys, etc. This matrix can be used to press records
simply by using it as a mold in a record press. It is then called a stamper.
About 500 to 1,000 records may be pressed from a stamper, after which it
is worn to the point where high-frequency response will suffer and the noise
level will increase.
If more records are required to be made from the same acetate, the neg-
ative of the master is again electroplated with nickel and the plating is
stripped away. The result is a positive, or a replica of the acetate, which can
be played like a regular record. It is called a mother, and it is in turn
electroplated and stripped. This nal plating is then used as the stamper.
The mother may be repeatedly plated, making as many stampers as
required. Also, the original negative can be replated to make more moth-
ers. This whole procedure is called three-step processing.
ProDigi A digital audio format used in stationary-head multitrack digital
tape recorders, similar to the dash system. There were two formats, one
for 14-inch tape (up to 8 audio channels) and one for 1-inch tape (up to 32
audio channels). Various sampling rates were accommodated, and the for-

310
 Pro Logic II

mat allowed editing of digital tapes by mechanical splicing with a refer-

ence analog track, and by punch-in recording. Abbreviated PD. PD was
promoted by Otari, Mitsubishi, and AEG. The system has been abandoned.
Program In radio and television broadcasting, the main part of the signal
from the mixing console or remote board that will be broadcast is called
program, as opposed to monitor, intercom, PA feed, etc. The program is
always isolated from the other audio outputs in order to prevent any inter-
ference or noise from the other connections affecting the broadcast signal.
Progressive Scan Video Rather than dividing each video frame into 2 elds
as interlace scanning does, progressive scanning paints complete
frames on the screen at the frame rate, which can be 23.976, 24, 29.97, 30,
59.94 or 60 frames per second in the countries that conform to the atsc stan-
dards. The result is much less visual icker and increased picture sharp-
ness. Certain dvd players provide progressive scanning, but the television
set used must be progressive-scan-ready to be capable of handling it.
Computer CRT monitors use progressive scan rather than interlaced, and
they operate at various refresh rates such as 60, 75, and 120 per second. A
computers refresh rate means the same as the TV frame rate .
Project Studio A smaller recording studio setup, often in a home, to do
recordings without the expense of going to a full-service studio. A project
studio may be a precursor to going to a full studio, eshing out a project
rst at home to save studio time and money. The distinction can sometimes
blur, as some project studios can be better equipped than some full studios.
The project studio has been popularized by modern technology that offers
good quality multitrack studio equipment the average person can afford.
This does not necessarily mean the average person can now make decent
recordings, however.
Pro Logic Pro Logic is a matrix decoder developed by Dolby Laborato-
ries that decodes the four channels of surround sound that have been
encoded onto the stereo soundtracks of Dolby Surround program material
such as VHS movies and TV shows. Dolby Surround is a matrix encoding
process that in essence folds left, center, right, and surround channels onto
stereo soundtracks. A Pro Logic decoder unfolds the four channels on
playback (without a Pro Logic decoder, the encoded program plays in reg-
ular stereo). Such matrixing is sometimes called a 4-2-4 system, as four chan-
nels are encoded into two, transmitted, then decoded into four.
Pro Logic II Pro Logic II is a second-generation matrix decoder from
Dolby Laboratories that derives ve-channel surround (left, center, right,
left surround, and right surround) from any stereo program material,
whether or not it has been specically Dolby Surround encoded. On
encoded material such as movie soundtracks, the sound is more like Dolby
Digital 5.1, while on unencoded stereo material such as music cds the effect
is a wider, more complex sound eld. Pro Logic II provides two full-range
surround channels, as opposed to Pro Logics single, limited-bandwidth
surround channel. A software algorithm is available for use in certain Dig-
ital Audio Workstations (daws) to encode multichannel audio to provide

311
Proscenium

Pro Logic II surround sound in game software. Pro Logic II encoding can
be decoded by virtually every dolby digital home theater system and
is backward compatible with Dolby Pro Logic receivers and decoders.
Proscenium The part of the stage of a modern theater between the curtain
or drop-scene and the auditorium, often including the curtain itself and
the enclosing arch.
When the curtain is closed, the proscenium in a theater acts as a sepa-
ration between the audience and the goings-on on the stage. The opening
in the proscenium is called the arch, even though it may be rectangular
rather than arched. The part of the proscenium at the sides of and above
the arch is frequently used for the installation of sound reinforcement speak-
ers, usually without regard to their appearance. In new theaters, the speak-
ers are usually recessed behind the surface of the proscenium walls.
ProTools A trademark for a family of software and hardware for audio
production from the Digidesign company. The products so named have
evolved over many years but essentially provide for multitrack recording
and editing. Often referred to as PT.
Proximity Effect Proximity effect is the increase in low-frequency sensi-
tivity of a microphone when the sound source is close to the microphone.
It is a characteristic of directional microphones, and some are much worse
than others.
Proximity effect is a shortcoming, but sometimes it can be used to advan-
tage. If a directional microphone is placed close to a bass instrument, the
low tones will be enhanced, which could be advantageous for some music.
A singer placed close to a directional microphone will sound much bassier,
an improvement in some voices, we suppose. Some of the early radio croon-
ers and radio announcers used proximity effect to deepen and enrich their
voices, and many frequently still do.
Proximity effect comes about because of the nature of the sound eld
close to a sound source. The magnitude of the velocity component of the
sound wave near a sound source is a function of the wavelength as well
as the distance from the source. At low frequencies, where the wavelength
is long compared to the dimensions of the source, this velocity component
rises as the distance decreases faster than the inverse square law would
predict. Thus, any microphone that senses the velocity component (or the
pressure gradient) will have increased low-frequency output compared to
high frequencies.
The effect is described in mathematical detail in Leo Beraneks book
Acoustics, which has been reprinted in paperback by the asa (1986).
Pseudostereo The idea of multichannel listening has been around since
at least 1881 (see ader, clement), and many experiments since the 1920s
have been conducted in stereophonic and binaural broadcasting and
recording. It is surely inevitable that there would be many ideas put forth
that attempted to realize similar results with only one transmission chan-
nel. F. M. Doolittle suggested in 1925 that a stethoscope could be substi-
tuted for the conventional earphones and that if one of its rubber tubes was

312
 Public Address System

longer than the other, a time delay would be introduced to one ear, pro-
viding a binaural effect. The effect was certainly audible but was hardly
binaural. Then, also in 1925, the Kluth system from Berlin used an elec-
tric network to change the phase response in one ear, providing the lis-
tener with plastic radio, as it was touted.
Then, the advent of commercial stereo recordings caused another urry
in the pseudostereo camp, avowedly for the purpose of updating the now-
obsolete monaural recordings, which were so numerous. The stereo-
phoner of Hermann Scherchen was one result, as was the 3D converter
of Chernov. Holger Lauridsen of the Danish National Broadcasting System
also worked on a pseudostereo system, as did Paul Weathers in the United
States. All these systems rely on short time delays and/or phase shifts and
equalization to achieve their effects.
Another idea for improved spaciousness in monaural recordings was
the addition of synthetic reverberation, and one of the earliest devices
for this purpose was the xophonic built by Radio Craftsmen in the mid-
1950s. The Xophonic contained a small loudspeaker connected to a
coiled tube with a microphone in the other end. The delayed sound was
then amplied and reproduced through another loudspeaker in the same
cabinet. Used in conjunction with a conventional single-loudspeaker sound
system, the Xophonic added a reverberant sound from another location.
The coiled pipe had many resonances, and the frequency response of the
system was far from uniform. The Xophonic died in the marketplace soon
after its introduction. Nevertheless, there were several other synthetic
reverberation devices soon placed on the market for home and automobile
use, most of them using the spring reverberator patented in the 1930s by
Laurens Hammond for the Hammond organ.
Psophometer A specialized instrument that measures noise in telephone
circuits, radio transmissions, and other audio-frequency communi-
cation systems. The thing that distinguishes it from a simple voltmeter is
the addition of weighting filters, which are designed to allow meas-
urements that correlate well with subjective perception of noise. The ccir
(now ITU-R) formulated the applicable specications for the psophome-
ter, and the rst commercially available psophometer was made by the Brel
& Kjaer company of Denmark.
Psychoacoustics Psychoacoustics is that discipline that treats the subjec-
tive, or psychological, aspects of acoustic phenomena. It is a branch of the
larger eld called psychophysics.
Psychoacousticians investigate such things as the ears ability to local-
ize sounds, the perception of phase shift in audio signals, the sensi-
tivity of the ear to various types and amounts of distortion, etc. The
eld is relatively young, and a great deal remains to be learned in it.
Public Address System An old term, abbreviated PA or P.A., coined in the
1930s, probably by somebody at the Bell Telephone Laboratories, for sound
reinforcement system. The term sound reinforcement was introduced in the
1960s to imply high quality and absence of distortion compared with

313
Puck

the older PA systems, many of which were notorious for unintelligibility.

Of course, perfect intelligibility and absence of audible distortion are not
very common characteristics of sound systems being built today. (We are
frequently bemused by the various clicks, pops, spurious noises, and dis-
tortions present in the sound systems used in the lecture rooms at the
national conventions of the Audio Engineering Society. Where would one
expect to go to hear state-of-the-art sound systems if not at the aes meet-
ings?) The term PA still seems to be preferred, especially among perform-
ers, even when referring to the most elaborate and complex sound
reinforcement systems. Anyhow, it is easier to say.
Puck In a tape recorder, the tape is pressed against the capstan by a rub-
ber wheel called a puck (or pinch roller). Possibly the name comes from its
resemblance to a hockey puck.
Some record turntables are of the rim drive type, and they use a rub-
ber wheel called a puck to engage and drive the rim at the proper speed.
Pulldown See 3:2 Pulldown.
Pumping Refers to the audible uctuations in the noise level of a signal
caused by poorly adjusted or unsuitable noise reduction systems; also
called noise pumping. Sometimes the term breathing is used synonymously.
quantization noise can also exhibit pumping.
Punch Block A type of terminal strip, rst used by the telephone com-
panies. It uses small, metal, V-shaped terminals into which the wire is sim-
ply pressed with a special tool. No screws are used, and the wire insulation
does not need to be stripped rst, for the V groove cuts into the wire. Wiring
up a punch block is much faster than making connections to a screw-type
terminal strip. Connectors not requiring wire stripping are generally called
IDCs, for Insulation Displacement Connectors.
Punch In The precise control of the onset of tape recording on one or more
tracks in the midst of an already existing recording. For instance, a uffed
word or phrase by an announcer can be corrected by listening to the play-
back and punching in at the exact moment. Punching in requires consid-
erable skill and also a tape recorder designed for it. See also purc. Today,
it is naturally much easier to do this on a digital editor.
PURC Pick Up Record Capability. This is a term coined by the Ampex com-
pany in the 1960s for the ability of a tape recorder to do punch-in recording.
Pure Tone A sound whose waveform is a sine wave; or a signal with
a single frequency and no harmonics. Pure tones have been used for
much psychoacoustic research, and some of the phenomena discovered
by using them do not apply to complex tones as are found in musical instru-
ments or elsewhere in nature. For instance, the pitch of a pure tone is found
to vary with its amplitude, but this is only slightly true for complex tones.
It is also difcult to apply the information from masking experiments using
pure tones to real world complex tones.
Push-Pull Push-pull is a type of amplifier that originated in the days of
tube-type power amplifiers and was patented by E. F. W. Alexander-
son of General Electric in 1913. In a push-pull amplier, there are two out-

314
 PZM

put tubes connected in such a way that while the current in one is increas-
ing, it is decreasing in the other. In a sense, one tube pushes and the other
pulls the current. The signals from the two tubes are combined in the out-
put transformer. The solid-state analog of the tube push-pull circuit
is the complementary symmetry circuit, where two transistors oper-
ate in a similar manner.
There are several classes of operation of push-pull circuits. In class a,
both devices carry signal all the time, and there is no transition of the sig-
nal from one device to the other. This conguration produces relatively lit-
tle distortion. In class b operation, one device is turned completely off
during part of the cycle, and the other one carries the entire signal. The sig-
nal is thus traded back and forth between the tubes or transistors as it goes
through the zero voltage point. This produces more distortion, but is more
efcient, permitting more audio power output from the given devices than
class A operation. To reduce the distortion in class B operation, relatively
large amounts of negative feedback are used. The different classes of
operation are obtained by the amount of bias applied to the stages.
Because push-pull circuits are symmetrical with respect to the signal volt-
age and current, the distortion they generate has a symmetrical waveform,
which means it consists only of odd-numbered harmonics. One difculty
in designing such circuits is to ensure that the signal is smoothly passed
from one device to the other as it goes through zero. Otherwise, crossover
distortion is introduced, which is worse at low signal levels, where it
represents a larger proportion of the signal. See also harmonic distor-
tion; power amplifier.
PWM Pulse Width Modulation. PWM is a method of encoding a signal
by using the length of a pulse as a measure of the height of a sample of the
waveform. PWM is not a digital encoding at all, but is all analog. Some
digital encoding systems use PWM as an intermediate stage between sam-
pling and analog-to-digital conversion. The video laserdisc system
uses PWM coding of the video signal, and thus is not a digital system.
PZM A registered trademark for a type of commercial microphone,
PZM stands for Pressure Zone Microphone. The PZM is a small electret
condenser microphone that is mounted very close to a small aluminum
plate. It is meant to be placed on a large at surface such as the oor or a
wall. The PZM is 6 decibels more sensitive than the same microphone
would be if it were in free space. This is due to the pressure doubling, which
occurs at a boundary when a sound wave is reected from it.
This increase in sensitivity is an advantage because it increases the
signal-to-noise ratio of the microphone by the same amount. How-
ever, the primary advantage of a microphone at a surface is that it is not
sensitive to phase cancellation of sound reected from the surface and inter-
fering with the direct sound. A free-standing microphone should never be
placed closer than 5 or 6 feet from a surface to avoid this effect, which is a
broad dip in the frequency response curve at a frequency whose wave-
length is twice the distance to the surface. Thus, a microphone 1 foot from

315
Q

a surface will have a dip centered at about 500 hertz, which is right in the
middle of the audible range.
The PZM has hemispherical coverage if mounted on a surface large com-
pared to the longest wavelength to be picked up. If it is mounted on a small
surface, such as a square piece of plexiglass, the sensitivity will drop by 6
dB at low frequencies because the sound wave is simply diffracted around
the barrier rather than being reected from it. This results in an apparent
treble boost.
PZMs are also somewhat sensitive to the vibration of the surface they
are mounted on, causing nonuniform frequency response. A good close-
miked piano sound can be obtained by placing such a microphone on the
underside of a raised grand piano lid. (PZMs are excellent at picking up
the sound of tap dancing.)
Although PZM is a recently introduced term, the technique of placing
microphones very close to surfaces has been used for many years, espe-
cially by laying microphones on the lip of a stage for unobtrusive rein-
forcement of musicals, etc. If a cardioid microphone is laid on a oor, its
directional pattern will be preserved, but with 6 dB greater gain. The true
PZM always has a hemispherical pickup pattern and 6 dB higher gain than
the microphone alone would have.

Q
Q In reference to a resonant mechanical or electrical circuit or a capac-
itor, Q stands for quality factor. In the case of a resonant system, Q is a
measure of the sharpness of the resonant peak in the frequency response
of the system and is inversely proportional to the damping in the system.
equalizers that contain resonant circuits are rated by their Q value: the
higher the Q, the higher and more well-dened the peak in the response.
In a capacitor, Q is a measure of its efciency and is dened as the ratio of
its capacitive reactance to its resistance at some specied high fre-
quency. The Q of a capacitor is also called the power factor, and is the recip-
rocal of the dissipation factor.
In loudspeaker systems, Q is a measure of the directivity of the sound
output. A Q of 1 means the system radiates energy equally in all directions,
or into 360 degrees of solid angle. A Q of 2 means the speaker radiates only
into a hemisphere, or 180 degrees of solid angle. Higher values of Q mean
the speaker radiates into smaller and smaller angles, or in other words
becomes more and more directional. Q is an important characteristic of
sound reinforcement loudspeakers because the more directional the speak-
ers, the less room reverberation will be excited by the sound system,
increasing the clarity of the perceived sound.
Unfortunately, the Q of a speaker depends on frequency. It always
increases at higher frequencies and is always near 1 at the lowest frequen-

316
 Quadraphonic Sound

cies. This must be carefully considered by the sound system designer. See
also directivity index.
QAVC Quiet Automatic Volume Control. A type of agc introduced about
1934 in radio receivers that reduced the sensitivity nearly to zero when no
carrier was received, resulting in the squelching of interstation noise.
Q-Biphonic See biphonic.
QS See quadraphonic sound.
Quadlet A 32-bit digital data quantity.
Quad Microphone Cable A superior type of microphone cable consisting
of four signal conductors and a shield. The four signal wires are twisted
together and opposite wires are connected together at each end of the cable,
giving a double twisted pair conguration. Quad cable has signicantly
better rejection of 60-Hz hum interference than conventional twisted pair
cables.
Quadraphonic Sound A stereophonic system consisting of four chan-
nels, usually two in the front as in conventional two-channel stereo, and
two in the rear, purportedly for the reproduction of reverberant sound.
Quadraphonic sound, or quad, enjoyed a brief period of popularity in
the early 1970s and has experienced a gradual but inevitable death since
then. The reason for its demise is that it did not provide the listener a con-
vincing illusion of being immersed in a reverberant eld; in short, it did
not work. Perhaps any system that mixes Latin and Greek to invent its name
is doomed from the start.
Throughout the history of audio, there have been many techniques put
forth that attempt to acoustically transport the listener to the room where
the recording was made. Even in the earliest days of the cylinder phono-
graph, multiple reproducers were sometimes used to introduce small time
delays when playing a record in an effort to produce greater reverbera-
tion. Articial reverberation has been added to recordings for many
years, and synthetic reverberation devices have been promoted for home
use to enhance the spaciousness of recordings. It has long been known
that two channels are not sufcient to convey a realistic sound perspective
to a listener-to-loudspeaker reproduction. Three-channel stereo, for instance,
is much better at simulating the positions of instruments across a stage. See
also stereophonic.
The advent of home-type, four-channel tape recorders helped to encour-
age the industry to produce four-channel recordings, and this was proba-
bly responsible for the birth of quadraphonic sound. Then someone
discovered that four channels of audio could be combined in a special way
(called matrixing) and recorded onto a two-channel disc. The composite
signal, containing the four channels, is fed into a de-matrixing network,
which recovers a reasonable facsimile of the original four channels. Thus,
the matrixed format of recording quadraphony on discs was born. The ini-
tials QS and SQ referred to competing schemes, each requiring its own type
of decoder, which the hapless consumer was expected to buy. Then, a multi-
plexing technique that used a high-frequency carrier modulated with

317
Quadrature

the two rear channels and recorded on the disc along with the two stan-
dard stereo channels was born. This was called CD-4 and was advertised
as discrete rather than matrixed quadraphony. Of course, a special de-
modulator was required to play it, not to mention a special cartridge and
stylus assembly. The CD-4 system was too expensive and suffered from
far too much background noise and distortion to be considered high-
fidelity, and it was short-lived in the marketplace.
Probably, if P. T. Barnum had lived in the 1970s, he would have promoted
a quadraphonic technique.
All these systems were rushed to the market, very often without suf-
cient engineering having been done. It was as if the manufacturers were
using the public as a testing agency for the technology, constantly coming
out with new improved models and expecting to sell them. Finally, the
long-suffering public gave up in disgust. Who was it who said you cannot
fool all the people all the time?
However, it must be said that todays multichannel surround sound sys-
tems found in most motion picture houses and many home theater setups
are really a logical outgrowth of the now obsolete quadraphonic systems
of the past. Of course, the systems of today are far more sophisticated and
the construction of the multichannel sound tracks used with them is an art
form in itself.
Quadrature Two signals which are 90 degrees out of phase with one
another are said to be in quadrature. Also, a signal or a function such as
impedance will have a phase angle that varies with frequency or with
time. This phase angle can be resolved into two components called the real
and imaginary parts, which have a 90-degree phase difference. They are
said to be in quadrature; or more commonly, the imaginary component is
called the quadrature part. See also Appendix 11.
Quad Track Optical soundtrack negative, and release print made therefrom,
which contains all three digital sound formats (dolby digital, DTS, and
SDDS) plus a standard sva analog track.23
Quantization The representation of a continuous voltage span by a num-
ber of discrete values. Quantization is inherent in any digital audio sys-
tem, and it adds quantization error, noise, and distortion to the signal.
The signal after quantization has a staircase shape rather than a con-
tinuous curve, and the difference between this and the original signal is
quantization error. The amount of error will always be within one lsb; there-
fore, the smaller the LSB, the better. In the quantization of a sine wave,
whose frequency is a submultiple of the sampling frequency, the error will
have a denite pattern that repeats at the frequency of the signal. Thus,
it will have a frequency content consisting of multiples of this frequency,
and it can be considered harmonic distortion rather than noise.
For music, however, the signal is constantly changing, and no such reg-

23. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

318
 Rack

ularity exists. The quantization error is then wideband noise, and is called
quantization noise. Quantization noise is difcult to measure because it does
not exist without a signal. A sine test signal is not good because sometimes
this results in distortion, not noise. If the sine wave frequency is chosen so
it is not a submultiple of the sampling frequency, the quantization errors
will be more nearly randomized and will resemble random noise.
Quantization Error See quantization.
Quantizing Step See lsb.
Quarter-Track Quarter-track, sometimes called four-track, refers to most
home-type, reel-to-reel analog tape recorders, which use one-fourth the
width of the tape for each recorded track. This allows stereo signals to be
recorded in both directions, effectively doubling the recording time. The
technique is simply to record two tracks in one pass of the tape, and then
reverse the tape reels and record the other side on another pass of the tape.
Professional stereo tape recorders use one-half of the tape for each track,
which results in better quality and reduced noise level. Quarter-track tapes
with both sides recorded cannot be edited by splicing the tape as can two-
track tapes. Four-track usually refers to an open reel format with all four
tracks recorded in one direction.
Quartz Control See servo.
Quiescent Current In transistor and tube circuitry, the direct bias cur-
rent is sometimes called the quiescent current because it still exists in the
absence of a signal.
Quiescent Noise The residual noise produced by an audio device at its
output terminals when no signal is present.
Quieting The decrease in the noise level at the output of an fm receiver or
tuner when tuned to an unmodulated carrier, as opposed to being tuned
between stations. The quieting, expressed in decibels, is one measure of
the performance of the receiver or tuner.
Quietness
Quintophonic An unusual surround-sound system for motion picture use
devised in Britain by the late John Mosely. It used both magnetic and opti-
cal sound tracks. Quintophonic sound never became a commercial success.
See also colortek.

R
Rack Many years ago, a standard method for mounting telephone equip-
ment was devised that used steel frames spaced 19 inches apart. They were
perforated with many tapped holes along the vertical axis, and equipment
with flanges 19 inches apart could be bolted in place. These frames were
originally called relay racks. The same design is still widely used to mount
commercial electronic equipment, including audio components. Even
some home-type components such as amplifiers and tuners are designed

319
Rack Space

for rack mounting. Some components are adapted for rack mounting by
the addition of rack anges, or rack ears.
Rack Space See u.
Radiation Impedance The acoustic impedance that acts as a load on a
loudspeaker, opposing the motion of the cone. The acoustic power out-
put of a speaker, especially a low-frequency speaker, is greatly affected
by the radiation impedance, and this depends on where the speaker is
placed. In free space, with no nearby reecting surfaces, the speaker will
see the highest impedance and will radiate the least amount of power. If
placed next to a large wall, it radiates into a hemisphere rather than a full
sphere, cutting its radiation impedance in half and doubling its power out-
put. The more restricted the volume the speaker sees, the more power it
will radiate. This is the reason a loudspeaker system has such an increase
in bass response when placed in the corner of a room. The radiation imped-
ance is also reected back to the electrical input impedance of the speaker
terminals, so the amount of power the speaker draws from the amplier
also varies with the speakers location.
Radiation Pattern The graph of a loudspeakers directional characteris-
tics, usually plotted for several specic test frequencies. Also called the polar
pattern. The term radiation pattern also applies to sound radiation of musi-
cal instruments.
Radio Frequency, RF An alternating current or voltage having a
frequency above about 100 kilohertz. It is so called because these fre-
quencies are radiated as electromagnetic waves by radio (and now televi-
sion) stations. Theoretically, RF could extend down into the audible range,
and some special-purpose radio transmissions are at about 20 kHz, but RF
is usually considered to be at least 100 kHz.
In radio and television broadcasting, the actual signal broadcast by the
station is called the radio frequency, as opposed to the audio frequency (AF)
or video frequency, which modulates it. The signal present at the receiv-
ing antenna terminals is also RF and remains so in the receiver until it is
heterodyned or detected.
Radio Frequency Interference, RFI The noise that can be induced in audio
systems due to radio and television broadcasting stations is called radio
frequency interference. One might think that the extremely high fre-
quencies radiated by broadcast stations would not cause problems at audio
frequencies, but most audio circuits are nonlinear at these high fre-
quencies, and this causes the interfering signals to be rectied, or
detected. The envelope of the RF signals can then be heard as the inter-
ference. RFI is a type of electromagnetic interference. See also
television interference.
Radiogram British terminology for radio-phonograph.
Radiophonics A term, coined by the BBC, referring to a type of electronic
music made up of multiple recordings of natural sounds that have been
copied and manipulated in various ways and spliced back together to make
a collage of sound. The more usual term for this is musique concrte.

320
 Rayleigh, Lord

Rails See supply rails.

Random Incidence Correlator A mechanical accessory placed over an omni-
directional measurement microphone to improve its performance as a ran-
dom incidence microphone.
Random Incidence Microphone An omnidirectional microphone hav-
ing essentially a constant sensitivity over a wide frequency range for
sounds that arrive at the microphone from all directions with equal prob-
ability. Such a microphone will have a rising high-frequency sensitivity for
sounds coming directly from the front. For recording a-b stereo in a large
auditorium, it is a good idea to use random incidence microphones, since
the increased frontal high-frequency sensitivity tends to compensate for the
high-frequency loss due to air absorption.
Raster The characteristic pattern of horizontal lines formed by the scan-
ning beam of a television or computer monitor picture tube, or the result-
ing image.
RASTI Short for Rapid Speech Transmission Index, which is a technique
for evaluating the intelligibility of speech in a given room. It is an objective
measure that correlates well with the subjective articulation index of speech.
The RASTI system consists of a box that produces two octave bands
of pink noise that are amplitude modulated at frequencies corre-
sponding to the uctuations in level of human speech.
This signal is projected into a room at a spot where a talker would nor-
mally be placed. Then, a microphone and analyzer pick up the signal at
a location where a listener would be. Room noise and reverberation of
the signal will reduce the amount of amplitude modulation. The analyzer
translates this reduction in modulation into a number from 0 to 1, which
is proportional to speech intelligibility.
The value of the system lies in the simplicity and rapidity with which
many tests can be made in a building. It is useful for sound system evalu-
ation, as well as for examination of acoustics.
Rated Bandwidth The frequency range, normally taken to be from 20
hertz to 20 kilohertz for audio equipment, over which the performance
of a device is rated with respect to characteristics such as power output,
distortion, etc.
Rated Characteristic A single value of a characteristic, such as power out-
put, distortion, etc., claimed as representative of the performance of the
device and measured in conformance with standardized test procedures.
Rated Load The load impedance into which a power amplifier is
designed to operate, and upon which other characteristics such as power
output, distortion, etc., are based.
Ratio Detector A type of detector used in fm radios that is sensitive only
to frequency modulation, as opposed to the frequency discriminator,
which is also sensitive to amplitude modulation. Most of the interference
heard by someone listening to FM is caused by amplitude modulation of
the carrier. The ratio detector is insensitive to this type of interference.
Rayleigh, Lord Lord Rayleigh (John William Strutt, 18421919), an English

321
RC

mathematician and physicist, is considered the father of modern

acoustics. He won a Nobel physics prize for the discovery of argon gas
and published a monumental two-volume treatise titled The Theory of Sound,
which is still the denitive work on the subject.
RC The magnitude of the time constant of a series resistance-
capacitance circuit is the resistance in ohms multiplied by the capaci-
tance in farads.
RCA Plug, RCA Jack The most common type of audio connector, next to
the standard phone plug, is the phono connector designed by RCA many
years ago. The input and output terminals on almost all audio pream-
plifiers are RCA jacks. The same jack is used for most consumer composite
video and S/PDIF connections, although the cable for such should have
75 ohms characteristic impedance. They have the advantage of small size,
shielded conguration, and low price, but the disadvantages of unreliability
and lack of ruggedness. If you step on an RCA plug, you can usually for-
get about using it again; and with continued use, the outer ground con-
nection is likely to lose its springiness and thereafter make poor contact,
resulting in hum and static.
R-DAT Original name for the common audio dat machine meaning Rotat-
ing-head digital Audio Tape. Since most common DAT machines are the
rotating head type and s-dat (stationary DAT) machines are referred to
by format (DASH, ProDigi), the term R-DAT has fallen into disuse.
Reactance See capacitive reactance; inductive reactance.
Reactive The term reactive describes an electronic component in an elec-
tronic circuit that has either capacitive reactance or inductive
reactance or both. Nonreactive components are sometimes called resis-
tive. See capacitor and inductor.
Read Only Memory A digital storage medium where the data cannot be
erased or recorded. The original audio compact disc was a type of read
only memory, as was the CD-ROM. ROMs are also silicon chips or mod-
ules that are an integral part of the computer, containing starting code that
never will be changed.
Real-Time Analyzer A special type of spectrum analyzer, which con-
sists of a group of bandpass filters, each having a constant percentage
bandwidth, such as one octave or one-third octave. All the filter
inputs are connected to the input signal, and the outputs of each lter
are passed through a detector and thence to an indicating device such
as a CRT or, more commonly, an array of LEDs or an LCD. The display is
a graph of amplitude of the signal versus frequency, which is a spec-
trum. The amplitude in each frequency band is almost always expressed
in decibels.
The real-time analyzer is so called because it does the entire frequency
analysis simultaneously, or in real time, as opposed to sweeping through
the frequency range with a single lter. The real-time analyzer lacks reso-
lution because the lters are too wide to separate narrowly spaced frequency

322
 Recording Lathe

components, but it is useful to indicate how a signal will sound subjectively,

especially if the analyzer is a one-third octave type.
Real-Time Dubbing Duplicating an analog tape at its normal playing speed
rather than at a higher speed. Real-time duplicating is generally of better
quality than high-speed duplicating.
Receiver An audio component consisting of an AM-FM or FM-only tuner
with a preamp and power amplifier all on one chassis. Nowadays
receivers tend to be home theater oriented, with multiple amps (e.g., 6),
several surround decoders, and AV switching as well as the aforementioned.
Recone To replace the cone of a dynamic loudspeaker. This involves
replacing the voice coil, surround and spider assemblies also. The
magnet structure of a typical loudspeaker is the most expensive part, and
this makes reconing of damaged units a practical proposition. When recon-
ing a speaker, the original response characteristics will be regained only if
the new parts used are the same as the originals!
Recording Horn A special horn used to collect the sound to be recorded
in the making of acoustic recordings. The recording horn somewhat
resembled an ear trumpet and was actually an acoustic transformer, match-
ing the acoustic impedance of the air to that of the recording mechanism
diaphragm. It thus greatly increased the efciency of the recording process.
Edisons rst phonograph had a rudimentary short horn that was not very
effective. It is said that one had to shout directly into it in order to make a
playable recording. Edison quickly realized the importance of the horn, how-
ever, and he performed endless experiments to determine the best shapes
and sizes. One Edison recording horn was reportedly 40 feet long.
It is interesting that all these early horns were conical in shape, and it
was not until the mid-1920s that workers at the Bell Telephone Labs devel-
oped the mathematical theory of horns in general and determined that the
exponential horn is much more efcient than the conical.

Recording Horn

Recording Lathe The recording lathe is the turntable assembly that is used
in the recording of the acetate master, which is the rst step in the mak-

323
Recording Lathe

ing of phonograph records. The lathe is more than a turntable, however. It

also includes the mechanism that advances the cutterhead across the
record as it turns, and in so doing determines the spacing of the grooves
or the pitch of the record. The lathe must be accurate and stable in its
motion, for any variations will show up in the nal record. This means that
the amounts of flutter and wow are very low, and the speed is very closely
controlled.
Early recording lathes were driven by weights and mechanical speed
governors because electric motors could not be made to have as stable a
speed. In the 1950s, the RCA Victor Record Company in Camden, New Jer-
sey, reportedly bought a manhole cover from the city of Camden and
machined it into a turntable for a recording lathe in order to attain sufcient
mass to ensure constant speed and negligible speed variations.
Although Edison chose 160 rpm as the speed of his 2-minute cylinder
records and 80 rpm for his 4-minute cylinders and Diamond Disc records,
the speed of 78.26 rpm was standardized about 1925 and used for many
years before the 3313-rpm speed was adopted for microgroove records.
You might well ask why the strange speed of 78.26 rpm was chosen. The
various record manufacturers in the 1920s did not agree on the optimum
speed for records to turn, and recording speeds varied from about 82 to
78 rpm. It was not considered a problem, because all phonographs were
variable speed, and the listener simply adjusted each record to suit his own
taste. In 1925, with the advent of electrical recording technology by the
Bell Telephone Labs, it became necessary to standardize the speed. Victor
Records was using 78 rpm at the time, and they were the largest record
company of the day. Victor was also the rst customer for the new elec-
trical recording apparatus, so the Bell people settled on 78 rpm. However,
it turns out that it is impossible to attain precisely 78 rpm with a syn-
chronous motor driving a turntable through a worm gear. The worm gear
is a very quiet drive mechanism compared to other types of gears. With a
synchronous motor running at 3,600 rpm and a gear box with 46 to 1 ratio,
the resulting speed is 78.26087. . . rpm, which is as close to 78 as possible
with a gear drive.
The 3313-rpm speed also came about in 1925, this time through the efforts
of the Western Electric Company, Inc. (a subsidiary of Bell Telephone). The
rst sound motion pictures had the sound track on a disc that was played
in synchronization with the movie projector. This was called the Vitaphone
system. A 1,000-foot reel of 35-mm motion picture lm ran for 11 minutes,
while the 10-inch standard 78-rpm record played for 3 minutes. In order
to get 11 minutes playing time, a speed of about one-half of 78 rpm was
sought. A worm gear of 54 teeth provides 3313-rpm at a 60-hertz line fre-
quency driving a synchronous motor. Then the record diameter had to be
increased to 20 inches to allow the needed time.
The Vitaphone system was used only a short time and was replaced by
recording the sound track directly on the lm, but the 3313 rpm format was
retained for the purpose of recording radio broadcasts for delayed trans-

324
 Re-entrant Horn

mission. These transcriptions were used until the advent of the tape
recorder in the early 1950s.
A third speed of 45 rpm was introduced about 1948 by RCA for their
new microgroove records. One wag has suggested that the speed was a log-
ical choice, since 78 less 33 equals 45.
Recordist A person (presumably) procient in the art of making record-
ings of music. The person who operates the recording device during a
recording session, who may also be called the machine operator.
Rectication The conversion of an alternating current to a direct
current. The AC of the power line is rectied in the power supply to pro-
vide DC for active circuit devices. Rectication is also used to recover
the signal from an amplitude modulated waveform. See also detec-
tion; detector.
Rectify See rectification.
Red Book A comprehensive manual published by Philips and Sony that
sets out the complete standards dening the compact disc format. When
buying a license to make CD players, a manufacturer obtains the rights to
the applicable patents and also a copy of the Red Book. This ensures that
all CD players will be compatible.
In like manner, the cd-rom standards are contained in the Yellow Book,
also published by Philips and Sony, but the le structure is not the same as
in the Red Book CD. The CD-ROM le structure conforms to the ISO 9660
specications, and this allows the CD-ROM to be read by many computer
operating systems. The Orange Book covers CD-R. The Blue Book is the
specication for CD-Plus, a multisession audio and data disc. The Green
Book covers the cd-i standards and the White Book covers Video CD. The
colors ostensibly come from the colors of the covers on the books.
Redundancy The transmission of more information (in a digital system,
this means more bits) than is necessary in order to improve the reliability
of the transmission. It must, of course, be accompanied by appropriate
decoding. error correction bits are examples of redundant information.
Human speech is highly redundant by nature. This is the reason it is
possible to understand what is said in the presence of large amounts of
noise or distortion. Music is also very redundant.
Reel Idler A rotating metal cylinder between the supply reel and the heads
of a magnetic tape recorder. It is usually connected to a ywheel under the
tape deck and thus serves to smooth out any irregularities in the tape motion
caused by tape sticking in the supply pack.
Re-entrant Horn A type of horn loudspeaker, usually made of metal,
in which the sound path is folded back on itself to make the overall horn
less bulky. Re-entrant horns (or re-entrant trumpets, as they are sometimes
called) are characterized by poor low-frequency response, poor very high-
frequency response, and many resonances in the audible range, making
them sound metallic or tinny. Their sole advantage is high efciency,
producing a lot of sound for a little input power. They are used only for
low-cost paging systems. See also folded horn.

325
Reference Level

Re-entrant Horn

Reference Level The reference level in an audio device is a signal level near
the maximum possible that the device can handle but low enough to ensure
low distortion. This difference is known as headroom. In audio devices,
it is desirable to attain the maximum signal-to-noise ratio, and to do
this, one needs to know the signal level in the device in comparison to the
maximum level possible without distortion. If the operator keeps the sig-
nal level as high as possible, the signal to noise ratio is maximized. Many
devices will show the reference level as 0 dB with an LED or on a mechan-
ical meter, and mark the onset of the maximum signal possible with a sep-
arate LED labeled clip or overload. Meters are usually inadequate to
show the maximum with any accuracy on program material because they
are too slow in response. It is up to the operator to determine if factors such
as the metering characteristics, or crest factor, dynamic range and
spectral content of the audio signal warrant adjusting the signal level above
or below the reference level. Operators may also interconnect devices and
use each devices reference level indication to estimate these optimal oper-
ating conditions. Unfortunately, devices seldom agree on the desirable head-
room, and this must be taken into account. See gain structure. Headroom
in many devices may be 24 dB or more, but you can get better signal-to-
noise ratio numbers by reducing headroom.
In the case of digital devices, the important reference level is the maxi-
mum, known as 0 dBFS for decibels full scale. Program should be recorded
to come as close to 0 as possible without ever going over. This is difcult
to achieve in practice, as even peak reading meters cannot always display
a brief level that is actually going over 0. Again, the operator must judge
how far below 0 dBFS to adjust her program level to avoid ever going over
and causing distortion. Some controversy exists over how many dB below
0 dBFS to record test tones intended to help subsequent analog equipment

326
 Relap

users set their levels. An accepted gure is 1218 dB, but since there is no
standard, a notation should be made for the benet of the next user. Any
decibel measurement requires a reference level that is 0 dB by denition.
This reference level is typically not chosen for any reason involving opti-
mizing signal to noise ratio as above.
In the case of analog tape recorders the term operating level is used.
Reex An early type of radio set that passed the rf signal through several
amplier tubes and then passed the audio signal after detection through
the same amplier tubes again. In effect each tube does the work of two.
Such sets usually had only one tuned circuit and were not very selective,
that is, they were likely to receive several strong stations at the same time.
See also bass reflex.
Regeneration A type of positive feedback in radio receivers invented and
patented in 1913 by Major Edwin Armstrong. Lee DeForest contested the
patent, claiming to have invented the same thing in 1912, although he didnt
publicize it. There was a great court battle, and several decisions each way
until nally the U.S. Supreme Court ruled in DeForests favor and awarded
him the patent in 1934.
The purpose of regeneration was to increase the gain of the RF ampli-
fier in the receiver and thus increase the sensitivity of the receiver. Arm-
strongs patent used a tickler coil inside the main tuning coil and con-
nected to the RF amplier. The tickler could be rotated by a knob, varying
the coupling between it and the coil and thus varying the feedback and the
resultant gain of the stage. Too much regeneration would cause the set to
oscillate, turning it into a miniature transmitter, to the consternation of the
neighbors who might be trying to listen to their radios! One method of oper-
ation, called dead beat tuning, was to increase the regeneration to get the
set to howl and then tune to a station. As the station was tuned in, the howl
frequency would fall, and there would be a bloop sound when it was
tuned in exactly. For this reason, these sets were also called bloopers.
Other methods of adding positive feedback were developed by differ-
ent makers in order to get around Armstrongs patent, but most were infe-
rior to the tickler coil. At least one maker built a set with a tickler coil that
was short-circuited by a heavy wire. Attached to the wire was a large red
tag cautioning the purchaser not to remove the wire under any circum-
stances, for this would result in patent infringement. However, the tag went
on to extol the improved reception that would result if the wire were
removed!
Regulation The ability of a power supply to maintain a constant volt-
age as varying current is being drawn from it. A regulated power sup-
ply is one that has a circuit specially designed to control the output voltage
as the current drawn varies. A conventional power supply, without a reg-
ulator circuit, can have good regulation, in which case it may be called a
brute force supply.
Relap To reshape analog tape recorder heads after they wear. For proper

327
Release Agent

operation, a tape head should have a smoothly curved surface where the
tape makes contact with it. As the head wears with use, the surface will
develop a flat, and this reduces the pressure of the tape against the gap
of the head, causing nonuniformity of the output.
The head can be restored to original performance by reshaping the curve
by a process called relapping. This is another word for a very ne-grain
grinding and polishing operation. Usually this can only be done once
because the pole pieces in the head will wear too thin and the gap will
increase in length. See also edge slotting.
Release Agent A chemical additive to the vinyl compound used for
pressing vinyl phonograph records to prevent them from sticking to the
stampers.
Remote A recording session that occurs at a location other than a record-
ing studio, such as a concert hall or church.
Most classical music recordings are remotes, for the acoustics of the
recording space are an important part of the musical experience. A remote
recording requires a good deal of planning and a large amount of work com-
pared to a studio recording, and unforeseen problems are always cropping
up. An example of such a problem occurred in Seattle about forty years
ago when a record company set out to record a large theater organ. micro-
phone cables were strung, all the equipment was brought in and set up,
and microphones were positioned for the rst test recording. The techni-
cian, needing AC power, pulled the end of his long extension cord back-
stage and plugged it into what looked like a standard power outlet. The
problem was that the theater was powered with DC rather than AC, and
all the equipment instantly failed, some by blown fuses and some by burned-
out power transformers.
Repeat Coil, Repeater Coil Another name for a one-to-one turns ratio trans-
former used in telephone and broadcasting applications in order to eliminate
the requirement for the ground connection between two audio devices
or systems. ground loops are thus eliminated, and certain noises such as
hum can be greatly reduced.
The repeat coil is typically designed for input and output impedances
of 600 ohms. It is also sometimes called an isolation transformer,
although some isolation transformers are designed for use in the power line
input of audio devices. These isolation transformers, however, are never
called repeat coils.
Reproduce Alignment Tape See alignment tape.
Resampling In digital audio systems, the changing of a signal encoded
at one sampling rate to a different sampling rate is called resampling. The
brute force way of doing this would be to convert the signal back to ana-
log with a dac and then re-encode it with an adc using a different sam-
pling rate. This, however, subjects the signal to distortions in the dual
conversion processes, so it is much better to do the job completely in the
digital domain.
If the two sampling rates are in a simple arithmetic ratio, it is relatively

328
 Resonance

straightforward to do this, but it gets much more complicated, although it

can still be done, when the two frequencies have a ratio of any real num-
ber. Because the compact disc system uses 44.1 kilohertz as a sampling
frequency, and the commonly used standard for master recording is 48
kHz, it can be seen that the resampling converter is a very useful device.
Also known as sample rate conversion.
Residue pitch The perceived pitch of a complex tone. The residue pitch in
many cases corresponds to a frequency that is not present in the com-
plex tone. An example of this is the pitch of a bell that is one octave higher
than the lowest frequency component in the tone.
Resist The coating applied to an unetched printed circuit board in the
places where the copper is to remain. The rest of the copper is then dis-
solved away in the etching process. See also printed circuit board.
Resistance Resistance is that characteristic of electric conductors which
resists, or opposes, electric current. The unit of resistance is the ohm (sym-
bol , the Greek letter Omega), after Georg Ohm, who formulated ohms
law. The reciprocal of resistance is conductance, which is measured in
units that used to be called mhos, but today are called siemens, after Ernst
von Siemens. Conductance is seldom used in practical applications. See also
impedance.
Resistor An electronic circuit component that has resistance but no appre-
ciable inductance or capacitance. Resistors are the most common type
of circuit elements, and they are available in many different standardized
resistance values. When an electric current exists in a resistor, a certain
amount of random noise is added to the current, and different types of resis-
tors contribute different amounts of noise. Carbon composition resistors
are very inexpensive but are notorious for being noisy, whereas metal lm
and wire wound resistors are very quiet. There is a theoretical lower lim-
iting noise that any resistor can produce, and this is the thermal noise, or
johnson noise, which is proportional to the resistance itself. Resistors
are made in different sizes to handle different amounts of power. The resist-
ance and tolerance values of the traditional axial resistors are generally not
printed on the unit but are encoded by painted-on colored bands using an
internationally agreed-upon code. The rst two bands encode the rst two
digits of the resistance value, and the third band represents the multiplier
in powers of ten, or simply the number of zeros to be appended to the rst
two digits. Surface mount and other forms will usually have the value
printed in code.
Resonance A resonance is the tendency of a mechanical or electrical sys-
tem to vibrate at a certain frequency when excited by an external force,
and to keep vibrating after the excitation is removed. A bell is a good exam-
ple. The Latin root of the word means to sound again.
All mechanical structures have resonances at many frequencies. Reso-
nances are most troublesome in audio transducers, where they color
the sound by adding their own natural frequencies. loudspeakers are
offenders to a large degree. Ideally, one would design structures of trans-

329
Resonator

Resistor

ducers so the resonances are outside the audible range, and this is done
in the case of most microphones. It is not possible, however, for loud-
speakers, so the resonances are damped so that they will contribute as lit-
tle distortion as possible. Resonances in tweeters are probably the most
insidious because the higher frequencies are where the ear is most sensi-
tive to tone color, or timbre, changes.
Some other mechanical resonances that cause problems are associated
with record players. The resonance between the tonearm mass and the sty-
lus stiffness can cause low-frequency distortion and mistracking.
Resonator An acoustic device that has a resonance. The most common
acoustic resonator is the helmholtz resonator. Virtually all musical
instruments have some sort of resonator as part of their tone-producing
mechanisms.
Return See send.
Reverb Chamber Short for reverberation chamber or room. A specially

330
 Reverberation

designed room whose surfaces are all carefully designed with regard to
placement and reectivity, into which a speaker and microphone are
placed for the purpose of creating controlled reverberation for adding to
recordings. Dry recordings are played through the speaker and the micro-
phone picks up the result created by the room, which can be judiciously
mixed back into the original dry recording. The reverberation chamber pre-
ceded purely electronic devices, and is thought by some to still sound the
best. Capitol Records in Hollywood has several famous reverb chambers
under their parking lot. Curiously, a concrete stairwell can sometimes serve
as a decent reverb chamber with the addition of the speaker and micro-
phone, so long as no one uses the stairs.
Reverberant Field In a room with reverberation, if a listener is close to
a source of sound, the direct sound will predominate, and the listener is
said to be in the direct eld of the source. At greater distances, the rever-
berant energy will predominate, and this region is called the reverberant
eld.
In general, the purpose of sound reinforcement systems is to send direct
energy into the reverberant eld to increase intelligibility.
Reverberation The remainder of sound that exists in a room after the source
of the sound is stopped is called reverberation, sometimes mistakenly called
echo. The time of reverberation is dened as the time it takes for the
sound pressure level to decay to one-millionth of its former value. This
is a 60-decibel reduction in level.
All rooms have some reverberation, and an important subjective qual-
ity of a room is its reverberation time, although other factors, such as ratio
of direct to reverberant sound, are probably more important. In a real room,
the sound heard by a listener is a mixture of direct sound from the source
and reverberant sound from the room. Reverberant sound is diffuse, com-
ing from random directions, and the direct sound allows us to localize the
source of the sound. As we move farther from the source, the direct sound
becomes weaker and the reverberant sound is relatively stronger. At a cer-
tain point, the two will be equal in strength, and this is sometimes called
the critical distance. This distance is surprisingly small for most
rooms. Even when we listen from distances greater than the critical dis-
tance, we can localize the sound because our hearing mechanism can dis-
tinguish the direct sound in the presence of stronger reverberation by
binaural hearing.
For good speech intelligibility, too much reverberation is a hindrance,
and can be considered noise, although some reverberation helps intelli-
gibility if the reverberation time is not too long. The objectively measured
reverberation time of a room is not necessarily heard by a listener, for the
reverberant sound may be high or low in level compared to the direct sound.
For instance, a large cathedral with six seconds reverberation time does not
sound reverberant to two people casually speaking inside it, but their speech
in a small chamber, say twenty feet on a side, with six seconds reverbera-
tion would be inarticulate. In the large space, the reections from the walls

331
Reverberator

are much later in arriving at the listener, and are at a lower level than in
the small rooms. For this reason, longer reverberation time is better toler-
ated in large buildings than in small ones; this also points up the fact that
the reverberation time alone does not give very much information on how
a room will sound to a listener.
Most music recordings have some reverberation recorded on them
along with the direct sound, and this causes a sensation of room ambi-
ence to be present. Usually, recording studios are lacking in reverbera-
tion, and so synthetic reverberation is mixed with the music signal when
making the master recording. This synthetic reverberation seldom sounds
the same as real reverberation, but the music consumer seems to have got-
ten used to it.
Reverberator A device for the generation of synthetic reverberation. See
also plate reverberation; spring reverb.
Reverse Current The current in a diode or other semiconductor junc-
tion that is opposite to the normal direction. Reverse current is sometimes
called leakage current and is usually extremely small, except in the case of
the zener diode.
Reverse Resistance The actual resistance of a diode or other semi-
conductor junction when conducting current in the direction opposite to
normal. Reverse resistance is normally very high, which causes reverse cur-
rent to be very low. But when the reverse voltage reaches the breakdown,
or avalanche, voltage, the resistance suddenly falls and the current increases.
Some diodes have a very nonlinear reverse resistance when operating in
the avalanche mode such that the voltage across them remains essentially
constant regardless of the current. Such a diode is called a zener diode.
Reverse Voltage See reverse resistance.
RF, or R.F. See radio frequency.
RFI See radio frequency interference.
Rhythmicon A clever musical keyboard instrument built in 1930 by Leon
Theremin, n Termen, the inventor of the theremin, at the request of com-
poser/theorist Henry Cowell. Each key of the Rhythmicon played a
repeated tone, proportional in pitch and rhythm to the overtone series
(the second key played one octave higher in pitch [double the frequency
of the rst key] and twice as fast as the rst key. The third key played three
times higher in frequency and repeated three times faster then the rst
key, etc.). The instrument used electronic oscillators to produce the
tones and a rotating disc with rows of holes in it at various spacing. The
rhythm was generated by photoelectric cells on one side of the disc and
neon light bulbs on the other side. When a hole let the neon light strike
the photocell, a sound was produced, and the keyboard determined
which lights were activated.
RIAA Short for Recording Industry Association of America. A long time
ago, the RIAA was active in setting standards for recording and playback
equalization of phonograph records. The universally used de-emphasis
curve for stereo record playback was standardized by the RIAA, and is often

332
 Ringing Out

referred to as the RIAA equalization. In recent years, the RIAA is con-

cerned with protecting musicians and music producers against copyright
infringement by illegal copying and dissemination of original recordings.
They also certify gold record status and present the awards for a gold
record.
Ribbon Microphone See velocity microphone.
Ribbon Tweeter The ribbon tweeter is a high-frequency loudspeaker that
works like a ribbon microphone in reverse. The audio signal is con-
nected to the ribbon, which is a very thin metal (usually aluminum) strip
suspended in a magnetic eld. The current in the ribbon establishes
another magnetic eld, causing the ribbon to move in synchrony with the
input signal waveform. The ribbon is a direct radiator of sound.
The ribbon tweeter, because of its small size, is only effective at very high
frequencies, and must be used with a crossover above 5 kilohertz or
so. It is characterized by uniform output up to ultrasonic frequencies.
A disadvantage is the very low impedance of the ribbon, which usually
requires that a step-down transformer be used with it. There have been some
recent developments of fairly large ribbon tweeters, which are capable of
response down to 1 kHz or so.
Rigging Points Special attachments on a loudspeaker cabinet designed
to safely hang the cabinet. Speakers should not be hung from unapproved
points, such as handles.
Ringing Any device, electronic or mechanical, is said to ring if it contin-
ues to produce a signal or to move after its input is stopped. Ringing is
caused by too little damping, and is particularly bothersome in audio
transducers such as loudspeakers and phono cartridges. Ringing
of low-frequency loudspeakers is called hangover. Ringing, of course,
causes transient distortion, and is to be avoided.
A filter with a very narrow bandpass or a very sharp cutoff will
ring for a certain time if it is presented with a sharp transient. This phe-
nomenon is common and is to be avoided if possible. Unfortunately, the
shape of the slopes of a lter is intimately connected to its inherent damp-
ing, and one cannot be changed without changing the other. This is the rea-
son lters and equalizers should have as gentle a slope as possible,
without extremely steep skirts, especially if they are in the audible range.
anti-aliasing and anti-imaging filters in digital systems have
very steep rolloffs, and they ring when excited by a transient signal. The
frequency at which they ring, however, is ultrasonic and so is not as
bad as one would expect from looking at the picture. See also resonance.
oscilloscope pictures of this ringing used to be published in equipment
reviews. See also ringing out.
Ringing Out One way to adjust a sound reinforcement system to maximize
its gain before feedback is to raise the level of each band of a graphic
equalizer until the system begins to feed back, then to reduce its level by
2 or 3 dB. This usually does not result in an optimally adjusted frequency
response curve of the system but it is widely used in practice.

333
Ring Modulator

Ringing of a Sharp-Cutoff Low-Pass Filter

Ring Modulator A circuit that allows one signal to amplitude modulate

another signal, resulting in a complex series of sum and difference fre-
quencies between the input signal frequencies. The ring modulator intro-
duces a gross nonlinearity and causes huge amounts of harmonic and
intermodulation distortion. It is sometimes used for special effects in the
generation of electronic music, usually in conjunction with a low-pass lter,
which reduces the high-frequency roughness of the resulting sound. It gets
its name from its circuit conguration, which is a ring of four diodes.
Ripping The extraction of audio tracks from cds, or audio/video tracks
from dvds, onto a computer hard drive is called ripping. The process is
more complex than one might expect, since the audio information is
encoded with other data to enable error correction when the CD is
played. Also called Grabbing.
Ripple Irregularities in the frequency response of a filter that has nom-
inally at response in its passband. Irregularity in the value of direct volt-
age from a power supply is also called ripple.
In some lter designs, such as elliptical and chebyshev, ripple in
the passband is traded for steepness of the slope of the response above the
cutoff frequency. Ripple is measured in decibels, and is the maximum
deviation from at response, regardless of frequency.
Power supply ripple is mostly at 120 hertz, but it may contain signicant
harmonics of 120. Ripple is reduced by lters consisting of large capac-
itors and resistors and/or inductors in the power supply. Most power
supplies in audio equipment today are electronically regulated, reducing
ripple even further. Power supply ripple is measured in millivolts rms.
Rise Time The time required for a device to allow a signal waveform to

334
 Router

rise from 10% of maximum amplitude to 90% of maximum amplitude, usu-

ally measured in microseconds. The shorter the rise time of a device, the
higher in frequency its response will extend. To pass a full-level 20-kilohertz
signal, the rise time must be less than 12 microseconds.
RLC Circuit A circuit usually found in equalizers or filters that uses
resistors (Rs), inductors (Ls), and capacitors (Cs).
RMS Root mean square (rms) is, literally, the square root of the mean of
the squares.
To nd the rms value of a signal, one must square the signal mathe-
matically, then average it over a time period, and then take the square root
of the average value. The result is a quantity which, when squared, is pro-
portional to the energy content (or power) of the signal. The rms value of
a sine wave is .707 times the peak value, but it is not usually possible to
deduce the rms value of an arbitrary waveform, especially a complex one
like a music signal. Most common AC voltmeters will not measure the
rms value of any signal except sine waves. A meter that actually does the
job is called a true rms voltmeter.
The term rms power is a misnomer. The rms value of the signal is used
to calculate the average power, but the power is simply the power, in spite
of misleading statements made by some manufacturers. A power ampli-
fier rated as having 200 watts rms power really puts out 200 watts of
continuous sine wave power.
Rock and Roll Editing One way of editing analog tapes is by listening
to the playback as the tape is slowly moved back and forth over the play-
back head to nd the best edit point. To do this, the tape reels are grasped
by hand and rocked back and forth, and the tape is rolled for short dis-
tances between the splice points. Although this is called rock and roll edit-
ing, it has nothing to do with the type of music being edited.
Roll off A filter that has a reduced output as the frequency is increased
or decreased is called a roll off lter. Bass and treble tone controls are roll
off lters when turned down. Sometimes the reduced high-frequency or
low-frequency content of the signal itself is called roll off.
Roll Surround See surround.
ROM See read only memory.
Room Leakage In recording studio jargon, room leakage is unwanted
transmission of sound from one room to the other.
Room Mode, Room Tone See eigentone.
Router A multiple selector switch that can send any of several input sig-
nals to any of several outputs. The equivalent of a patch bay. The actual
switching mechanism in a router is sometimes called a crossbar switch
from telephone exchange usage. A crossbar switch has the input circuits
on a series of bars running at a right angle to a series of bars represent-
ing the output. To send any input to any output requires a connection
between the bars where they cross. Modern routers are almost invariably
electronic, rather than mechanical, switches. Special routers are also used

335
RPM

for switching digital audio and video signals as well as computer network
switching.
RPM Revolutions per minute, the standard unit for measuring rotation
speed.
RTA See real-time analyzer.
Rub and Buzz Distortion A type of distortion found in some dynamic
loudspeakers caused by the voice coil assembly rubbing against one
of the magnetic pole pieces. The detection of rub and buzz is one of the
quality control test procedures used by loudspeaker manufacturers. It is
more likely to occur in high-efciency speakers, where the distance between
the coil and pole pieces is very small to maintain a strong magnetic eld
in the gap.
Rumble Turntables for the making and playing of phonograph records are
designed to operate as smoothly and noiselessly as possible, but no mat-
ter how careful the design, the unit will exhibit some noise, caused by bear-
ing irregularities, shaft and pulley eccentricities, belt inconsistencies, etc.
This noise is mostly low-frequency, and is called rumble. The frequency
content affects the audibility of rumble, higher frequencies being more
annoying than lower ones.
The advent of direct drive turntables did not eliminate rumble, but sub-
stituted a different kind, caused by motor vibration and irregularities, or
cogging, of the drive system. Rumble is measured by playing back a silent
groove record, supposedly recorded without rumble, and measuring the
low-frequency noise output of the phono cartridge. It is expressed as an
rms value in decibels below a reference level.

S
Sabin The unit of acoustical absorption, equivalent to 1 square foot of a
perfect absorber. An open window of 1 square foot would have 1 sabin of
absorption. There are no perfect sound absorbers (other than open win-
dows), and sound absorbers are rated in terms of their absorption as a per-
cent of a theoretically perfect absorber. The sound absorption coefcient is
actually this ratio. For instance, an absorber with a coefcient of .5 will
absorb 50 percent of the sound that reaches it. Two square feet of this mate-
rial will have a total absorption of 1 sabin. One square meter of absorption
is called a metric sabin.
Absorption coefcients are used in the calculation of reverberation
times in rooms. The absorption of any material will differ at different fre-
quencies, and acoustical materials are measured at many frequencies to
determine their absorption characteristics.
The sabin is named after Wallace Clement Sabine, a Harvard professor
of physics in the early 1900s. He is credited as being the rst scientic acousti-
cian. He dened reverberation time, developed methods to measure and

336
 Sampling

predict it, and wrote extensively about architectural acoustics. He de-

signed Symphony Hall in Boston, and this is considered to be the rst music
hall to be designed using scientic principles. It is regarded today as one
of the nest halls in the world for symphonic music.
SACD, Super Audio Compact Disc SACD is a joint venture of Sony and
Philips for a new format for CD recording. It uses one-bit encoding of the
audio signal at a sampling rate of 64 times 44.1 kHz or 2,822,400 samples
per second. This encoding process is called dsd. The high sampling rate
does away with the need for an anti-aliasing filter at the input and
an anti-imaging filter at the output, eliminating the phase distor-
tion that these lters produce.
The SACD specication allows the discs to be encoded with both con-
ventional CD encoding and SACD encoding by using two layers of data
the top layer is for conventional CD encoding and the bottom layer is for
SACD encoding. The discs can be played on existing CD players (top layer
only) as well as on SACD players. It is also possible to encode up to six-
channel audio as well as conventional two-channel stereo on SACDs. See
also dsd.
Sample and Hold The part of the analog-to-digital converter that
actually does the job of sampling the signal is called a sample and hold
circuit. It measures the instantaneous signal voltage at a particular time
and holds this level constant for the duration of the sampling interval. This
level is meanwhile converted into a digital code of the sample bit size
(for example, 16 or 24 bits) before the sample and hold moves to the next
sample. This code is a digital (arithmetic) expression of the signal level at
that instant. Sample and hold circuits are subject to certain inaccuracies.
The sample and hold function is much simpler in the case of delta-sigma
encoding because each sample consists of only one bit. It can be positive,
negative, or zero, but its magnitude is xed and does not have to be trans-
lated into a digital word. See also aperture time errors.
Sampler A device used in electronic music synthesis that digitizes and stores
a short musical tone or phrase. The sampled signal may be the sound of an
instrument played at a particular pitch, and it can then be subjected to var-
ious digital signal processing techniques to change its duration, pitch, tim-
bre, etc.
Many of the sounds heard in electronic music productions are actually
the sampled waveforms of actual instruments. There has been much legal
turmoil in recent years about the use of even very short pieces of copyrighted
material that has been sampled and used in otherwise unrelated commer-
cial releases.
Sampling In a digital audio system, the analog audio signal must be
fed into an analog-to-digital converter to be changed into a series
of numbers for further processing by the system. The rst step in this is
sampling, where the instantaneous signal amplitude is determined at
very short intervals of time. Sampling must be done very accurately to
avoid adding distortion to the digitized signal. The sampling rate, which

337
Sampling Rate

is the number of samples per second, must be uniform and precisely

controlled.
The choice of the sampling rate is somewhat complicated. For traditional,
linear pcm, the Nyquist critieria states that the sampling rate must be twice
the highest frequency of interest. For audio, this is considered 20 kHz, so
the sampling rate must be at least 40 kHz.
Since perfect anti-aliasing lters cant be built, a slightly higher
sampling frequency is needed as a practical matter.
There are many standard sampling rates in general use. The two most
common for linear PCM are the professional sampling rate of 48 kilo-
hertz, which is often used for master recordings and digital sound with
video. The consumer sampling rate is 44.1 kHz, as on the common com-
pact disc.
This frequency was chosen because early recorders used videotape tech-
nology, which conveniently accomodated a 44.056 kHz sampling rate. 44.1
kHz was chosen for the compact disc merely because it is a simpler fre-
quency. As a practical matter, recordings are often originated and processed
at 44.1 kHz to eliminate having to convert the sample rate if the end prod-
uct is a CD. See also oversampling; delta-sigma modulation
As technology progresses, higher sampling rates and sample sizes are
being used for better sound quality. Some would view this as odd consid-
ering the publics acceptance of reduced delity in the form of compressed
digital audio from MP3s, DVDs, and motion pictures. To transfer a pro-
fessional digital recording at the 48-kHz standard to CD is possible by
purely digital means, and this process is called resampling. The other alter-
native is to convert the signal back to analog and then pass it into another
adc operating at the new sampling rate. This is not a good solution, how-
ever, because the conversion from digital to analog and back to digital is a
source of signal degradation.
Sometimes, the sampling rate will be notated as fs, as in frequency of
sampling.
Sampling Rate See sampling.
Sampling Rate Converter, SRC A somewhat complex digital device or com-
puter program that accepts sampled inputs at one sampling frequency
and converts them to sampled outputs of a different sampling frequency.
It may seem that once a signal is sampled at a particular rate, it would not
be possible to resample it at a different rate unless it were rst converted
back to an analog voltage, but it is theoretically possible to convert from
any sampling rate to any other while still in the digital domain. Of course,
the initial sampling will determine the maximum bandwidth of the sig-
nal, and if the new sampling rate is lower than the original, the signal band-
width will be reduced. The sampling rate converter is useful because there
are several standardized sampling rates in use for various digital audio sys-
tems. See sampling.
SAN Storage Area Network, a central, large, and high-speed digital data
storage system used for audio and video workstations.

338
 Saturation

Sanders Theater Seat Cushion The rst unit of acoustic absorption used
by Wallace Sabine in his experiments in auditorium acoustics at Harvard
University. He used the seat cushions from the Sanders theater to reduce
the reverberation in the Fogg Art Museum and, along with organ pipes as
sound generators and his ears as microphones, determined the reverberation
time as a function of frequency and the amount of added absorption. Later,
he dened the unit of absorption as 1 square foot of complete absorption,
and this has become known as the sabin in his honor.
Sand Filling A technique for increasing the mass and damping of the pan-
els of loudspeaker cabinets by forming the sidewalls from two panels
separated by about an inch or so and lling the interior space with dry sand.
It greatly reduces the amount of energy the sides of the cabinet radiate,
reducing the coloration of the output sound. The technique was rst
used by Gilbert Briggs of Wharfdale Wireless Works in England. In recent
times, other methods have been developed to add damping to the cabinet
sidewalls.
SAP Acronym for Second Audio Program in the ntsc analog TV standard.
SAP is a monophonic audio signal broadcast by a television station in addi-
tion to the standard television sound. One use is to broadcast the same
program in a second language.
The SAP signal is frequency modulated onto a 78.67-kilohertz sub-
carrier that is then added to the TV sound signal before the subcarrier
modulates the 4.5-megahertz sound carrier. To receive the SAP, the TV
receiver must be equipped with a device to select and demodulate the sub-
carrier. SAP is part of mts, added to NTSC television in 1984.
Satellite Radio Systems that transmit radio-like channels of digital audio
programming via satellites to receivers in automobiles or homes. In 1992,
the fcc (Federal Communications Commission) allocated a spectrum in
the 2.3-GHz band for nationwide broadcasting of satellite-based digital
audio; CD Radio (now Sirius Satellite Radio) and American Mobile Radio
(now XM Satellite Radio) are the two systems in operation in the U.S. Pro-
grams are beamed to three Sirius satellites and two XM satellites, which
then transmit the signals to the ground, where your radio receiver tunes in
to one of the channels within the signal. Signals are also beamed to ground
repeaters for listeners in urban areas where the satellite signal cannot be
received directly.
A third system called Worldspace is broadcasting now in Africa and Asia,
and is planning to begin broadcasting in South America. All three systems
require a subscription fee. See also dab.
Saturation In magnetic tape recording, saturation is the maximum mag-
netization that a tape can attain. In analog audio recording, actual recorded
levels are less than saturation because distortion is introduced if satu-
ration is approached, especially at low frequencies. At high frequencies, it
is not possible to reach tape saturation because the signal itself acts to par-
tially erase itself as it is being recorded. This is called self-erasure, and
limits the maximum level attainable in a tape recorder at high frequencies.

339
SAW Filter

Some FM recording methods record to saturation in normal operation, and

may be referred to as saturation recording.
The core of a transformer can also be saturated with magnetic ux by
excessive signal level, and the result is severe distortion. For this reason,
the maximum signal level that an audio transformer is capable of handling
is part of its specication.
Saturation also applies to active audio devices such as transistors or
tubes. A transistor is saturated if it is carrying as much current as pos-
sible with the given power supply voltage. In other words, an increase
in current at the base will not result in any output current increase.
SAW Filter A type of high-frequency bandpass filter that uses surface
acoustic waves on a small piezoelectric crystal. SAW lters are very pre-
cise and can be designed for very at bandpass and very steep skirts. They
are excellent for use as if lters in FM receivers.
SCA Acronym for Subsidiary Carrier Authorization. In FM radio trans-
mission, in addition to the 38-kilohertz subcarrier used for stereophonic
signals, an SCA subcarrier at 67 kHz can also be included and used for such
things as foreign languages or background music. In 1955 the fcc author-
ized SCA channels working with mono FM, and fm stereo with SCA was
approved in 1961. See also musicasting.
SCA Adapter A type of multiplex adapter for use in decoding sca
broadcasts, such as musicasting.
Schematic The wiring diagram of an electronic component or system is
called a schematic. It has nothing to do with the physical location of the
parts in the circuit, but is concerned only with their function. A drawing
showing the physical parts layout of a circuit is called a pictorial diagram.
SCMS Acronym for Serial Copy Management System, often pronounced
scums. SCMS is a special circuit built into consumer digital recorders such
as r-dat, minidisc, and cd-r machines that are to be sold in the United
States. The SCMS prevents unauthorized digital-to-digital replication of
compact discs. The riaa, the ifpi, the eia, and several manufacturers
agreed to use the system in 1989.
The introduction of the r-dat into the United States was delayed sev-
eral years because of a dispute between the makers of the machines and
the record industry regarding a circuit to inhibit multiple-generation copy-
ing of CDs, ostensibly to prevent copyright infringement. Several schemes
were proposed by the record industry, but they were found to impair the
quality of the resulting rst-generation copies.
The SCMS allows a CD to be directly copied digitally onto the recorder,
without using digital-to-analog conversion, ltering, copying, or analog-
to-digital conversion.
The system deactivates the digital anti-copy flag in the CD, allowing
it to be copied once. Another ag is inserted in the copy that prevents it
from being copied; that is, second-generation copies cannot be made. The
SCMS also prevents third-generation digital copying of R-DAT recordings
originally made from the recorders analog inputs.

340
 SCSI

The SCMS system is an outgrowth of the solo copy protection scheme,

which was devised by Philips but never put into practice. SCMS is rendered
impotent by computer drive audio extraction, also called ripping, which
ignores anti-copy bits. In addition, equipment designed for professional
use does not have SCMS.
Scoring Stage A large recording studio equipped with interlocked lm and
video playback, in addition to audio recording equipment. Motion pictures
and video production scores are recorded while the conductor watches the
image and conducts the performers so the nished recording is synchro-
nized properly with the video or lm footage.
Scrape Flutter In an analog tape recorder, the tape must pass over several
heads and tape guides on its way from the supply reel to the capstan,
and it will sometimes vibrate something like a violin string due to friction
from the heads and/or guides. This vibration is actually a high-frequency
flutter, and results in a frequency modulation of the recorded or repro-
duced signal. The frequency is too high to be heard as a pitch variation,
but is instead perceived as added noise to the signal. To combat this phe-
nomenon, some of the better tape recorders place a small precision roller
between the heads. This roller acts as a mechanical filter to reduce the
vibration of the tape. The roller is called a scrape utter roller.
Scratch Filter A type of equalizer that consists of a fairly sharp cutoff at
high frequencies to reduce the effect of record surface noise; usually called
a high-cut filter today. The name scratch lter was more appropriate in
the days of shellac 78-rpm records than it is today because of the much
lower surface noise on vinyl compared to shellac.
Screen Grid In the tetrode tube, a second grid is placed between the con-
trol grid and the plate. It is maintained at a high positive potential, and
it increases the gain of the tube.
Scribe When the mastering engineer nishes cutting an acetate mas-
ter record, he will scratch some identication letters on the uncut surface
near the label area with a sharp stylus. This is called scribing, and it is used
by the processing personnel to identify the record. The scribed code can
also be seen on the nished record.
Scrub To scrub is to listen to a part of a recorded audio signal at a variable
speed forward or backward under manual control to nd the exact loca-
tion in the music where an edit is to be made. In analog editing, the tape
itself is scrubbed back and forth across the playback head, and in digi-
tal editing, a daw is used.
Scrubbing Motion Relative back-and-forth motion along the record groove
of the playback stylus as it plays an analog record. It is caused by a com-
bination of the warp of the record raising and lowering the cartridge and
the vertical pivot of the arm being located above the plane of the surface
of the record. It causes a frequency variation of the music at the rate of the
warp, which is called warp wow.
SCSI, Small Computer Systems Interface A family of parallel computer
data interconnection formats. Many daws and computer audio users, espe-

341
S-DAT

cially Macintosh users, used SCSI hard drives for mass storage for many
years because of its high performance. SCSI has since been eclipsed by other,
cheaper systems such as firewire.
S-DAT Acronym for Stationary-head digital Audio Tape. See dat; dash;
prodigi.
SDDS Sony Dynamic Digital Sound is a motion picture digital sound sys-
tem capable of eight discrete digital channels. The sound track is on both
edges of the lm and is read optically by a device using a CCD (charge
coupled device) such as used in the late 1970s in the ill-fated colortek
4-track optical lm sound system. The soundtrack consists of an array of
microscopic dots (or pixels) much like those recorded on a CD. Both edges
are used to provide two continuous streams of data interleaved using a
cross-redundant error correction technique to further prevent drop-
outs from lm damage or scratches. The atrac technique is used to
reduce the bit rate of the digital data by as much as ve to one. As the lm
runs, red leds are used to illuminate the SDDS soundtrack. CCDs read the
SDDS data and convert the stream of dots on the lm into digital informa-
tion. This information is preprocessed in the reader and passed on to the
SDDS decoder.
SDIF-2 Short for Sony Digital Interface-2. This is an interface format devel-
oped by Sony for digitally transferring audio signals between different Sony
professional audio products. Each audio channel is transmitted over its own
cable.
SDMI Secure Digital Music Initiative. A working group formed by the riaa
to develop a voluntary method for protecting copyrights of music distrib-
uted on the Internet. The SDMIs mandate is to legitimize music distribu-
tion and prevent copyright infractions. It is backed by the major labels. The
similar group in England is called the Creative Industries Taskforce (CIT).
See also drm.
SDTV, Standard Denition Television Several formats for digital broad-
cast television have been put forth by atsc, the Advanced Television Sys-
tems Committee. SDTV standards include 480 lines interlaced scan and 480
lines progressive scan, as compared to 1080 lines for high denition.
SECAM Systme En Coleurs Mmoir. The French standard for color tel-
evision transmission. Sometimes it is referred to as System Essentially Con-
trary to American Methods, which it is. The SECAM system is used in
Hungary, Algeria, and the former USSR, besides France. It uses 625 lines
in the picture and a frame rate of 25 per second.
Secondary In a transformer, the winding that supplies the output signal
or current is called the secondary, whereas the input signal or current
enters the primary winding. There is sometimes no fundamental differ-
ence in the two windings; either one can be used as a primary or secondary.
Secondary Emission In a tetrode vacuum tube, the high-speed electrons
striking the plate dislodge other electrons, which interfere with the action
of the screen grid. This is called secondary emission, and its effect is
reduced or eliminated by placing a third grid between the screen and plate.

342
 Semicapacitor

This grid is at cathode potential and is called a suppressor grid. It is the

dening element of the pentode tube.
Section See l-section; pi-section; t-section.
Segue A quick, smooth switch from one musical selection to another with-
out a gap; the absence of a cross-fade. English speakers usually pronounce
it, seg-way.
SEL Acronym for Sound Exposure Level. The SEL of a noise event is the
a-weighted sound pressure level lasting 1 second that would have
the same acoustic energy as the event itself. SEL is a way of comparing the
noisiness of events that have different durations, such as airplane yover
noise or motorcycle passby noise.
Selectivity The ability of a radio receiver to tune to or select one station
to the exclusion of others. A receiver with poor selectivity will pick up a
very strong station even when tuned to another station. This is often noticed
in car radios when driving near transmitting stations.
Self-Erasure The tendency of a high-frequency signal to erase itself while
it is being recorded on an analog tape recorder. Self-erasure usually limits
the maximum signal level that can be recorded on tape at high frequencies,
and saturation of the tape cannot be achieved. The signal itself causes
the record head to act like a tape degausser, partially wiping out the
recorded signal. See hx; hx pro.
Self-Noise See noise floor.
Self-Powered Speaker A loudspeaker system is called self-powered when
it has its own power amplifiers inside the cabinet. This usually also
implies that special circuitry for power limiting, crossover, and equali-
zation are included. The advantage is that the entire system of speakers,
ampliers, and processors can be optimized, and no power is lost in long
speaker wires or passive crossovers. A line-level audio cable and AC
power cable are simply attached to the speaker system.
Sel-Sync When the tracks of an analog multiple-track tape recording are
recorded one after the other so as to build up a musical composite from
several parts, one must listen to the previously recorded tracks while the
next one is being recorded. If the standard tape playback head is used for
this, there will be a time delay between the tracks when played back because
of the physical distance between the recording and playback heads in the
machine. To solve the problem, the record head gaps of the previously
recorded tracks are used as playback heads while the adjacent tracks are
being recorded. Since all the record head tracks are in line, there will be no
delay. This technique was developed by Ampex in the 1950s, and was rst
used commercially by Les Paul. Sel-Sync does not apply to digital recorders
because new tracks are synchronized to played-back tracks digitally.
The term is short for selective synchronization, and was invented at
Ampex by Ross Snyder. The idea was trademarked, but not patented, and
others have used the idea in their own recorders, using such names as
simul-synch.
Semicapacitor An electrical circuit whose impedance is inversely pro-

343
Semiconductor

portional to the square root of frequency rather than inversely propor-

tional to frequency itself has been called a semicapacitor. The phase of the
current in a semicapacitor leads the voltage across it by 45 degrees, rather
than the 90 degrees in a true capacitor. The proposed unit of semicapaci-
tance is the semifarad. If a semicapacitor is connected in parallel with a semi-
inductor, the impedance will have a broad peak rather than the sharp
peak that results in an lc circuit. This broad peak can be called a semi-
resonance.
A resistive transmission line that has distributed shunt capacitance along
its length exhibits semicapacitive behavior.
Semiconductor A semiconductor is technically any material whose elec-
trical conductivity is somewhere between that of a conductor and an insu-
lator. Semiconductors used for making solid-state electronic devices are
the elements silicon and germanium. They are not semiconductors in their
pure form but must be subjected to a complex procedure in which certain
impurities are diffused into the pure crystals. This is called doping and
increases the conductivity of the elements. Doping can add an excess of
either positively charged ions or negatively charged ions to the pure mate-
rial. These two types of doped elements are called P-type and N-type semi-
conductors. diodes are made by combining P-type and N-type materials
to form P-N junctions. transistors are made in many forms by combin-
ing junctions in more complex ways, some of the simplest of which are PNP
and NPN.
Although strictly speaking the materials involved are the semiconduc-
tors, some devices themselves, such as transistors and integrated circuits,
are often collectively called semiconductors.
Semifarad See semicapacitor.
Semihenry See semi-inductor.
Semi-inductor A device in which the electrical impedance is proportional
to the square root of frequency rather than directly proportional to fre-
quency. In a semi-inductor, the phase of the current lags the voltage by 45
degrees; whereas in a true inductor, the phase lag is 90 degrees. A semi-
inductor is said to possess semi-inductance, and its units are semihenrys.
The concept of semi-inductance was developed by John Vanderkooy of
the University of Waterloo, in Canada. It is useful in the analysis of the
behavior of dynamic loudspeakers. The current in the voice coil
induces eddy currents in the iron magnetic structure around the coil,
and these eddy currents modify the electrical impedance of the coil so it is
dominated by a component proportional to the square root of frequency.
The effect occurs much more strongly at high frequencies than at low fre-
quencies, where the coil impedance becomes simply inductive.
The semi-inductive behavior of the coil is benecial to the operation of
the loudspeaker, for it reduces the rise in impedance at high frequencies,
improving the high-frequency response.
Some manufacturers have inserted a copper or silver ring around the

344
 Sensurround

central magnetic pole piece to increase these eddy currents and further
reduce the high-frequency impedance. This works because copper and sil-
ver are much better electrical conductors than the iron of the pole piece.
However, the conductive ring displaces some iron, and the magnetic eld
the voice coil sees is reduced, so the overall sensitivity of the speaker is
reduced somewhat.
Semiresonance See semicapacitor.
Semitone The musical interval one note in the chromatic scale. The inter-
val from C to C sharp is a semitone, as is the interval between C sharp and
D, etc. In the equal temperament system of tuning, all semitones are exactly
the same interval, with a frequency ratio of the twelfth root of two, or slightly
less than six percent in frequency. Equal temperament is almost universally
used in the tuning of pianos world wide, but some instruments, such as
harpsichords and organs, may be tuned to one of several systems of
unequal temperament. These tuning schemes have interesting effects on
the music being played. For instance, a selection played in the key of C will
sound different than if it is played in the keys of D or G sharp. There is
much discussion and sometimes even gnashing of teeth and wringing of
hands among certain musicians over the desirability of using these tem-
peraments, but that is another story. See also Appendix 8.
Send An output on a recording or sound reinforcement console for a sig-
nal to be sent to another device, such as a reverberator or equalizer.
The signal is returned to the console via the return connector. Typical con-
soles will have several sends and returns.
Sensitivity The minimum required signal at the input of an audio
device in order to produce the rated output is generally called the sensi-
tivity of the device. The higher the sensitivity, the lower the signal required
at the input.
The sensitivity is usually specied for a particular output. For instance,
the sensitivity of a power amplifier is that input voltage that will result
in the rated output power. In an fm tuner, the sensitivity is the input sig-
nal level which will result in a specied signal-to-noise ratio of the
output signal.
Sometimes sensitivity is used in a relative way. A microphone may
be said to be very sensitive, or a phono cartridge might have low sensi-
tivity, without reference to numbers.
Sensurround A now obsolete motion picture sound system that used very
strong, very low frequencies to simulate the effects of explosions, earth-
quakes, etc. Sensurround was developed by the sound engineers W. O.
Watson and Richard Stumpf of Universal Studios in 1974 for the movie
Earthquake. Sensurround essentially created subsonic, low-frequency
vibrations between 5 and 40 hertz at sound pressures of 110120 deci-
bels, causing the audience to feel low vibrations during the main earth-
quake and dam collapse. The process was used for two subsequent lms,
Rollercoaster and Midway. The rst lm simply triggered a noise genera-

345
Sequencer

tor during the earthquake sequences, although later versions of Sensur-

round did record very low-frequency information on the lm. The actual
setup consisted of six speakers; four large speakers behind the screen, and
two on a platform in the back of the theatre. To give you an idea for the
size of the speaker cabinets, two featured Cerwin-Vega Model W horns
8 ft. long, 4 ft. wide, and 4 ft. high. The other two had Cerwin-Vega Model
C horns 1 ft. wide and 5 ft. high. Each speaker was driven by a 1000-watt
amplier, which was controlled by a special optical control track in the
lm.
Sequencer A module in analog electronic music synthesizers that generate
a repeating sequence of predetermined control voltages. These
voltages can be used to control pitch, loudness, and timbre, etc., of the syn-
thesized signals. The more common usage today refers to the MIDI
sequencer.
Series An interconnection of circuit elements one after another in such
a fashion that the current in all of them is the same; the voltage across
each may be different, however. This is unlike parallel connection,
where the voltage across each element is the same, but the currents may
be different. An example of series connection is the old-fashioned Christ-
mas tree lights that all went out if one burned out. In some cases, loud-
speakers may be connected in series and then to a single power amplier
to ensure that the amplifier is driving an adequately high load imped-
ance. However, as with the Christmas tree lights, if one fails, the others
also stop working.
Servo Short for servomechanism, which is a control system that uses feed-
back of an output signal to compare to a reference signal. The difference
between the two is an error signal, which is used to change the output sig-
nal in such a way as to reduce the error. In a servo system, if the reference
is changed, the output will rapidly change to a new value, and so it is con-
trolled by the reference.
Servo systems are of many different types and are quite common. One
example of a mechanical servo is the power steering apparatus in a car. The
steering wheel position is the reference and the mechanism moves the wheels
so as to reduce the difference between this reference and the actual posi-
tion of the wheels to zero.
The rst servomechanism is generally thought to be the centrifugal y-
ball speed governor that James Watt built into his steam engine. The speed
of the engine causes the yballs to rise, which in turn reduces the steam
input with a valve connected to the ball assembly. If the load on the engine
increases, the speed decreases a little, causing the valve to open more and
speed up the engine. Thus, the speed will remain almost the same over vary-
ing loads.
This, however, was not the rst servo, for organ builders had been using
servos to control the wind pressure in an organ for centuries before Watt.
The pressure in the windchest causes a bellows assembly to rise and close
a valve from the air source. More demand for air, as when playing with

346
 Sharp

more stops on, causes the bellows to fall a little, opening the valve more,
restoring the pressure.
What has all this to do with audio, you ask? Servos are used to control
speeds in tape recorders and record turntables, and are used to position
the laser beam in compact disc players. Some speed servos use a vibrat-
ing quartz crystal as a reference, and they are often advertised as being
quartz controlled or some other catch phrase. The quartz oscillator
is very stable over long periods, and makes a good reference for speed control
under varying loads.
The tape tension in some tape recorders is servo-controlled also, usually
by means of a mechanical linkage that uses a spring tension as a reference.
These servos prevent the tape recorder from running at different speeds
with different amounts of tape on the reels.
Negative feedback in an amplifier is an example of an electronic
servo.
ServoDrive A registered trademark for a novel subwoofer system orig-
inally sold by Intersonics, and now by Sound Physics Labs and ServoDrive,
Inc., which uses small rotary DC electric motors to drive the loudspeaker
cones rather than voice coil and magnet assemblies. The rotary motion
of the motor is coupled to two cones by a system of small belts in such
a way that the cones move in opposite directions, canceling any recipro-
cating force on the motor shaft. The two cones are disposed on opposite
sides of a closed cabinet so they radiate low-frequency sound in phase with
each other. The motor is specially designed to have very little inertia and
very high torque. The motor is driven directly by the amplifier output,
and its alternating clockwise and counterclockwise motion follows the
audio waveform; the belts resolve this motion into back and forth motion
of the cones.
The advantage claimed for this system is an increase in linearity for very
large cone excursions, allowing very high sound level output with low dis-
tortion. The frequency response extends uniformly from 25 to 100 hertz.
The name ServoDrive suggests that a servomechanism is used, which
implies motional feedback in the system. This is in fact not the case,
for the designers believe the system is sufciently linear that it does not
need the feedback to reduce distortion. In any case, this is an unfor-
tunate and somewhat misleading choice of words (but not unprecedented
in the audio business!).
SFX Shorthand for sound effects.
Sharp Higher in pitch, as opposed to flat, which means lower in pitch.
In reference to musical scales, sharp indicates one-half step higher; for
example, F sharp is one-half step higher than F.
In high-pass, low-pass, and bandpass filters, sharp refers to the
rapidity with which the response of the lter falls off in the lter stopband.
Very sharp lters have a very steep slope of decline in response. In gen-
eral, the sharper the lter, the greater will be its phase shift. The sharp-
ness of a lter is commonly described in decibels per octave. Six dB per

347
Shellac

octave of slope is a gentle lter, while 24 dB per octave is a sharp cutoff,

and 96 dB per octave is a very sharp cutoff. Filter slopes usually work out
to be multiples of 6 dB per octave because one lter section provides this
slope, and complex lters are made by cascading several sections, each
adding attenuation at 6 dB per octave.
Shellac An organic material excreted by certain tropical insects, shellac is
used as the active ingredient in some varnishes and in shellac. Shellac
was the material used for making 78-rpm records. It is breakable and varies
widely in quality. Pure shellac is quite smooth and fairly noise free when
pressed into a record, but pure shellac is difcult to obtain. Most of the mate-
rial used to make records contained signicant impurities, including twigs,
insect fragments, sand, etc. The introduction of vinyl chloride, which is the
material currently used for records, was a great improvement, especially
in uniformity.
It is interesting that Thomas Edison did much experimentation to nd
a suitable material for making records. For his Diamond Discs, he invented
a plastic made from formaldehyde, which he called densite. It was al-
most identical to Bakelite, a plastic invented by Dr. Leo Baekeland in 1907
and commercialized in 1909.
Shelving Sometimes an equalizer is required wherein the response is at
above and below a certain transition frequency range and needs a grad-
ual slope within the transition range. Such an equalizer is called a shelv-
ing equalizer or shelving control because of the resemblance of the response
curve to a shelf. An example is a low-frequency reduction control for use
when a person is speaking very close to a microphone. Such an equalizer
might have a 20-decibel shelf, and a 6 dB per octave roll off below
200 hertz. See also proximity effect.

Shelving

Shibata An LP phono cartridge stylus that has a shape designed to

reduce pinch effect and tracing distortion. Its edges that make con-
tact with the record groove walls are sharply curved in the horizontal axis

348
 Shock Mount

but are nearly straight lines in the vertical axis. This makes the area of con-
tact almost a line along the groove wall, which reduces the wear on the
groove as well as the stylus.
The shape was developed by the Japanese designer Shibata. It can be
said to be a logical extension of the elliptical stylus shape.
Shield, Shielding A shield is an enclosure that protects its contents
against the inuence of magnetic elds or electrostatic elds or both. Some
audio circuit elements, such as transformers and tape recorder
heads, are sensitive to magnetic elds. A varying magnetic eld induces
currents in wires, and especially in coils of wires, at the frequency at
which it is varying. The most common magnetic elds found in house-
holds are caused by the 60-hertz power line, and they are responsible
for most of the 60-Hz hum heard in audio systems. Some components,
most notably power transformers, generate 60-Hz hum elds, and if
a shield of soft iron, or other metal of high magnetic permeability, is placed
around them, the elds are greatly reduced. This is one common type of
magnetic shielding. The other type of shielding is placed around the sen-
sitive circuit element. Thick shields are placed around tape heads to reduce
induced hum.
Audio circuit elements are also sensitive to electric elds. The 60-Hz
power lines also produce strong electric elds. Any electric conductor may
be used as an electrostatic shield, and copper is generally used. One place
where electrostatic shields are used is in signal cables between audio
devices, such as microphone and phono cartridge cables and cables
between preamps and power amplifiers. These shields are usually of
braided copper wire, and do not reduce magnetically induced hum. To
reduce magnetic induction, the shields would have to be of iron, and it is
difcult to make such shields exible. Other techniques such as twisting
the signal wires together and using differential input ampliers are used
to reduce magnetic interference.
In general, all the chassis and shields in any given audio system are
connected together and sometimes are grounded to earth through a water
pipe or other suitable connection. The actual connection to earth is not as
effective an interference reducer as is generally supposed. More important
is the integrity of the shield between devices and in cables. Sometimes con-
nection to earth will cause a ground loop, which can increase induced
noise rather then reduce it. See also balanced line.
Shift Register A type of integrated circuit time delay device in which
a small quantity of electric charge is transferred, or shifted, along a series
of small capacitors in response to a timing signal. The charges usually rep-
resent digital binary bits rather than analog signal levels. See also bucket
brigade.
Shock Mount A exible mounting arrangement for placing a microphone
on a stand such that tapping or kicking the stand will not transmit mechan-
ical shocks to the microphone, which would cause noise in the output. Shock
mounts are made of rubber or some similar material or sometimes springs.

349
Shock Switch

Some microphones are much more sensitive to shocks than others. omni-
directional ones generally are much less sensitive than cardioids and
velocity microphones.
Vibration isolators used to reduce acoustic feedback in record-
playing turntables are also sometimes called shock mounts.
Shock Switch A special switch found on some compact disc players that
makes the unit less susceptible to losing the track in the disc due to mechan-
ical shocks.
It is possible, with the formidable low-frequency performance of the com-
pact disc system and some woofers, that the vibration of the player caused
by the reproduced sound itself can cause the laser to lose its place while
tracking the series of pits on the disc. The shock switch tightens up the laser
servo, making it stiffer. It will then be less apt to lose track, but then if
a sufciently strong shock does occur, it will take a longer time to re-establish
tracking.
This phenomenon is largely taken care of now by larger data buffers,
such that some portable CD players can hold nearly a minute of sound dur-
ing mistracking.
Short Abbreviation for short circuit, which is a direct connection between
two points in an electric or electronic circuit. Shorts are usually unde-
sirable and inadvertent, but sometimes a purposeful connection is called
a short. A defective component, such as a transistor or capacitor, etc.,
is said to be shorted if its terminals are directly connected internally. Short-
ing is a common failure mode for many components.
Short Line An audio line, such as a telephone line, whose electrical length
is shorter than the shortest audio wavelength to be transmitted. See also long
line.
Shotgun Microphone See line microphone.
Shot Noise A type of random noise generated in nonmetallic media such
as active devices like transistors or vacuum tubes. It is proportional to the
current in the device and its load resistance. Shot noise arises because cur-
rent in the device is made up of discrete charges and is thus somewhat
grainy and not continuous and smooth. The total noise in the output of
an active device is the sum of the shot noise of the device and the thermal
noise, or johnson noise, at its input. It is useless to reduce the input ther-
mal noise if the shot noise overwhelms it. It turns out that different devices
have different optimum input impedances: for vacuum tubes, about 1
megohm, and for most bipolar transistors, about 500 or 1,000 ohms.
Shufer Circuit A type of matrixing circuit that includes equalization
of the sum and difference components. It allows the apparent width of a
stereo image to be varied as a function of frequency. It works on almost
any stereo signal. Its invention is credited to the multitalented British engi-
neer Alan blumlein in the 1930s, and was recently revived by Richard
Kaufman and David Griesinger in the United States. See also m-s stereo;
blumlein, alan.

350
 Signal

Shuttle To wind the tape on a tape recorder back and forth in order to locate
a specic selection.
SI Systme Internationale, or the international system of units.
In the eld of architectural acoustics, SI is also used, by Barron and
Ando, among others, to indicate spatial impression, which is similar to
envelopment. SI is correlated with iacc.
Sibilance Vocal recordings, especially if made with very close micro-
phones, are often characterized by excessive loudness of the voice sibi-
lants, and this effect is sometimes called sibilance. The most difcult
sibilants to reproduce accurately are the sounds s and sh. The effect is
accentuated by high-frequency peaks in microphones and in many loud-
speakers; it is reduced by the use of a de-esser.
Sibilant See sibilance.
SID Short for Slew Induced Distortion. See transient intermodula-
tion distortion.
Sidebands When a signal of frequency F amplitude modulates a car-
rier of frequency C, the resultant waveform consists of the carrier and
other frequency components at C minus F and C plus F. These added fre-
quency components are called sidebands because in a graph of the spec-
trum they appear on the left and right sides of the carrier frequency.
The sum of the two sidebands and the carrier looks like a single frequency
that is changing in amplitude at the modulation rate; however, it is actu-
ally the simple addition of the sidebands and carrier, with the carrier
unchanged in any way. This means in an am broadcast station, all the broad-
cast information is contained in the sidebands and none in the carrier. To
maximize efciency, the carrier can be suppressed and not transmitted at
all, although it is much easier to recover the modulating signal in the receiver
if the carrier is present.
The upper and lower sidebands of standard AM transmission contain
the same signal information, but it is possible to modulate a carrier so that
the upper and lower sidebands carry the left and right stereo channels.
This is the principle of one type of AM stereo broadcasting.
In the case of frequency modulation (fm), the situation is somewhat dif-
ferent. The FM sidebands extend farther from the carrier frequency and dif-
fer in phase from AM sidebands. In commercial FM transmission, only
sidebands extending plus and minus 75 kilohertz from the carrier are
transmitted in order to prevent the transmissions from overlapping and
interfering with each other. Thus more transmitters can operate within a
given frequency band.
Side Tone See hybrid transformer.
Siemens The si unit of conductance.
Sigma-Delta Modulation See delta-sigma modulation.
Signal A signal is an electrical phenomenon, usually a voltage but some-
times a current, that contains desired information, as opposed to noise,
which is undesired. Audio signals are generally electrical analogs of the

351
Signal Processing

corresponding sound waveforms. The signal-to-noise ratio (S/N)

is the ratio in decibels of the signal power to the noise power at a point
in a circuit.
Signal Processing Signal processing is the modication of an audio sig-
nal in a generally desirable way by any device inserted in the audio path.
Examples of signal processing devices are reverberators, noise reduc-
tion units such as dolby, dbx, etc., and equalizers.
Signal-to-Noise Ratio Signal-to-noise ratio is the ratio of the signal power
at a certain reference point in a circuit to the noise power that would exist
there if the signal were removed. This ratio is expressed in decibels. For
instance, if an analog tape recorder has a signal-to-noise (S/N) ratio of 50
dB, the signal power at the output is 50 dB above the noise power.
Measurement of S/N ratio is complicated by several factors. In an ana-
log tape recorder, for example, a signal is recorded at a high level, approach-
ing the onset of audible distortion, and its voltage is measured with
a voltmeter. Usually the signal is a sine wave. Then the signal is removed
from the input, and the remaining noise consisting of tape noise, hum, elec-
tronic noise, etc., is measured with the voltmeter. Because the noise will be
mostly random in nature, the voltmeter should be a true rms meter. The
ratio of these two measurements expressed in dB is the signal-to-noise ratio.
This measurement ignores any type of noise that is present only when the
signal is present, such as modulation noise.
The subjective audibility of the measured noise will not be well corre-
lated to the rms measurement of it because of our ears varying sensitivity
to different frequencies. Therefore, a filter or weighting network is
usually placed in the meters input. The characteristics of this weighting
network are not universally agreed upon, so when comparing noise
specications, as between European and American equipment, care must
be taken to see that the measurements were similarly made.
In digital audio systems, the signal-to-noise ratio is dened as the ratio
of the maximum possible sine wave signal power to the quantization
noise power, also expressed in dB. This is an unambiguous measure in lin-
ear pcm systems; but in all other PCM schemes, the quantization noise
depends strongly on the level of the signal being encoded. See pcm for a
discussion of nonlinear PCM systems. See also dynamic range.
Silicon One of the more abundant chemical elements, with atomic num-
ber 14. Pure silicon (Si) is an insulator, but when it is doped, or infused
with small quantities of various impurities, it becomes a semiconductor
and is the basic material from which almost all solid-state electronic
devices are built.
Silicone Any one of a group of chemical compounds that contain silicon
and oxygen, not to be confused with silicon. Silicones are extremely var-
ied in their properties; some are solids and some liquids. Most of the vis-
cous damping uids are silicones.
Simulcast The broadcasting of one program over more than one radio or

352
 Single-System Sound

television station at the same time. An example is broadcasting a live con-

cert with monaural sound on television and at the same time broadcasting
the same concert in stereo from an fm station. The listener can thus watch
the video on a television set with the volume turned down while listening
to the program in FM stereo. Simulcasting has been rendered obsolete by
the advent of television broadcasting with stereophonic sound.
Sine Wave The sine wave, or sinusoid, is the simplest possible periodic
waveform. It consists of a single frequency and has a musical pitch
but a neutral timbre or tone quality. It is called a sine wave because it has
the same shape as the mathematical sine function, familiar to most of us
from trigonometry.
Sine waves are commonly used as test signals for audio equipment
because they consist of only one frequency. If a sine wave is passed through
a device, it will always have some distortion added to it, and this is
detected by looking for other frequencies, such as harmonics of the orig-
inal sine wave frequency. See also Appendix 5.
Single-Ended Single-ended can refer to an amplifier output circuit
using a single vacuum tube or transistor, as opposed to a push-pull
circuit, which uses two active components operating in opposition to each
other.
Single-ended more commonly means a circuit interconnection that is not
balanced. Short lines interconnecting audio components are usually
single-ended rather than balanced for simplicity, but low-level lines such
as microphone cables are almost always balanced for reduced noise. Tele-
phone lines are also balanced for the same reason.

Single-Ended Amplier

Single-System Sound A way of producing sound motion pictures where

the sound is recorded directly onto the lm in the camera at the time of
shooting. This is the least expensive way to make sound movies, but is

353
Sinusoid

Skating Force

also much less exible and of much poorer quality than double-system
sound.
Sinusoid See sine wave.
Sirius Satellite Radio See satellite radio.
Skating Force The force on the stylus of a record player that causes the
tonearm to move toward the center of the record. The skating force is caused
by friction between the record and the stylus and is a result of the fact that
the end of the arm is offset at an angle to the arm itself. Therefore, placing
the stylus on a smooth surface to adjust anti-skating force is not quite right.
The Johnny Winter LP Second Winter is somewhat famous for having a blank
side 4, often used for this purpose. See also anti-skating device; off-
set angle.
Skew Skew is the motion of recording tape past a record or reproduce head
at an angle different from 90 degrees. Tape skew can be caused by loose or
poorly adjusted tape guides, or it can be caused by the tape itself having
been improperly slit during its manufacture. It causes high-frequency loss.
See also azimuth.
Skin Effect The tendency of high-frequency current to travel near the out-
side of an electric conductor rather than all through its cross section. Skin
effect increases the effective resistance of a wire at high frequencies. It
is not noticeable at audio frequencies but becomes troublesome at radio
frequencies. There are certain people who believe the skin effect is actu-
ally audible at audio frequencies, but this has not been proven to be true.
Skirt The slope of a filter response curve outside its passband is some-
times called a skirt, especially in the case of a bandpass lter.
Slap Echo The single repetition of a signal at a xed time delay to simu-
late an echo from a single reecting surface, as opposed to a multiple echo
from a time delay, where the delayed signal is repeatedly fed back into the

354
 Slew Factor

delay input. Under certain acoustical conditions, a room can exhibit a sim-
ilar sounding phenomenon when two walls are parallel and spaced apart
by about ten to thirty or so feet apart. This is called a utter echo.
Slate To identify the various takes in a recording session by announcing
the take numbers and recording them on one track of the tape. Slating and
the notes taken at the time of recording are important once tape editing
begins; it would be almost impossible to nd any particular take otherwise.
Sometimes a low-frequency, high-level tone is recorded on analog
tapes before each take to aid in nding the beginning of the take when spool-
ing the tape past the heads. The low frequency is heard as a beep even
though the tape may not be contacting the playback head. This tone is called
a slate tone and is usually about 20 or 30 Hz.
The term originated from motion picture usage, where a small chalk-
board was photographed to identify the beginnings of various takes when
lming. With the advent of motion picture sound, the slate became a clap-
board when it acquired a hinged and striped appendage that is clapped
against the board itself at the beginning of the take. During lm editing,
the sound of the clap is synchronized with the image of the clapper mak-
ing contact. A slate at the beginning of a take is called a headslate. Some-
times another slate will be added at the end of a take to verify that sound
synch has been maintained. This is called a tailslate and is made with the
clapboard held upside down. In this way, the lm editor can tell the begin-
ning of a shot from the end. The advent of smpte time code has made
the clapboard nearly obsolete, however its modern variant is still used to
identify takes visually with written data, and to generate and display the
timecode for the camera and sound recorders. In an emergency, the clap-
stick can still be used to synch sound and picture.
Slew Factor Dened originally by the ihf as the ratio of the highest fre-
quency that can be applied to the input terminals of an amplifier at a
signal level that produces rated output at 1 kilohertz, and that can be
reproduced at the output with acceptable linearity to 20 kHz.

Skirts of a Bandpass Filter

355
Slewing

For example, if an amplier can produce a 10-kHz tone with acceptable

linearity at the same level as the maximum it can produce at 1 kHz, its slew
factor is 12. Slew rate is a measure of the high-level high-frequency per-
formance of an amplier. See also tim.
Slewing, Slew Rate The slew rate of a signal in an audio device is the
rate, usually expressed in volts per microsecond, at which the signal is
changing. It depends on the amplitude and frequency of the signal; high-
amplitude, high-frequency signals have the highest slew rates. The slew
rate also varies with time along the waveform, usually being the highest
as the waveform passes through zero.
An audio device will have a maximum slew rate above which it cannot
operate. This maximum slew rate limits the high-frequency power output
of ampliers and limits the high-level, high-frequency handling capacity
of all audio devices. Too low a maximum slew rate results in slew-induced
distortion. See also tim.
Slim Wire System See fat wire system.
Slip Cue A method of cueing for records where the stylus is placed on
the record at the beginning of a selection and the record is held stationary
while it slips on the turntable. At the moment of truth, the record is let go,
and it almost instantly comes up to the speed of the turntable. Slip cueing
was done by radio disc jockeys merely to start songs on cue, but the mod-
ern club DJ takes slip cueing to new levels.
Slope The portion of the frequency response curve of a filter or
equalizer that indicates the response is falling off as the frequency is in-
creased or decreased. Filter slopes are characterized by their steepness and
are usually 6, 12, 18, or some other multiple of 6 dB per octave. Filter slopes
are sometimes called skirts.
Slope Detector A primitive type of FM detector that uses a low-pass
or high-pass filter to convert the frequency-modulated signal into
an amplitude-modulated signal. The carrier frequency is tuned to
be on the slope, or skirt, of the lter, and as the frequency varies with
the modulating signal, the varying sensitivity of the lter passes more or
less of its amplitude, resulting in the amplitude modulation. This modu-
lation is then detected by a simple diode detector.
Small Signal Bandwidth The bandwidth an audio device will exhibit at
relatively small signal levels. Most audio devices will have a wider fre-
quency range at low signal levels than at high signal levels because such
effects as slew limiting occur at high frequencies and high levels.
In devices using transformers, high-level, low-frequency signals can
be limited by saturation of the transformer core. In a power ampli-
fier, the large signal bandwidth is called the power bandwidth. The
small signal bandwidth is important, but the power bandwidth of a device
is far more so.
Smith, Oberlin In the September 8, 1888, issue of The Electrical World, a mag-
azine published in New York at the time, Oberlin Smith wrote the rst known

356
 SMPTE Time Code

description of magnetic sound recording under the title, Some Forms of

Phonograph.
In this article, Mr. Smith describes hypothetical magnetic recording on
a steel wire using a telephone transmitter as a microphone. As an alter-
native to wire, he says a string or thread, suitably impregnated with iron
lings or other small magnetic particles, might serve better. He also men-
tions illicit recordings of telephone conversations, an 1888 precursor to
todays bugging!
It was eleven years later that Valdemar Poulsen built his telegraphone,
which was a device for magnetic recording of telegraph messages, and
which started the evolutionary process that led to the tape recorder.
Smoke, aka Magic Smoke From the phlogiston theory of electronics: it
is smoke that makes ICs (integrated circuits) and transistors work. The
proof of this is self-evident because every time you let the smoke out of an
IC or transistor it stops workingelementary. This has been veried
through exhaustive testing, particularly regarding power amplier ICs and
transistors. (Incidentally, wires carry smoke from one device to another.)
Origin unknown but classic.24
SMPTE Society of Motion Picture and Television Engineers. The SMPTE
is a long-respected technical organization, and has been active in many
aspects of audio engineering. Founded in 1916 in the U.S., SMPTE has spread
throughout at least eighty-ve countries worldwide.
The classical method of measuring intermodulation distortion
was devised by the SMPTE.
SMPTE Time Code (TC) A digital code, standardized by smpte and the
ebu, consisting of 80-bit digital words and recorded on a video recorder
or lm camera for identifying each frame. The time code is also recorded
in audio recorders used for double-system sound and allows the syn-
chronization and the precise editing of the program at any frame line with
computer editors. Sometimes two multitrack audio recorders would be syn-
chronized with TC to gain more audio tracks. The basic version of the code
repeats in a 24-hour cycle, and within the 24-hour period, each frame is
identied by an exact time (HH:MM:SS:Frames). There are many other
specications to accommodate analog and digital technologies and differ-
ences in lm, video, and international standards. Originally used on ana-
log video and audio recorders and lm cameras, the code would be
recorded on a special analog track on the video recorder, on a mag stripe
or optical area on lm, and on one track of a multitrack analog audio recorder
or a special center track on a stereo audio recorder. This allowed recogni-
tion at faster than normal play speeds. A method called vitc (vertical
interval tc) inserts the TC data into the vertical interval of the television
signal, which has some advantages and disadvantages over longitudinal
TC (LTC). See also drop-frame time code.

24. Thanks to Dennis Bohn, Rane Corp.

357
Snake

Snake A snake, or mike snake, is a multiple-pair cable with several

microphone connectors on each end. Snakes are used when several
microphone cables must be run a fair distance, such as between a stage and
the sound control console. The snake replaces an unwieldy collection of
cables. The snake is also called a breakout cable.
Snake Track A special optical sound track on a test lm for adjusting the
alignment of the optical sound system in a motion picture projector
equipped to reproduce sva sound tracks. The snake track is a narrow sound
track with a recorded tone. The entire track weaves back and forth over the
width of a normal sound track, and the alignment is adjusted until the tone
is heard alternately and equally loudly on the two output channels.
Soft Clipping A form of clipping where the edges of the clipped wave-
form are rounded rather than sharp; compare to hard clipping. Soft clip-
ping is much easier on the ears and on tweeters than hard, for it contains
much less very high frequency energy. In the case of power amplifiers,
ones using vacuum tubes tend to exhibit softer clipping when overloaded
than do transistor ampliers. Small signal ampliers such as microphone
preamplifiers generally produce hard clipping, especially transistorized
ones.
Some ampliers and receivers exist with a switch to turn a soft clip-
ping circuit on or off.
Soft Knee A term usually used to describe the response of a compressor/
limiter when it begins to reduce the gain of the signal. The input vs. out-
put graph of the compressor/limiter makes a bend (looking somewhat like
a knee) when it starts to act. A soft knee characteristic smoothes this bend
so the onset of compression or limiting is gradual instead of sudden and
is not so harsh sounding.
Solder A metal alloy used to attach electrical conductors, such as electronic
components to circuit boards and wires to connectors. The alloy is com-
monly 60% tin and 40% lead, but variations are seen as well as small amounts
of other metals such as silver. The known toxicity of lead has given rise to
other alloys as well. Solder has a fairly low melting point and certain ow
characteristics that make it practical for electronic bonding. Electronic sol-
dering uses a rosin material to clean the surfaces to be bonded, called rosin
ux. If the solder is in the form of a wire, rosin ux is usually already in-
side the interior of the solder wire, and this is called rosin core solder. Acid
uxes exist but are never used in electronic work. Highly automated sol-
der machines can solder an entire circuit board at once with a wave of solder.
Hand soldering utilizes a heated stylus called a soldering iron, which the
user wields to heat the desired joint just enough to melt the solder and allow
it to cleanly ow around the joint. This can be quite a skill, and something
quite useful when making or repairing circuits or connectors. Americans,
at least, pronounce the material sodder.
Solid State Solid state refers to electronic circuits that use transistors
and integrated circuits instead of vacuum tubes as active elements.
Solo A special circuit conguration developed by Philips for use in dat

358
 Sounding Vessels

recorders that permits the recording of a perfect digital copy of a compact

disc by simply copying the digital code from the disc directly, without going
through the digital-to-analog conversion and subsequent ltering and
other analog circuitry. The Solo system would insert in the DAT recording
certain inaudible digital information (an anti-copy flag) that would not
allow the recorder to make another digital copy of the DAT copy. The pur-
pose is to prevent the unauthorized digital replication of CDs by unscrupu-
lous persons, but it allows the consumer to make a DAT recording of various
selections from CDs and still maintain the original quality. The riaa, the
British Phonographic Industry (BPI), the International Federation of the
Phonographic Industry (IFPI), the eia, and several Japanese and European
manufacturers all agreed in mid-1989 to implement a system somewhat sim-
ilar to Solo. See also scms.
Solo can also refer to certain buttons on recording consoles; most have
a solo button associated with each console input. Pressing a solo button
mutes the other channels and allows the recording engineer to hear only
the signal from that channel over the monitor speakers while the record-
ing or sound reinforcement program goes on unaffected.
Sone The subjective unit of loudness. A sound has a loudness of 1 sone
if its loudness level is 40 phons. This means a sound of 1 sone loudness
sounds equally loud as a 1,000-hertz sound whose spl is 40 decibels. A
doubling of loudness corresponds to an increase of 10 sones.
Sonic An adjective meaning of sound; the audio equivalent of visual.
Sonic refers to any sound, not just audible sound. Outside the audio eld,
the term is used to mean the speed of sound in air. Thus supersonic is faster
than the speed of sound, and subsonic is slower than the speed of sound.
These terms are not to be confused with ultrasonic and infrasonic.
Sound Effects Filter A special-purpose bandpass filter. For creation of
special effects for the stage and motion pictures, etc., sound effects lters
are often used to modify the frequency response of sound systems. An
example is a steep high-frequency rolloff to simulate the sound of a tele-
phone, or a voice on short-wave radio. Sound effects lters were very much
used in the radio broadcasts of the 1940s and 1950s.
Sound Field Microphone See ambisonics.
Sound Head The sound-reproducing device in a motion picture projector.
The traditional analog optical version consists of a roller over which the
lm passes, with one or two photocells inside. Associated with it is a lens
assembly that focuses the light from the exciter lamp through the small area
of the lm sound track and onto the photocells. Newer units may use an
led and ccd devices. The commercial theater lm projector sound head
also usually includes the drive motor for the projector.
Sounding Vessels As described by the Roman architect vitruvius, sound-
ing vessels are large urn-like helmholtz resonators placed in certain
Roman theaters to improve the acoustics. Not much is known about ancient
sounding vessels, and none survive to the present day. There is at least one
Roman theater that has niches in the audience area that probably held sound-

359
Sound Intensity

ing vessels. Also, in Saint Sophias cathedral in Kiev, Ukraine, which dates
from the thirteenth century, there are many Helmholtz resonators embed-
ded in the interior walls, which were placed there to amplify sound,
according to the local tour guide.
Of course the resonators, which are really resonant absorbers, cannot
amplify sound but merely absorb energy at well-dened frequencies and
then reradiate the energy later, effectively increasing the reverberation time
of the room. Vitruviuss account of them implies that they were tuned to
several notes of a musical scale, which would cause a musical chord to sound
after being excited by a loud noise. If some sound-absorbing material is
inside the resonator, it becomes a strong absorber at its resonant frequency.
This type of specic absorber has been used in some modern auditoria.
An interesting use of Helmholtz resonators is in the assisted resonance
system installed in the Royal Festival Hall in London. Many resonators,
tuned to a broad range of frequencies in the lower musical registers, are
placed above the ceiling of the auditorium. Each resonator has a small loud-
speaker in it and a microphone associated with it. The microphone picks
up sound at the resonant frequency and amplies it into the speaker, pro-
longing the reverberation time of the resonator and thus prolonging the
reverberation time of the auditorium itself. This is one of the few musically
successful attempts to electronically alter the acoustical properties of an
auditorium. A similar system was installed in the Hult Center auditorium
in Eugene, Oregon.
Sound Intensity Sound intensity is dened as a measure of the net ow
of acoustic energy in a sound eld. The units are watts per square meter.
Because the energy moves in a particular direction, sound intensity is a vec-
tor quantity, i.e., it has magnitude and direction. Sound intensity cannot
be measured directly, and it should not be confused with sound pressure
level, which is what a sound level meter measures.
Some confusion exists about sound intensity because it used to be dened
in terms of the energy content of a sound eld, without regard to the move-
ment of the energy. A much better concept for this is sound power.
It is interesting and important to note that there may exist a large sound
pressure at a point, but the sound intensity may be very small, or even zero.
This is the case for instance in standing waves, where the energy is slosh-
ing back and forth within a distance of a half-wavelength or so, but does
not progress through the medium.
A microphone always measures sound pressure, or sometimes it may
measure the particle velocity, but it can never measure sound intensity
directly.
Sound Level Meter A meter that measures sound pressure level. It con-
sists of a pressure microphone, an amplifier, an rms detector, a
logarithmic amplier, and a meter or other indicating device. It is battery-
powered for portable use. Usually, it will contain at least one weighting
filter to make its response more or less conform to the sensitivity of the
human ear. See also a-weighting.

360
 Sound Reinforcement

The microphone on a sound level meter is omnidirectional, with

wide and uniform frequency response, and must be very stable in its
sensitivity to maintain accurate calibration. An accessory to the sound
level meter is the acoustic calibrator, which presents the microphone with
a known sound pressure level to check the sensitivity of the microphone
and meter electronics.
A high-quality sound level meter will meet the ansi and iso speci-
cations for frequency range, time constants, microphone directivity, and
lter response curve accuracy, and is always expensive. There are several
very inexpensive meters available that masquerade as true sound level
meters, but they seldom meet the standardized performance requirements
for precise measurements. They may be useful for measuring relative lev-
els, such as estimating the noisiness of a machine by measuring when it is
running and comparing the noise level when it is turned off.
Sound-on-Sound A popular name for overdubbing.
Sound Power Sound power is the amount of energy radiated by a sound
source per unit time. Its units are joules per second, or watts.
Sound power is a measurement of a characteristic of a sound source,
and has no direct relation to sound pressure level (SPL). This is anal-
ogous to a light bulb, which may be rated at 100 watts. This does not tell
how bright a room will be with the light bulb in it; that depends on the size
and color of the room, etc. Thus, the SPL of a sound in a room is related to
the sound power of the source in a complicated way, depending on the envi-
ronment and microphone placement, etc.
Sound power measurements are made when the sources of sound are
being investigated, for instance in noise control investigations, whereas SPL
is measured if the effects of sound on a human subject are being studied.
Sound power cannot be measured directly, but must be calculated from
a series of SPL or sound intensity measurements and a knowledge of
the environment. See also power response.
Sound Pressure, Sound Pressure Level A sound wave progressing through
air causes the instantaneous air pressure at any given point to vary above
and below the barometric pressure in accordance with the waveform of
the sound. This variation in pressure is used as a quantitative measure of
the strength of the sound, and is called sound pressure. This is the quan-
tity that a pressure microphone measures, and if it is expressed on a
decibel scale and referred to a pressure of 20 micropascals, it is called sound
pressure level. The reference pressure of 20 micropascals is 0 dB on the scale,
and corresponds to the threshold of hearing at 1,000 hertz for a normal
human ear.
Sound pressure level is what is measured by a sound level meter. See
also level.
Sound Reinforcement The general term for a sound system designed to
amplify the voice and/or music to improve its intelligibility to an audience.
A sound reinforcement system always consists of at least one microphone,
associated preamplifiers, control console, amplifiers, and loud-

361
Sound Stage

speakers. The old term for such a system is pa system, standing for pub-
lic address, and it is still in common use.
Traditionally, sound systems have been monaural, i.e., they have had
only one independent channel of audio. In recent years, there has been a
strong interest in stereophonic sound reinforcement, and when prop-
erly designed and operated, these systems offer a large degree of natural-
ness. The best stereo reinforcement systems have three or ve independent
channels or more. It is possible to perform multichannel stereo reinforce-
ment of classical music programs for very large audiences in such a way
that the reinforcement itself is not noticed by the audience, although this
is rarely attained in practice.
The advent of large rock concerts has spawned the development and
renement of very high power sound reinforcement systems, and they can
be extremely complex and expensive, in addition to being very loud. Such
systems are really part of the musical ensemble, and the operator often has
as much musical talent as the performers.
The primary problem that has to be overcome in any sound reinforce-
ment system is acoustic feedback, which is an oscillation of the sys-
tem due to the microphone picking up the amplied sound and the system
amplifying it again into a continuous howl. To minimize feedback, direc-
tional microphones and loudspeakers are used, and much attention is paid
to their placement in the room and their proximity to the performers.
equalization is often used to reduce the system gain at frequencies
where the acoustic gain may be high due to peculiarities in the room
acoustics, the loudspeakers, and/or the microphones. See also sound sys-
tem equalization.
Sound Stage A theatrical stage for lming motion pictures or video pro-
ductions that is specially treated for the simultaneous recording of dialogue
or music. A sound stage must have a low level of background noise and a
low reverberation time. The cameras and other equipment used must be
specially designed for quiet operation.
Sound Stripe A narrow strip of magnetic sound recording material
applied to motion picture lm for recording of the sound track, in the man-
ner of a tape recorder. In the case of Super-8 and 16-mm lm, two stripes
were used, one for the sound and one just to balance the thickness so the
lm would wind smoothly on the reels. 35-mm magnetic stripe prints
could have multiple tracks and stripes on both sides of the lm, as well
as the traditional optical sound track. 70-mm release prints of motion pic-
tures exclusively used magnetic sound tracks until recent digital systems
became available. 35-mm mag prints are essentially obsolete and are
rarely seen.
Sound System Equalization The equalization of a sound rein-
forcement system, either to increase its amount of gain before feed-
back, or to make its overall frequency response more nearly flat.
Sound system equalization is a complex and tricky business, because the

362
 Source

acoustics of the room, as well as the frequency responses of all the audio
devices, are involved, in addition to the characteristics of the microphone
being used. Actually, it is only possible to attain a given frequency response
curve at one place in an auditorium and for a narrow range of microphone
locations at a time. Therefore, many curves are measured and an average
is made to represent the best compromise to the desired curve. Equaliza-
tion requires measuring the frequency response while the equalizer is being
adjusted, and this requires the use of a real-time analyzer (RTA) or an
fft analyzer.
Some years ago, it was thought by some audio people that if a sound
system had a series of very narrow band notch filters, each tuned to a
frequency at which the system would be likely to go into acoustic feed-
back, the available gain would be increased. This is true to a certain extent,
but it would only work for a single microphone location, and also the dis-
tortion of the tonal quality by the phase shift introduced by the lters is
disturbing to many listeners.
Quite a few years ago the idea of putting a frequency shifter that in-
creased the frequency of the signal by a few hertz in the sound system
was tried to reduce the tendency for feedback. The theory was that the
amplied sound picked up by the microphones would not be at the same
frequency as the original sound that was being amplied, and therefore,
with repeated trips through the microphones, ampliers, and speakers,
would not build up energy in the room at a specic frequency. Unfor-
tunately, it was soon found that instead of conventional feedback, the
sound system would emit chirps when the gain was increased to high
levels. This was as obnoxious as feedback, and these systems met a rapid
demise.
Today the narrowest frequency band in common use for xed sound
system equalization is one-third octave, or about 23 percent of the cen-
ter frequency. There is a relatively new technique, however, which uses a
dual-channel FFT analyzer to continuously measure the frequency response
of the entire system, including the room, and using live music as the test
signal. Equalization in narrow bands can be performed in real time dur-
ing the progress of a concert without the audience knowing about it. The
advantage of this is that no matter where the system is used, and no mat-
ter where the microphones are, the optimum equalization is rapidly
achieved. An example of this real-time approach is the popular Sabine FBX
Feedback Eliminator See also feedback.
Source One terminal of a field effect transistor, the other two of
which are the gate and the drain. Also, some tape recorders with 3
heads or condence heads have a switch labeled Source/Tape for
switching the output between the source of the signal and the playback
from the recorded tape. Source is sometimes labeled input in that you
monitor the input signal, as opposed to the recorded playback. Video people
refer to source as EE for Electronics to Electronics.

363
Source Follower

Source Follower A circuit element conguration, usually using an

opamp, which has a gain of 1 (unity). The input impedance of a source
follower is very high, and the output impedance is very low. This makes
the circuit useful for driving audio lines and other low impedance devices.
The source follower is similar in its action on signals to the cathode fol-
lower and emitter follower.
Spaced Microphone Stereo See a-b stereo.
Spaced Omnis See a-b stereo; omnidirectional.
Spaciousness A subjective quality of reproduced music that is related to
envelopment in room acoustics. Spaciousness comes from the ability
of the recording to capture the acoustical effect of the room in which the
music was being played.
SPARS Acronym for Society of Professional Audio Recording Studios. This
group is responsible for the so-called SPARS code. The three-letter SPARS
code appears on records and compact discs and indicates whether the
original recording, the editing, and the nal production were done by ana-
log or digital means. For instance, AAD means the original recording was
analog, the editing was done before digitizing, and the nal product was
digitally mastered. The use of the SPARS code has come under re in recent
years, usually because it does not accurately indicate the number of A to
D and D to A conversions the signal has undergone in its travels from micro-
phone to mixing console to tape recorder to the CD mastering facility. The
Society of Professional Audio Recording Services met during an Audio Engi-
neering Society conference and decided to recommend that the SPARS code
be discontinued. They issued a statement that said, in part, But by that
time, the digital/analog technical scene had become so cluttered with con-
versions and algorithms for interface as to resemble rocket science, and many
felt the SPARS code too simple to carry enough information to be mean-
ingful. SPARS withdrew endorsement of the code in 1991. But many labels
continued to use it, and the organization renewed its endorsement of the
code in 1995.
Spatializer A proprietary digital signal processor that modies a stereo
input signal to simulate a stereo image wider than the speaker spacing,
developed by Desper Products, Inc., a subsidiary of Spatializer Audio Labs,
Inc. Widely licensed in both the consumer audio and multimedia comput-
ing markets, the Spatializer process is normally used as a postprocessor. It
claims to place sounds in front of the listener in an arc of 180 degrees, with
excellent imaging and delity.
S/PDIF Sony/Philips Digital Interface or Sony/Philips Digital Interface For-
mat, a serial digital audio interconnection scheme found on consumer
equipment. It uses RCA plugs and jacks with 75-ohm coaxial cable, or ber
optics. The format was eventually recognized as IEC958, and later as IEC
60958-3 when other formats such as AES3 (IEC 60958-4) were included by
the iec specication.
SPDT Single Pole Double Throw. This refers to a switch that has a single
current path but is capable of routing the input to either of two outputs.

364
 Spectral Recording

Speaker See loudspeaker.

Speaker Sensitivity A number that quanties the efciency of a speaker
or speaker system, usually expressed as spl (sound pressure level),
produced by the speaker at one meter directly in front of the speaker when
driven by one watt of input power. The higher the number, the louder it
will sound at a given volume control setting. Speaker sensitivity is not a mea-
sure of the quality or fidelity of a speakerit only refers to the speakers
loudness.
Speakon A trademark for a high-power audio connector from the Neu-
trik company used for loudspeakers. The previous lack of a good solution
has led to wide industry acceptance. Speakons can have several contact
options, allowing for easy biamping for example, and a locking collar.
Spectral Contamination A sensitive technique, developed by Gary Sokolich
and the late Deane Jensen in the 1980s, for measuring cross-modulation
products produced by nonlinearities in a dut that is passing many fre-
quencies at once. The DUT is excited at many simultaneous well-dened
frequencies, and the distortion components are measured at frequencies
where no excitation energy exists. Their method detects distortion com-
ponents as low as 110 dB below the level of the signal.
The test signal can have many forms but is usually a collection of evenly
spaced sine waves spread over a broad frequency range. One signal of this
type has frequency components from 10 kHz to 25 kHz spaced 120 Hz
apart. All these frequencies interact in the DUT to produce energy in the
0-Hz to 10-kHz band, where no test signal existed. This frequency band
in the output of the DUT is passed into an fft analyzer for spectrum
analysis.
It is interesting that this technique is able to separate linear errors, such
as frequency response, magnitude, and time delay errors, from the non-
linear errors. The results are said by the inventors to correlate well with lis-
tening tests. At present, there are no established standards for spectral
contamination measurement.
Spectral Recording, SR A somewhat complex type of compander sys-
tem invented and marketed by Dolby Laboratories that achieves a greater
degree of noise reduction than Dolby A, B, or C systems and also results
in reduced distortion in most cases. The system was described by Ray
Dolby at the Audio Engineering Society convention in November 1986. It
is intended for use with professional-quality analog recording media, such
as tape recorders and motion picture sound, rather than digital systems,
whose low distortion and wide dynamic range are already adequate to
the task of recording music.
In the SR system, low-level components in the music signal are com-
pressed in a frequency-selective way during recording, and high-level parts
of the signal at the very high and very low frequencies are attenuated. The
attenuation of the signals at the frequency extremes reduces the distortion
added by the recording medium, and the compression of low-level signals
increases their signal-to-noise ratio. In playback of the recorded sig-

365
Spectrum

nal, these effects are reversed in order to restore the original dynamics to
the signal.
For the system to operate properly, the signal levels in the encoding and
decoding sections must be matched, and this is cleverly accomplished by
the encoder generating a calibration tone for signal level and a band of pink
noise for frequency response matching. The SR system is very well
conceived and engineered, and probably represents the state of the art in
compander systems.
Spectrum When a time-varying signal is subjected to frequency analy-
sis, it is transformed from the time domain to the frequency domain.
The frequency-domain representation of the signal is called the spectrum,
and the time-domain representation is called the waveform. The two
quantities contain the same information, and one can be converted into the
other by a mathematical operation called the fourier transform.
Spectrum analysis, frequency analysis, and Fourier analysis are
synonymous.
Spectrum Shifter See frequency shifter.
Specular According to the OED, Having the reecting property of a
mirror; presenting a smooth, polished, reective surface; of a brilliant glassy
or metallic luster. In acoustics, specular reection is a discrete reection of
a sound where the angle of reection is equal to the angle of incidence. An
example is an echo reected off a large plane surface like a wall. In gen-
eral, specular reections are avoided in rooms designed for music listening
or recording. See Appendix 4 for more details. The opposite of specular
reection is diffuse reection, where the reected sound is scattered in all
directions.
Speech Coil The same as voice coil. This is primarily British usage.
Speech-Music Switch A special type of equalizer that adds a low-
frequency reduction to the signal when in the speech position. The rea-
son is that most broadcast and recorded speech has too much low-frequency
content because of too-close microphone position and proximity effect
when directional microphones are used.
Speech-music switches are often found on sound reinforcement sys-
tem preamplifiers. The low-frequency cut usually amounts to 6 deci-
bels per octave below 100 hertz or so.
Speed Word yelled by the production sound mixer when the production
recorder is in record and up to speed (indicated by ags on a Nagra),
indicating to the camera crew and the assistant director that he or she is
recording. While Nagras get up to speed quickly, and digital machines are
always at speed, the term derives from older Hollywood technology in
which a common motor system drove cameras and lm sound recorders
(originally optical, and later either 1712-mm or 35-mm mag) and sometimes
even turntables for music playback.25

25. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

366
 Splitter

Spider The assembly that holds the voice coil of a dynamic loud-
speaker centered in the magnetic gap. The spider is a corrugated circular
piece of specially treated fabric. The name comes from the early days of
loudspeakers when it was made of a plastic material that resembled the
legs of a spider.
SPL Common abbreviation for sound pressure level.
Splatter A type of distortion of an audio signal caused by hard clip-
ping of the waveform, usually because of the overload of a device. The
term splatter is most often used for the distortion caused by overmodula-
tion of am transmitters. See also asymmetrical limiter.
Splice In reference to magnetic recording tape, a splice is a discontinuity
in the tape itself occasioned by cutting the tape and pasting it back together
with thin adhesive splicing tape in a different sequence. If carefully done,
splices can be inaudible.
Splicer A mechanical device for cutting and taping together magnetic
recording tape. Many designs have been produced, some with various
attachments for cutting the tape and applying splicing tape. There was a
guillotine-type tape splicer on the market for quite a few years that had a
pair of curved blades that trimmed the tape on both sides at the vicinity of
the splice to make the spliced tape a little narrower than 14 inch. The pur-
pose of this was to assure that no sticky splicing tape would overlap the
tape edges and gum up the guides and heads in the tape recorder. The splicer
was called the Gibson Girl, by analogy to the hour glass gure of the Gib-
son Girl style. The OED denes Gibson Girl as follows: A girl typifying
the fashionable ideal of the late 19th and early 20th cents. as represented
in the work of Charles Dana Gibson.
While splicing analog audio tape is quite effective and was essential until
computer editing, it doesnt work very well on videotape due to the way
the signal is recorded and the nature of the video signal.
Splicing Block A small rectangular block of aluminum, or sometimes
plastic, which has a shallow groove designed to hold magnetic tape while
it is being cut and spliced. The splicing block also has a slot to guide the
razor blade used to cut the tape.
Splicing Tape Sticky tape for holding together splices in analog audio tape.
Splicing tape comes in different widths to accommodate tapes of 2-inch, 1-
inch, 12-inch, 14-inch, and 18-inch widths. Splicing tape is slightly narrower
than the tape it is designed to join, and has a special adhesive that is designed
not to slide or bleed out of the area where it is applied. This is very impor-
tant; it is not advisable to use standard Scotch tape for splicing magnetic
tape for this reason. Unfortunately, over decades, all splicing tape seems
either to dry up or ooze anyway.
Splitter A small device used in recording studios to accept a signal from
one device, usually a microphone, and split the signal and feed it to
two or more other devices. Splitters are also used for connecting several
television sets to a single coaxial cable. The simplest type of splitter is
the so-called Y-cord, but most splitters contain series resistors in the signal

367
Spot Microphone

paths to avoid excessive loading of the device being split. This causes the
split signals to be about 6 dB lower in level than the signal being split.
Spot Microphone A microphone purposely aimed toward a certain sub-
ject, like a spotlight. Usually the spot mic augments a stereo microphone
array such as when recording an orchestra. A soloist or soft instrument may
get a spot mic. The concept is essentially foreign in the multitrack, close
microphone world.
Spring Reverb A type of synthetic reverberator that uses a vibrating
spring as the reverberating element.
The rst commercial articial reverberator was introduced in the late
1930s by the Hammond Organ company. It used a somewhat complex series
of springs that were driven by the vibration of a loudspeaker voice coil
that was connected to the organs preamplifier output. This vibration was
sensed by a transducer similar to a phonograph cartridge whose out-
put was mixed with the organ signal as it entered the power amplifier.
The reverberation of the springs was thus added to the organ signal. The
vibration of the springs was damped by their ends being submerged in a
bath of oil.
This system was improved over the years by the Hammond Organ com-
pany, and was the subject of several patents. Later spring reverberators
applied a twisting, or torsional, motion to the springs, which made them
much less sensitive to external vibrations.
Springs are among the least expensive types of synthetic reverberators,
and they do not sound very much like real reverberationbut no synthetic
reverberation really does. See also plate reverberation.
Sprocket A toothed wheel found in motion picture cameras and projectors
to control the motion of the lm. The teeth in the sprocket interlock with
holes in the lm in the manner of a sprocket and chain of a bicycle. At one
time, audio portions of a lm soundtrack were transferred to sprocketed
audio lm known as fullcoat or magstock. This allowed editing while
keeping synch with the picture. Sprocketed magstock players were used
for the sound in theater systems such as Cinerama. Occasionally a sprock-
eted audio recorder would be used just for audio recording, a technique
championed and used by the late recording engineer Bert Whyte, owner
of Everest Records.
Sprocket Run Sometimes the lm in a motion picture projector can become
misaligned such that the sprockets punch a series of holes in the lm along-
side the sprocket holes, damaging the optical sound track. The audible result
of this is a 96-Hz buzz superimposed on the sound. Such a lm is beyond
repair.
SPST Single Pole Single Throw. This describes a switch that has only one
current path through itself and only on or off positions.
Sputter Microphone diaphrams are often made of polyester (trademarked
Mylar by DuPont) coated with gold a few molecules thick. Gold is used
because of its inert nonoxidizing character and its ductility. The coating

368
 Square Wave

Square Wave

process is called sputtering, and the result is a sputtered diaphragm. The

Mylar is placed in a vacuum chamber, and a small piece of pure gold is
heated until its atoms boil off and impinge on and stick to the Mylar.
SQ See quadraphonic sound.
Square Wave A square wave is a periodic waveform that consists of a
fundamental and all the odd-numbered harmonics. The harmonics
gradually decrease in amplitude as frequency rises, and they are in
phase with the fundamental.
The square wave, along with the sine wave, is one of the classic test
signals used for evaluating audio device performance. It contains energy
up to at least the twentieth harmonic, so an amplifier that passes a 1,000-
hertz square wave without distortion has a bandwidth of at least 20
kilohertz. Because the shape of the wave is sensitive to the phase of the

369
Squawker

harmonics, the phase accuracy can also be determined. The square wave
is a good test signal because it exercises the device at many frequencies at
the same time, as does music.
Squawker An unattering name sometimes given to mid-range loud-
speaker units in three-way or more systems.
Squelch A circuit, commonly found in two-way radio systems, that mutes
the audio signal when the RF carrier is not present. This prevents noise
from being received when the transmitter is not operating. Squelch circuits
are also used in some audio devices to reduce the output noise in the absence
of an input signal. See also muting.
SR See spectral recording.
SRC Stereo Reception Control, a proprietary system used in Mitsubishi FM
car radios in which the stereo separation is reduced in response to noise
and distortion caused by multipath reception.
Because of the short wavelengths used in FM broadcasting, and the fact
that radios in cars are in motion, reected signals cause rapid variations in
received noise, or picket fencing. The SRC circuit senses the noise level
and gradually blends the stereo channels into monaural as the noise
increases. Several other manufacturers use somewhat similar techniques.
See also sampling rate converter.
SSFS Stereophonic Sound Film System, developed during the 1930s by Har-
vey Fletcher and others at the Bell Telephone Laboratories and rst pub-
licly demonstrated in New York in 1937 and 1939. See also fantasound.
The SSFS was an extremely innovative experimental system, using four
optical tracks on a separate 35-mm lm. Three of the tracks carried
stereophonic sound. The fourth, or control, track controlled the gain in
the sound tracks. This was done by recording a mixture of different-fre-
quency tones on the control track and, in playback, separating these tones
with lters and using their individual levels to vary the amplication in
the appropriate channel. The system was thus a three-channel compan-
der and was able to deliver a dynamic range of 80 dB. The optical track
itself was only capable of about 50-dB dynamic range. The original stereo
tracks were compressed in the recording process to compensate for the
later expansion.
SSFS was a remarkably good system considering when it was developed.
It was capable of 20 Hz to 14 kHz response with quite low distortion.
Stage In an amplier circuit, a unit of amplication, such as provided
by a transistor or vacuum tube, is called a stage, or sometimes a gain
stage. There is no xed amount of amplication provided by a single stage;
some stages may provide 40 decibels of gain, while others may provide
less than 1 dB.
Stamper The stamper is the master record from which records are pressed.
The stamper can be made by direct plating onto the acetate master but
is usually made from a mother in the technique known as three-step
processing. About four or ve hundred records can be pressed with one
stamper if it is not damaged. After this number, wear on the stamper

370
 Standard Tuning Frequency

decreases the quality of the pressings. Normally, only one good-quality

stamper can be made from an original acetate, so if more than ve hundred
records are to be made, three-step processing must be used.
Standards In order to ensure interchangeability of such components as
records among record players, tapes among tape recorders, etc., they must
conform to certain shared characteristics. Examples are record turntable
speeds, tape speeds, tape widths, tape track widths, and playback equal-
ization curves.
Many of these standardized values are not signicant in themselves; the
only important thing is the existence of the conformity between manufac-
turers. For instance, record rpm might just as well be 30 rpm or 35 rpm as
3313 rpm and no difference in the quality would be noticed. Other types
of standards, however, are selected to optimize the performance of certain
types of products or processes, and they no longer apply when the prod-
uct or process has become obsolete. An example is the standard nab record-
ing and playback equalization curve for tape recorders, which was adopted
by the NAB in 1954. It made optimum use of the then-available recording
tape, but new technology has revolutionized tape formulation since 1954.
A new equalization scheme would greatly improve the signal-to-noise
ratio in tape recorders, but it would make all existing recorders and tapes
obsolete.
This points up the perennial problem of standardization. Standards
should be set early to allow exploitation of the technology, but they should
be set late to take maximum advantage of that technology. Some feel that
standards are set too early in many cases. With the public crying for
improved technology, and the manufacturers anxious to take advantage of
the new market, the temptation is to standardize early and rush to market.
with new products, even before the engineering is completed. Probably the
worst case of poor planning in this regard was the birth and quick death
of commercially available quadraphonic sound, where products, both
hardware and software, were heavily promoted while the industry was in
the midst of great turmoil over standardization. Even the basic theory of
surround sound was not well established, as system after system was
promoted, sold, and made obsolete almost overnight. Another example is
the standardized sampling rates of digital audio systems, which are
thought by many to be too low for optimum performance.
American organizations interested in the determination of standards are
the ansi (American National Standards Institute), the ftc (Federal Trade
Commission), the nab (National Association of Broadcasters), the smpte
(Society of Motion Picture and Television Engineers), the aes (Audio Engi-
neering Society), and the asa (American Standards Association). There are
correspondingly many such groups in Europe. With so many standardi-
zation groups, there is inevitable overlap and disagreement in some areas.
Standard Tuning Frequency The internationally agreed upon standard for
musical instrument tuning is 440 Hz for the note A4. The frequencies for
the equal tempered scale based on A-440 are shown in the following table.

371
Standing Wave

Note Frequency Note Frequency Note Frequency

B2 123.47 B5 987.77 B8 7,902.10

# 116.54 # 932.33 # 7,458.60
A2 110.00 A5 880.00 A8 7,040.00
# 103.83 # 830.61 # 6,644.90
G2 97.999 G5 783.99 G8 6,271.90
# 92.499 # 739.99 # 5,919.90
F2 87.307 F5 698.46 F8 5,587.70
E2 82.407 E5 659.26 E8 5,274.00
# 77.782 # 622.25 # 4,978.00
D2 73.416 D5 587.33 D8 4,698.60
# 69.296 # 554.37 # 4,434.90
C2 65.406 C5 523.25 C8 4,186.00
B1 61.74 B4 493.88 B7 3,951.10
# 58.27 # 466.16 # 3,729.30
A1 55.00 A4 440.00 A7 3,520.00
# 51.91 # 415.30 # 3,322.40
G1 49.00 G4 392.00 G7 3,136.00
# 46.25 # 369.99 # 2,960.00
F1 43.65 F4 349.23 F7 2,793.80
E1 41.20 E4 329.63 E7 2,637.00
# 38.89 # 311.13 # 2,489.00
D1 36.71 D4 293.66 D7 2,349.30
# 34.68 # 277.18 # 2,217.50
C1 32.70 C4 261.63 C7 2,093.00
B0 30.87 B3 246.94 B6 1,975.50
# 29.14 # 233.08 # 1,864.70
A0 27.50 A3 220.00 A6 1,760.00
# 25.96 # 207.65 # 1,661.20
G0 24.50 G3 196.00 G8 1,568.00
# 23.13 # 185.00 # 1,480.00
F0 21.83 F3 174.61 F6 1,396.90
E0 20.60 E3 164.81 E6 1,318.50
# 19.45 # 155.56 # 1,244.50
D-0 18.35 D3 146.83 D6 1,174.70
# 17.32 # 138.59 # 1,108.70
C0 16.35 C3 130.81 C6 1,046.50

Standing Wave A phenomenon in room acoustics whereby a sound is

reected back and forth between two parallel surfaces, such as two side-
walls. The sound waves interfere with one another to produce a series of

372
 Stems

places where the sound pressure level (SPL) is high and another series
of places between them where the SPL is very low. It is as if a sound wave
were stationary in the space between the surfaces.
A standing wave exists between surfaces only when the frequency is
such that the distance between the walls is an integral multiple of one-half
the wavelength. For a given distance, there will be many frequencies that
will generate standing waves, each an integral multiple of the lowest, or
fundamental, frequency. Standing waves are created by room modes,
which are modes of vibration of the air in a room.
Standing waves are always detrimental to the acoustics of a room, and
are avoided by careful design in music listening rooms. They are the cause
of irregularities in the bass response of most home sound systems. In record-
ing studios, special constructions called bass traps are sometimes used
to add low-frequency absorption in order to prevent or reduce the forma-
tion of standing waves. See also eigentones.
All musical instruments with strings or columns of air operate on the
principal of standing waves. They set up standing waves that cause vibra-
tions of the instruments and radiation of sounds at all the resonant fre-
quencies of the standing waves.
State-Variable Filter A type of active lter that uses resistors, capac-
itors, and three opamps to provide simultaneous low-pass, high-pass, and
bandpass functions of the input signal. The state-variable lter has the
advantage that the q and the cutoff frequency are independently adjustable
by varying resistance values in the circuit. It is a simple matter to arrange
voltage control of these two parameters, and this is often done. The state-
variable lter exactly simulates the response of an equivalent RLC lter and
has been used for some time in analog computer circuitry.
Static Any high-frequency intermittent noise is called static. The term comes
from the early days of radio broadcasting when distant lightning strokes
caused such noise in reception.
Static is distinct from such continuous noises as hum or buzz, and can
be difcult to diagnose because of its intermittent quality. It can be caused
by loose or oxidized connections in audio devices as well as from rfi.
It is curious that it was called static in as much as it is a dynamic type
of noise.
Stems The three or more nal components of a stereo lm mix, usually
comprising three multichannel mixes, one each of dialog, music, and
sound effects that, combined, make up the nal mix of a lm. Minimal
(hopefully no) additional level changes, equalization, etc., should be
needed to create a printmaster, although of course a 6-track printmaster
will sometimes have different requirements than a print master for stereo
analog uses.
The separation of elements afforded by stems allows Music and Effects
(M&E) mixes to be easily derived from the original stereo mix. For this rea-
son, dialog stems are sometimes comprised of multiple centers.
The word stem should not be used for any other element prior to the

373
Ster-Bin

nal mix masters; it is a common mistake to refer to the various premixes

as stems.26
Ster-Bin A circuit that converts a stereophonic signal into a quasi-
binaural signal to make it more suited to listening via headphones.
If conventional stereo music is heard through headphones, the subjec-
tive effect is usually one of great spaciousness and exaggerated dynam-
ics and stereo separation. The effect may be pleasant with some types of
music, but is not at all natural; that is, it does not resemble the sound of
live music. Binaural recordings, on the other hand, are eminently suited to
headphone listening by their very nature. In order to simulate binaural from
stereo, crosstalk that varies with frequency combined with a time delay
must be introduced between the signals. When listening to loudspeaker
reproduction of stereo, each ear hears both channels, but the right-channel
signal heard by the left ear is modied by diffraction of the sound about
the head. This causes a relative time delay between the left and right ears
hearing of the signal, and also causes a modication of the frequency
response of the left ear signal compared to the right.
The Ster-Bin circuit introduces this frequency-dependent crosstalk and
time delay. Stereo recordings heard through the Ster-Bin sound more com-
pact and realistic, and cause less listener fatigue than if it is not used,
although they do not sound as realistic as true binaural recordings. Some
simple systems have been built that introduce crosstalk between the chan-
nels without the equalization and time delay, and they are sometimes
called cross-feed circuits. Their effect is not nearly as convincing as that
of the Ster-Bin.
The Ster-Bin was invented by Benjamin Bauer of CBS Laboratories, and
was marketed for a time by the Jensen company. Improvements in the sys-
tem were suggested by M. V. Thomas of Boston University. There seems to
be little interest in it today.
Stereo See stereophonic.
Stereophoner A commercially unsuccessful device invented in the late 1950s
by the eminent orchestra conductor Hermann Scherchen for producing a
pseudostereo effect from monaural recordings. It embodied time-delay
and equalization networks to derive two different signals from the sin-
gle source. It was intended to provide a means for improved listening to
the vast number of monaural recordings in existence at the time that com-
mercial stereo records were being introduced.
Stereophonic In common usage, stereo has come to mean any sound sys-
tem with two loudspeakers. Stereophonic, however, refers to a sound
system that provides the listener with an illusion of directional realism,
regardless of how many channels are used.
Probably the rst large-scale test and demonstration of stereophonic
reproduction of music was carried out by Harvey Fletcher and his cowork-

26. This entry is copyright 19992003 by Larry Blake and is reprinted with permission.

374
 Stereophony

ers at the Bell Telephone Labs in 1933. In this historic experiment, the
Philadelphia Orchestra, playing in Philadelphia, was picked up by three
microphones carefully placed in front of the proscenium. The microphone
signals were amplied and sent over telephone lines to Constitution Hall
in Washington, D.C., where they were further amplied and sent to three
specially designed high-power loudspeaker systems on the stage. Leopold
Stokowski was the conductor of the orchestra, but on this occasion, he was
in Constitution Hall, where he controlled the volume of the three repro-
duced channels. In this way he controlled the musical dynamics of the
performance in a way he could not do simply by conducting the orchestra.
As part of the demonstration, a number of experiments were performed
by Dr. Harvey Fletcher, who had charge of the development work. On the
stage in Philadelphia, a carpenter hammered and sawed, while conversing
with his helper; a soloist sang as she walked across the stage; and nally a
trumpeter in Philadelphia played antiphonally with another trumpeter on
the opposite side of the stage in Washington. It was only when the curtain
that hid the loudspeakers was raised that the audience in Washington could
believe that what they had heard had not happened on the stage before them.
To present this demonstration in 1933 was a monumental task, for suit-
able amplifiers and loudspeakers did not exist, and had to be designed
and built just for the event. Great care was taken to ensure high-quality
reproduction, and I am sure the standards of excellence set then are sel-
dom equaled today. In any event, the experiment established the fact that
three independent microphones, ampliers, and loudspeakers are sufcient
to provide a convincing illusion of spatial perspective.
So successful was this production that when a few years later Dr. Sto-
kowski was asked to direct a local orchestra in the Hollywood Bowl, he
insisted that a stereophonic system be installed to reinforce the music. The
Bell Laboratories had pioneered the development of sound reinforcement
systems and, with the assistance of Electrical Research Products, Inc., un-
dertook this project as a further research in that art. Again, the results were
successful; some 25,000 people in the open air were able to hear a program
of vocal and instrumental music with ease.
The fact that present-day stereo consists of only two channels is simply
due to cost considerations and the relative ease of recording two channels
of sound in the single groove of a record. With the advent of the digital
compact disc, there is no such theoretical limitation, and the design could
have been done so that three channels rather than two could be recorded.
This would have resulted in a true improvement in the realistic reproduc-
tion of music in the home, rather than the marginal improvement provided
by using digital technology with only the two channels of conventional
stereo.
Stereophonizer A device introduced by Kintek that converts a monaural
signal into a 3-channel stereo effect. It is used in certain movie theater sound
systems to add width and depth to the sound eld in the theater.
Stereophony A term coined by Western Electric Corporation, rst used in

375
Stereosonic

the 1920s to mean the transmission of sound from one place to another by
a stereophonic system. Stereophonic audio transmission actually dates
from 1881. See also ader, clement.
Stereosonic A term, mostly used in Britain, for intensity stereo, espe-
cially the technique of using two coincident figure-8 microphones.
The term was also used in 1950 by Murray Crosby for a multiplex FM-
stereo system of broadcasting. See also fm stereo.
Sticky Shed A malady that affects certain magnetic tapes, especially those
with back coatings and urethane binders. When stored for many years,
the binder that holds the magnetic oxide material together becomes soft
and sticky, and when the tape is played, some of the oxide rubs off onto
the tape recorder tape guides and heads. The oxide adhering to the heads
prevents the tape from contacting the head near and over the gap causing
the high-frequency response to be degraded and the tape to squeal, and
leaving the sticky material on everything that touches the tape. The prob-
lem has been studied for some years, and luckily, it was discovered that
if the tape is roasted in a 130-degree oven for several hours, the binder
regains its integrity and the tape behaves almost as if it were new for a
month or so.
Stinger In motion picture music, a short emphasized passage that calls atten-
tion to an important event in the story.
Stomp Box A oor-mounted effects device for use in the cable between a
guitar and an amplifier. It has foot-operated switches to turn the effects
on and off.
Stopband The frequency band that is not passed by a filter is called
the stopband, as opposed to the passband. A lter can have more than
one stopband; for instance a bandpass lter will have a high stopband
and a low stopband, above and below the lters passband, respectively.
Storage Cell Any type of electric cell that can be recharged after being
discharged. Examples are the nickel-cadmium cell and the lead-
acid cell.
Strapping The interconnection of transformer windings is often referred
to as strapping. In many audio and some power transformers, there
are several primary and secondary windings. They can be connected in
series to increase the impedance or they can be connected in parallel
to reduce the impedance. These connections are frequently made by short
pieces of wire between the screw terminals of a terminal strip. The wires
are called straps, hence the term strapping.
line level inputs and outputs of professional audio devices often have
transformers with two windings. When connected in series, the impedance
is 600 ohms, and when connected in parallel, the impedance will be 150
ohms. This offers the user the choice, depending on application. Power trans-
formers sometimes have a similar arrangement so they can be strapped for
120 volts or 240 volts input.
Streaming Transmission of digital audio or video data to a client (the lis-
tener or viewer) in a way that requires it to be monitored in real time as it

376
 Stylus

arrives. An example is listening to a radio station via the internet. If the

stream of data stops, so does the program. The alternative is to require the
client to download the entire media le rst. Streaming, however, can be
more efcient at distributing the media to many clients at the time of their
own choosing, and affords some copy protection.
Strike To remove sets, props, etc., from a theater stage or sound stage. Sim-
ply put, it means to clear the stage in preparation for the next act or next
production. The term is also used to refer to a single or several items in the
set, like strike the piano.
Stroboscopic Disc A disc made of cardboard or other plastic typically with
three circles of dots or bars around the periphery. When the disc is placed
on a record-playing turntable and the disc illuminated with uorescent or
neon light, one of the circles of dots appears stationary if the turntable speed
is correct. The light ashes at the rate of 120 ashes per second and the time
between ashes is just enough for each dot to move one space to take the
place of the previous dot. The three circles are for 78.26 rpm, 45 rpm, and
3313 rpm, and they contain 92, 159, and 216 dots, respectively. Some turnta-
bles have the stroboscopic circles of dots on their underside, with a small
observation mirror and neon lamp.
If the turntable has variable speed (which means it is probably never
running at the correct speed), the stroboscope is a real help in setting it. If
the turntable does not have variable speed, the stroboscope can be used to
determine if it is defective or needs lubrication.
In most European countries, the power line frequency is 50 Hz rather
than our 60 Hz, and the stroboscopes have fewer dots.
This strobe principle works whenever the light and rotation rates are
known, and at least one tape recorder, the Nagra 4 series, had stroboscopic
marks on the top of one the rotating tape rollers.
Stylus The business end of the phono cartridge is the stylus. It consists
of a small arm called the cantilever and the stylus tip, which makes con-
tact with the record groove. The ideal shape for a stylus tip is a subject of
much controversy. The rst styli were conical, with a spherical section at
the end.
The conical stylus does not do a perfect job of tracing the stereo record
groove due to the pinch effect, and it has a very small area of contact
with the groove. The elliptical stylus was introduced to reduce this effect.
The top of the elliptical stylus is narrower in the front-back direction than
in the side-to-side direction, so it ts better into the groove. The elliptical
stylus, however, also has a very small area of contact with the groove, caus-
ing relatively rapid wear.
The shibata stylus is somewhat like an elliptical, except the shape is
such that the contact area on each side of the groove is a small line rather
than a point. It is capable of excellent high-frequency response, and was
in fact developed for use in playing CD-4 quadraphonic records, which
had a high-frequency carrier signal that had to be picked up. Since that
time, several styli have been developed with quite complex shapes, each

377
Subcarrier

claiming to be the best compromise for low distortion, low wear, and
best frequency response. One of these that was popular was known as the
Van den Hul stylus after its inventor.
There is a certain (small) group in the audio eld who insist the old con-
ical stylus is best.
Subcarrier In multiplex radio broadcasting, the subcarrier is an ultra-
sonic signal modulated by another signal and then used to modulate
the transmitted carrier along with the main modulating signal. The sub-
carrier is normally frequency modulated, but could also be amplitude mod-
ulated. See also fm stereo.
Subcode In digital audio systems, particularly cd and r-dat, additional
data are interleaved with the audio information, which carry synchroniza-
tion and user information such as tags and comments that are independent
of the audio data. CD subcodes consist of digital data included on the CD
that contains such information as track numbers, track playing times, copy-
right information, and copy inhibit codes, etc. In the compact disc for-
mat, eight additional bits (one byte) containing no audio information are
added to each frame of data. This means a byte of information is available
from the disc every 136 microseconds. Each bit in the added byte is given a
one-letter name: P, Q, R, S, T, U, V, and W. Thus eight separate subcodes can
be recorded on and recovered from the CD. So far, only subcodes P and Q
are used; the P subcode is used for the pause signal between musical tracks
and at the end of the last track and the Q subcode tells the player if the record-
ing is two- or four-channel. (No quadraphonic player is yet available,
however.) The Q subcode also contains timing information about the tracks
and identies the country of origin and date of the recording. The P and Q
control bytes also contain the timing information that allows the CD player
to cue instantly to the beginning of each selection, display the selections
number and running time, and provide a continuous display of elapsed time.
No standard has been dened for the use of the other six sub codes.
The three main types of dat subcodes are Start IDs that mark the begin-
ning of each song, Program Numbers that are the ordinal numbers assigned
to each Start ID, and Skip IDs that cause a player to skip the current Start
ID and go to the next one.
Subgroup A term used in mixing consoles where several input channels
can be grouped together and their output levels controlled by one gain con-
trol that is called a Submaster.
Subharmonic A submultiple, usually one-half, of a fundamental fre-
quency. Sometimes subharmonics are produced by loudspeakers that
have poorly controlled cone resonances. The audible effect is a distor-
tion component one octave lower than the input signal frequency.
Submaster A control on a recording or sound reinforcement console that
controls the level of a mixture, or subgroup, of signals. Several submas-
ters may be fed into a master control for nal level control of the console
output signal. The use of submasters makes it easier to handle a large num-
ber of input signals.

378
 Superheterodyne

Submini Plug A very small version of the phone plug, with a diameter of
2.5 mm. It is commonly seen on cellular phone earsets and some other audio
and control devices where space is at a premium.
Submix The mixture of signals fed into a submaster control in a record-
ing or sound reinforcement console. The submix is usually a mix of signals
that remains constant over a period of time, and it is convenient to control
it as a single signal.
Subsequent Reverberation Time Similar to the classic reverberation
time except it is measured from the time of arrival of the rst reected sound
rather than from the time the sound source is stopped.
Subsonic Literally, under sound. Actually, subsonic means slower than
the speed of sound, but it is often misused to mean sound having frequencies
below the human hearing range. The proper term for this is infrasonic.
Subtractive Synthesis The technique of generating a desired musical tim-
bre by ltering complex signals generated electronically. Typical wave-
forms used in this way are the square wave and the sawtooth wave, both
of which sound bright and raucous before ltering to subtract, or attenu-
ate, some of the upper harmonics.
Subwoofer A loudspeaker system designed to reproduce the very low
frequencies from about 16 hertz to 100 Hz or so. A subwoofer must be
capable of large amounts of power output because of our ears relative
insensitivity to low frequencies. Use of a subwoofer also allows using stereo
speakers with modest bass response, and our insensitivity to directional-
ity of low bass means only one subwoofer is needed and placement is not
stereophonically sensitive.
Super-Cardioid A modication of the cardioid microphone to reduce
sensitivity to sounds coming from the sides. It is something like a cross
between a cardioid and a figure 8 in that it has a somewhat large rear
lobe. Its advantage is that it reduces the apparent reverberation by about
1.3 decibels if placed at the same location as a true cardioid. It will pick
up the same direct-to-reverberant ratio as an omnidirectional micro-
phone when it is 1.9 times as far away from the source.
Like all directional microphones, it suffers from proximity effect.
Superheterodyne A revolutionary type of radio receiver, invented in the
late 1920s by Major Edwin F. Armstrong, who also developed fm broad-
casting much later. Superheterodyne is a contraction of supersonic heterodyne,
which is actually a misnomer; it should have been called ultrasonic hetero-
dyne. In the standard AM superheterodyne (superhet) receiver, the radio
frequency signal is combined with a local oscillator signal in a
mixer to produce a beat frequency of 455 kHz. The 455-kHz signal is
called the intermediate frequency (IF) and is always the same regardless
of the station being tuned in. This is because the local oscillator is tuned
such that the difference frequency between it and the indicated broadcast
frequency is 455 kHz. The IF is then amplied through several stages and
is detected by a standard detector of one sort or another. The mixer stage
(also call the rst detector) is not a linear mixer as is implied but actu-

379
Supersonic

ally performs a multiplication of the two signals, similar to amplitude

modulation. One of the resulting sidebands is at 455 kHz and is the IF
signal. The actual signal detector is called the second detector.
All modern radios, AM and FM, and analog television sets use the super-
het principle. The IF in a standard FM receiver is 10.7 MHz rather than the
455 kHz used in AM receivers. The main advantage of the superhet design
is that the amplifier stages always operate at the same frequency and so
can be optimized for gain and signal-to-noise ratio performance.
Supersonic Faster than the speed of sound, as in supersonic aircraft. The
term is sometimes mistakenly used to mean sound higher in frequency than
the audible range; the correct term for this is ultrasonic.
Supply Rails The output terminals of the power supply in an audio device
are commonly called rails. Each stage of the device is connected to the
rails via some sort of decoupling circuit. Most modern devices utilize two
supply rails, one at a positive (B+) voltage, and one at a negative (B-) volt-
age, although there may be several rails at different voltages in certain
devices. The common, or ground, terminal of the power supply is called
a ground bus rather than a rail.
Supply Reel The reel on a magnetic tape recorder from which is unwound
the tape to be recorded or played back. It normally resides on the left-hand
side of the machine. The other reel is the takeup reel.
Supraaural Literally, on top of the ear. This term refers to headsets that
rest on the outer ear, or pinna. These headsets are necessarily light in weight,
and do not exclude external sounds very well. See also circumaural.
Surface Mount An electronic circuit board construction technology, using
specially formed electronic components such as resistors, capacitors,
and integrated circuits. The components are designed to have their
electrical contacts soldered directly to the surface of the circuit board, as
opposed to the traditional method of wire leads going into holes in the board.
This technique greatly decreases manufacturing cost as it is highly automat-
able, and greatly increases component density on circuit boards.
Surround The mechanical suspension that holds the outer edge of the cone
of a dynamic loudspeaker is called the surround. Loudspeaker cone sur-
rounds can be made of many different materials, such as corrugated cloth
or paper, polyurethane or other plastic foam, and rubber, and some, called
roll surrounds, have a single large semicircular corrugation. Roll sur-
rounds allow a greater amount of cone motion and are relatively linear in
their springiness, reducing distortion. Most surrounds are treated with
some type of damping material to reduce the effects of cone resonances.
See also spider.
Surrounds In many modern motion pictures, the sound tracks will include
some sound intended to come from behind or beside the viewer. These
channels are called surround channels, and the loudspeakers are called
surround speakers, or simply surrounds. They are also found in many
home theater setups.

380
 S-VHS

One of the aims of quadraphonic sound was to create surround

sound for the listener.
Surround Sound The sound effects in motion pictures that come from the
sides and/or rear of the theater. In the case of the original analog Dolby
optical sound tracks, the surround sound was recorded on the lm by
matrixing the sound effects signal into the two stereo channels using the
qs technique. In the theater the signal was decoded and sent to the rear
and side speakers after passing through a time delay device. The time delay
makes the resulting effect sound more distant, although one also needs to
take the physical distance of the surround speakers into account. Dolby
Stereo movies that were transferred to laserdisc or video tape may contain
the original coded sound track intact, although dvds usually have the sound
tracks remixed for home use.
In the case of stereophonic magnetic and digital sound tracks for movies,
there have been several schemes employed over the years. Some of these
sound tracks were encoded like the Dolby optical sound tracks, but some,
especially 70-mm presentations and all the digital systems, have been
recorded with at least two independent surround tracks, allowing true
stereophonic surround sound. The current 5.1 type surround systems in
theaters and DVD all provide for true stereo surround.
Suspension In a dynamic loudspeaker, the spider and the surround
are collectively called the suspension, or suspensions, because they effec-
tively suspend the voice coil in the gap of the magnet.
SVA, Stereo Variable Area A motion picture sound track on lm system
consisting of two closely spaced optical variable-area sound tracks. SVA
sound tracks have existed since the 1930s, but were never a commercial suc-
cess until Dolby Laboratories applied their type A compander system to
the two tracks, greatly improving the signal-to-noise ratio and reduc-
ing the distortion. The Dolby system also includes matrixing to synthe-
size a third, or center, channel for three-loudspeaker reproduction and a
fourth channel for ambience, or surround sound effects. Most modern
movies with analog stereo sound tracks employ the Dolby SVA system with
their SR companding. See also dolby digital.
S-VHS, Super Video Home System An improved videocassette recorder
that will record a horizontal resolution of 400 lines, whereas the standard
VHS recorder only manages 250 lines. S-VHS accomplishes this with wider
video bandwidth and by separating the chrominance (the part of the video
signal that carries the color information) and the luminance (the part of the
video signal that carries the brightness information), and records them on
separate tracks.
The video cassette recorder (VCR) for home use was pioneered by Sony
and introduced in 1975 as the Betamax system, and the Japan Victor Com-
pany introduced the VHS system in 1976. S-VHS was introduced several
years later. Both Betamax and VHS systems used 12-inch recording tape,
but were not compatible with each other, which is not surprising. There

381
Swarf

ensued a battle for market dominance, which VHS nally won, even
though the Betamax system was believed to have superior quality by many
acionados.
S-VHS tape has slightly different magnetic characteristics than ordinary
VHS tape and has a sensing hole in the bottom of the cassette which S-VHS
machines use to determine standard or S-VHS mode. The popular ADAT
mdm recorders use S-VHS tape for digital audio. All S-VHS machines
include the vhs hi-fi FM analog audio system as well as a linear audio
track. Ordinary VHS machines can use S-VHS tape to no detriment, how-
ever they cannot play back an S-VHS tape unless specially equipped with
a special conversion circuit and then, only at standard resolution.
Swarf British usage for the chip of acetate material removed from the sur-
face of the acetate disc during the cutting process.
Sweep When a test tone (usually a sine wave tone) is smoothly varied in
frequency from low to high, or vice versa, it is called a frequency sweep
or simply a sweep.
The use of the sine sweep is the classical method for measuring the fre-
quency response of a device.
Sweepable EQ An equalizer section whose center frequency can be
adjusted. See also equalizer; parametric equalizer.

SVA

382
 Synthesizer

Sweetening A recording session in which a previously made recording is

augmented by the addition of other instruments, usually strings or a vocal
chorus. The original recording will have lead vocals and instruments as well
as the rhythm tracks. The requirements for sweetening are less demand-
ing on the musicians, and the sweetening can usually be done in a shorter
time at a later date. See also overdub.
Sweet Spot The listener position in front of stereophonic loudspeakers that
provides the optimum listening conditions for tonal balance, stereo sepa-
ration, etc. In the case of transaural systems, which allow loudspeaker
listening to binaural recordings, the sweet spot is likely to be very pre-
cisely dened, with essentially no binaural effect to be heard at other
locations.
In the case of 2-channel stereo, the sweet spot is generally located equi-
distant from the two speakers and preferably directly in front of each
speaker, implying that the speakers should be toed in so each one points
directly to the listener in the sweet spot.
Swinger An analog phonograph record that has been pressed either with
the center hole not exactly centered or with the center hole too big. When
played, such a disc will introduce a wow component at about .5 hertz.
As small an error as .1 mm will cause audible wow to careful listeners.
Swinging Choke An iron-core inductor designed so that its inductance
varies inversely with the amount of current it passes. Swinging chokes are
used as ripple lters in power supplies that are subjected to various load
currents. The varying inductance maintains better regulation of the output
voltage.
Switched-On Bach An album of the music of J. S. Bach arranged for elec-
tronic synthesizer, produced in 1969 by Wendy Carlos, ne Walter Car-
los. This album was probably the most inuential of any electronic music
recording, for it was the rst purely electronic album of real music. It had
a tremendous popularity and greatly assisted in the promotion of synthe-
sizers as serious instruments.
Synchronous Refers to a type of motor used in many tape recorders and
some turntables. A synchronous motor is locked to the 60-hertz (U.S.)
power line frequency and always runs at a constant speed, regardless of
any voltage uctuations. Early AC-powered electric clocks operate with
synchronous motors for this reason. Many European turntables and elec-
tric clocks also use synchronous motors, but with different gear ratios to
run at the correct speed for 50-Hz power.
Synthesizer Strictly speaking, a purely electronic musical instrument
which can be used to imitate many different conventional instruments. The
accuracy of the imitation varies from synthesizer to synthesizer.
The rst synthesizer was built by RCA Research Labs in the 1950s, was
called the Mark II, and is a very large affair, using multitudes of vacuum
tubes. Although primitive in terms of the capabilities of todays solid-
state machines, the original synthesizer established the feasibility of
music synthesis by electronic means, and it has an interesting characteris-

383
Syntony

tic sound. It is still used by some composers to write serious music. It was
donated by RCA to the Columbia-Princeton electronic music laboratory in
New York many years ago, where it still resides.
Modern solid-state synthesizers are seldom used to imitate natural
sounds; it has been demonstrated adequately that in most cases, synthetic
imitations of musical instruments are rarely convincing. Synthesizers are
used much more for generating special musical sounds that could not be
made by conventional instruments. They are also used to modify the tim-
bre of existing instruments, for instance the singing voice.
digital audio technology has been applied to synthesizers, and some
of these units are able to sample or digitize the sound from an actual
instrument by use of an analog-to-digital converter. In digital
form, the sound can then be manipulated in a great many ways, which
would be impossible in the analog domain. Digital synthesizers are very
popular, and have essentially replaced analog models.
Syntony The principle of tuning a radio receiver to the same frequency
as that of the transmitter. The concept of syntony was rst explained by Sir
Oliver Lodge before 1900, but Guglielmo Marconi perfected and patented
the technique in 1900 for use in wireless telegraphy, as radio was called then.
This is one of the most important patents in the history of radio. It may
seem obvious now, but it was a revolutionary idea at the time, for it allowed
a receiver to select between several transmitting stations. Up to that time,
all stations broadcasting were received at once!

T
Tails Out A tape that has not been rewound after being played is said to
be tails out. In general, it is a good idea to store tape recordings tails out
and rewind them just prior to playing them. This helps to reduce the effects
of print-through.
Take A recorded performance, or part of a performance, that is to be kept
for possible use rather than being recorded over is a take. The various takes
are then edited into a complete composite performance. In the case of solo
performers, it is not unusual to record several hundred takes during the
course of a single piece of music.
Take Sheet A sheet of paper on which the recording engineer makes notes
about each take as it is recorded. The notes are useful to speed up the edit-
ing process.
Take-up Reel On a reel-to-reel tape recorder, the tape moves from the left-
hand reel, called the supply reel, to the head stack to the capstan and
pinch roller and then onto the take-up reel.
Talkback Microphone A microphone in the control room of a recording
studio or by the control console in a sound reinforcement system to allow
the engineer to talk to the performers.

384
 Tape Head

Tangency The orientation of an analog magnetic tape head such that the
tape contacts the head for an equal distance on either side of the gap. Poor
adjustment of tangency causes uneven head wear on the sides of the gap.
Tape See magnetic tape.
Tape Head In an analog tape recorder, the transducers that magnetize
the tape in response to the input signal and reconvert the magnetization
into an electrical signal are called the tape heads. For each track, a tape head
consists of a small two-part core of soft iron or other suitable ferro-mag-
netic material with a coil of wire wound around it. Between the two parts
of the core is a very short gap, over which the tape is moved. current in
the coil causes the tape near the gap to be magnetized in proportion to its
strength. Record and reproduce heads are similar in construction, except
that the reproduce head has a narrower gap and many more turns of wire
in the coil. This increases the sensitivity of the head in picking up sig-
nals recorded on the tape.
Another type of tape head is the erase head, which is similar to a record
head, but has a wider gap. The erase head is fed with a very high fre-
quency, high-amplitude signal, usually derived from the bias oscilla-
tor, which effectively demagnetizes the tape just before it reaches the record
head. This eliminates any previously recorded signals from the tape.
Some early tape recorders used a permanent magnet as an erase head,
and although it did a good job of erasing the signal on the tape, it left the
tape in a magnetized condition that greatly increased the noise level of the
subsequent recording. Some inexpensive small portable cassette recorders
still use permanent magnet erase heads.
Helical-scan recorders such as video, adat and r-dat have several small
heads placed near the surface of the rotating aluminum head drum. These
small heads, or pole pieces, write and read the magnetization on the

Tape Head

385
Tape Hiss

tape in slanted tracks. Digital recorders have no erase heads, merely over-
writing old data at tape saturation level. Video recorders usually have a
very traditional analog erase head, but some have a so-called ying erase
head that is mounted in the rotating head drum and allows accurate eras-
ing and editing.
Tape Hiss The residual noise of a blank tape being played; also the tape
noise of a recorded tape.
In general, the tape hiss of a blank tape will be less than that of a recorded
tape because the recording process, including bias and erase, adds some
noise to the tape. Tape hiss is distinguished from modulation noise. The
primary cause of tape hiss is the barkhausen effect.
Tape Lifter A device in an analog reel tape recorder that lifts the tape from
the heads during rewind and fast forward to decrease the head wear and
sound output. The magnetic coating on tape usually consists of iron oxide,
which is an abrasive and will wear the tape heads, especially at high
speeds. Audio cassettes move an assembly of all heads and the pinch roller
away from the tape during the fast wind modes. r-dat type recorders
dont bother lifting the tape, leaving it on the head drum (with a bit less
tension) to enable enough data pickup to provide timecode or ABS time
readings.
Tape Pack The smoothness with which magnetic recording tape is wound
onto the hub of a reel is called the tape pack, or simply the pack. If the tape
wanders around when being wound, the exposed edges of tape are sub-
ject to damage, especially if left in this condition for an extended time. This
is one reason it is desirable to store tapes tails out without rewinding.
Most poor tape pack occurs with high-speed winding.
Tape Speed In analog tape recorders, the velocity with which the tape
moves across the tape head is the tape speed. It is measured in inches per
second (ips) in English-speaking countries, and in centimeters per second
elsewhere. Quite important for analog recorders, the actual tape speed is
often obscured to the consumer for digital devices in preference for run-
ning time.
It is interesting that the tape speed in the rst successful commercial tape
recorder, the German Magnetophon from the 1930s, used a tape speed of
78.5 centimeters per second, which is close to 30.9 ips. This came about by
the use of a capstan motor turning 1,500 rpm (a four-pole motor operat-
ing at 50-hertz line frequency), and a capstan diameter of 1 centimeter.
(According to Friederich Karl Engel, of BASF AG in Germany, the Magne-
tophon motor was not synchronous, and its actual speed was about 1,470
rpm. This gave an actual tape speed of about 30.2 ips.) When Ampex made
the rst American-built tape recorder (the model 200, introduced on April
25, 1948), they chose 30 ips in order to use English units and still be close
to the German practice, which was known to work well. In a similar situ-
ation, the German machine used tape of 6.5 millimeters width, and this is
very close to 14-inch, so Ampex and the 3M company chose 14-inch as a
standard width, and it has remained so ever since.

386
 TEF

As tape was improved in quality and as Ampex gained expertise in tape

recorder electronics and tape head design, they found the same fre-
quency response could be obtained at half the speed, so the next pro-
duction machine operated at 15 ips. It also was able to operate at 7.5 ips
and give acceptable results for noncritical use. The nab, feeling it was about
time to set some standards for tape recording, adopted the 15-ips tape speed
and the Ampex-designed de-emphasis curve as standards. At this time
the rest of the world followed suit; European and Japanese machines were
also standardized at 15 ips and 7.5 ips, although they did not copy the NAB
de-emphasis curve.
With continued improvement in tape formulation and tape head design,
comparable performance in frequency response was attainable at 334 ips
and even 178 ips. The Philips company, in setting the standards for the cas-
sette recorder, chose 178 ips as the speed, and so it remains to this day.
It is important to note that although the successive halvings in tape speed
allowed the frequency response to be maintained, other parameters such
as headroom (especially at high frequencies) suffered and distortion
increased. For this reason, professional-quality analog tape recording is done
almost exclusively at 15 ips, with some manufacturers returning to 30 ips.
In analog audio tape recording, the tape speed must be very accurately
controlled so the reproduced pitch of the music is the same as the pitch of
the source, and the uniformity of the speed is also very crucial (See flut-
ter and wow). In digital audio recording, the actual tape speed is not nearly
as important because the audio signal is reconstructed from the digital data
on the tape, and is not directly related to the speed of the tape. See also head
bump.
Tape Tension Tension is pulling force, and in any tape recorder, tension on
the tape is a critical operating factor. Tension must be just right to maintain
proper head contact and mechanical control of the tape. Elaborate mechan-
ical and electronic means have been employed to control tape tension.
Tape Weave An improper gradual up and down motion of magnetic tape
as it passes over the tape heads in a tape recorder, causing tape skew. It
is usually caused by the tape not having been properly slit, but it can also
result from worn tape guides.
TDFD See total difference frequency distortion.
TDS See time delay spectrometry.
TEF Originally a trademark of the Crown company that stands for Time-
Energy-Frequency and refers to a measurement instrument based on time
delay spectrometry.
It is possible to display a family of curves that graph the relationship of
three different parameters of a signal. A familiar frequency response
curve plots two parameters: magnitude of the signal as a function of fre-
quency. If a series of these curves are plotted next to each other on a page,
and if each one is measured at a slightly different time, the relationship of
frequency, response, and time can be visualized. The curves merge into a
three-dimensional surface that is represented in perspective on the paper.

387
Telephone Hybrid

If instead of simply plotting amplitude response versus frequency and

time, the actual energy content of the signal is plotted, the TEF surface is
obtained. Such a representation of a signal reveals much information to the
trained eye. The performance of a loudspeaker during the time just after
a signal is applied to it is an example of the use of TEF, and these meas-
urements of loudspeakers reveal that various frequencies are produced at
different times; in other words, the time delay inherent in a loudspeaker is
a function of frequency.
It is also possible to obtain the TEF measurement by means of a dual-
channel fft analyzer.
Telephone Hybrid A device designed to couple the audio of a telephone
line to a broadcast or recording system, while keeping them electrically iso-
lated and protected from each other. See also phone patch.
Telescope A device having a relation to the eye similar to that of the tele-
phone to the ear, allowing distant objects to plague us with a multitude of
needless details. Luckily, it is unprovided with a bell summoning us to the
sacrice.A. Bierce.
Television Interference, TVI Analog television broadcast stations trans-
mit very large amounts of energy in the form of synchronizing pulses that
are used by the TV receiver to determine when each eld of the picture is
to start. These synch pulses, as they are called, occur at almost exactly
60 per second in ntsc, the same as the AC power line frequency. (Actu-
ally, the line frequency is a little different from the synch frequency, but it
is very close.)
If the transmitted TV signal is picked up by audio equipment, usu-
ally by wires and cables leading to amplifier inputs and outputs, and if
the signal is detected, or rectied, by the rst transistor or tube it en-
counters, the audio signal will be contaminated by the synch pulses and
will be heard as a 60-hertz buzz. Very often, the amount of buzz heard
is dependent on the exact location of the cables, which act like antennas. If
such a buzz occurs in a sound system, and it changes with movement of
the cables or with a person walking about in the room, you can be sure the
cause is TVI.
There are many cases where hum exists in sound systems and is blamed
on the sound system components resulting in heroic efforts to eliminate it,
but elimination of TVI can be very difcult. It involves ltering out the rf
signal before it reaches any active component like a transistor. The signal
can enter the equipment through the output wires or the power cord, as
well as through the input wires.
Telharmonium The rst purely electrical musical instrument, invented by
Thaddeus Cahill in 1903. It worked on the same principle as the Hammond
organ, introduced in 1935, using a group of alternators or AC generators,
that produced all the frequencies of the musical scale. Unlike the Hammond
organ, which used conventional amplifiers and loudspeakers to pro-
duce the sound, the Telharmonium had no ampliers. (The triode vacuum
tube was not invented until 1906.) The tones were produced at kilowatt lev-

388
 Termination

els by the alternators. These signals were controlled by a keyboard, and

switches allowed various tones to be mixed together to make complex wave-
forms. Cahills idea was to transmit the resulting music over telephone lines
to subscribers, but he had few takers and eventually went bankrupt.
Temperament In the tuning of a musical instrument to a scale, tempera-
ment is the deliberate mistuning of pure or just intervals so the various
frequency ratios between notes of the scale are compatible with perfect
octaves. See Appendix 8.
Temporary Threshold Shift, TTS An upward shift in the threshold of
hearing caused by exposure to a loud sound. The human ear is remark-
ably tolerant of abuse, and it has several ways of protecting itself from dam-
age from very loud sounds. One of these is to reduce its sensitivity,
causing the hearing threshold to shift upward. If the ear is repeatedly
exposed to very loud sounds over a period of years, the temporary shifts
will become permanent, and this is permanent hearing loss, or a perma-
nent threshold shift.
Tennis Shoe Remote In the making of some remote recordings, especially
of classical music, often the microphones must be hung from the ceiling,
or from some superstructure over the stage. This requires much running
back and forth and up and down, stringing microphone cables, suspension
cords, and power cords, etc. Such a recording session is called a tennis shoe
remote.
Tension Ring Many microphones use a round diaphram of thin plastic or
metal, that must be precisely stretched and held taut and stable. The ten-
sion ring is the circular piece that uniformly draws the diaphram down over
a cylinder, then holds it in position.
Terminal The point at which the signal either enters or leaves an audio
device is a terminal. Strictly speaking, terminals are screws under which
wires are placed, but all types of audio connectors are also commonly
referred to as terminals
Terminal Impedance Terminal impedance is the impedance seen from
either the output or the input terminals of an audio device such as an equal-
izer when connected to their rated impedance. The device is said to be ter-
minated when connected to the impedance it was designed to work with;
sometimes the terminal impedance matches the source impedance, but more
often not. See also impedance; Appendix 11.
Terminal Strip A series of connections, usually screw terminals, arranged
in a line to permanently connect multiple audio lines to such devices as
recording or broadcast consoles. It is essentially obsolete today in favor of
faster methods. When many lines are connected to the same terminal strip,
the individual wires are often laced together with linen lacing cord in such
a way that they are fanned out between the lacing to match the distance
between the terminals. Sometimes the terminal strip was called a fanning
strip for this reason.
Termination A transmission line is said to be terminated if it is connected
to an impedance equal to its characteristic impedance. Under these con-

389
Terrestrial

ditions, the signal will enter the terminating impedance as if it were an

innitely long extension of the line, and no energy will be reected. If the
termination is not correct, a reection of the signal will occur, and this effect
was sometimes heard as an echo in old analog long-distance telephone
conversations. In the case of television, it is important that the antenna lead-
in wire be terminated to prevent reections that will be seen as ghosts on
the screen. Video lines in closed-circuit television systems also need to be
properly terminated for the same reason. The termination is frequently a
resistor of the proper value, usually 600 ohms for analog telephone lines,
300 ohms for TV antennas using twin lead, and 75 ohms for video signals
using coaxial cables.
There exists a certain amount of confusion about the need for termina-
tion of audio transmission systems such as microphone lines. Generally, it
is not desirable to terminate a microphone line with an impedance match-
ing the output impedance of the microphone, because the signal-to-
noise ratio will suffer. This is because the termination loads the line and
reduces the signal voltage level, which is what the microphone pream-
plifier is sensitive to. If the microphone line were many miles long, it would
be necessary to terminate it to prevent reections, but in relatively short
lines and at audio frequencies, reections do not cause any trouble. See also
balanced line; matching.
The term termination is also (mis)used to indicate the proper resistance
to which an audio device is intended to be connected. The 47,000-ohm input
resistance of phono preamps is an example. The phono cartridge requires
the 47,000 ohms for proper damping, but this is not the same thing as a
line termination.
Terrestrial The word terrestrial as used by TV people means conventional
wireless TV transmission. This is to differentiate it from satellite or cable
transmission.
Tertiary A third winding in an output transformer, in addition to the nor-
mal primary and secondary, sometimes used in power amplifiers to
provide a negative feedback signal. The classic McIntosh tube ampliers
were some of the rst to use tertiary winding.
Tesla, Nikola A Serbian-American scientist who conducted many experi-
ments in the transmission of high-frequency radio waves. Had he added a
detector to his apparatus, he would be credited as the inventor of radio.
He invented the Tesla coil and many other electrical devices and spent many
years of his life attempting to transmit electric power over long distances
without wires. He was a great champion of alternating-current power dis-
tribution as opposed to direct current, which Edison favored. Tesla is respon-
sible for the establishment of 60 Hz as the standard frequency for power
distribution in the United States. He also invented the induction motor,
which is the type of motor used today almost to the exclusion of all other
types. In his honor, the unit of magnetic ux density is called the tesla. It
is equal to 10,000 gauss, which is the older unit of measure.
Tetrode A vacuum tube containing four elements, as opposed to the tri-

390
 3:2 Pulldown

ode, which contains three. The extra element in the tetrode is a screen grid
placed between the control grid and the plate. The screen has a high pos-
itive voltage on it, and it increases the efciency of the tube, but at the
expense of higher distortion and a higher plate impedance. Most
modern power output tubes used in audio ampliers are tetrodes with
added beam-forming plates. See beam power tube.
THD See harmonic distortion.
Theremin An unusually clever electronic musical instrument invented in
1919 by Leon Termen, a Russian migr and engineer living in New York.
He changed his name to Leon Theremin in the 1920s. The Theremin had two
antennas, and it was played by moving ones hands nearer to and farther
from them. One antenna controlled the volume of the sound produced by
the instrument, and the other varied the musical pitch. The pitch was contin-
uously variable, and the sound had a somewhat eerie, other-worldly character.
It attracted the attention of a few serious performers for a while, and the idea
was brought up to date by Robert Moog with his modular synthesizer.
The Theremin was the subject of a fascinating documentary movie titled
Theremin: An Electronic Odyssey made in 1993, directed by Steven Martin.
Mr. Theremin lived into his nineties and died shortly after the documen-
tary was released.
Thermal Noise See johnson noise.
Thermionic Emission The process by which the cathode in a vacuum
tube emits electrons by being heated to a red-hot temperature. In Britain,
tubes are sometimes called thermionic valves.
Thiele-Small Parameters Characteristics of direct radiator loud-
speakers that can be used to predict the low-frequency performance of the
loudspeaker under various conditions of bafing (see baffle).
Australian researchers Neville Thiele and Richard Small identied
several parameters that can be used in the analysis of both closed-box and
ported loudspeaker systems. These parameters can be deduced from two
impedance measurements of the loudspeaker itself, one in open air and
one with the driver mounted in a closed box of known volume.
The Thiele-Small parameters are: DC resistance of the driver, resonant
frequency of the driver, the impedance at resonance, the mechanical
q of the driver, the electrical Q of the driver, the total Q of the driver, the
equivalent volume of the compliance of the driver, the acoustic compliance
of the driver suspension, the acoustic mass of the driver diaphragm assem-
bly, the acoustic resistance of the driver suspension, the product of the mag-
netic ux density and coil length, and the acoustic efciency.
Third Octave Filter Commonly heard corruption of one-third octave
filter.
Three-Step Processing See processing.
3:2 Pulldown A clever method of matching the 24 per second frame rate
of motion pictures to the 30 per second frame rate of ntsc television. Sim-
ply put, 24 frames of lm must be stretched out to ll 30 frames of video.
This is equivalent to stretching 4 frames of lm onto 5 frames of video. This

391
Threshold of Hearing

3:2 Pulldown

is possible because 30 frames of video contain 60 interlaced elds. Here is

the procedure:
The rst and second frames of lm are used for both elds of the rst
two video frames. So far, so good. Video frames 3 and 4 are treated differ-
ently, as follows: Field 1 of video frame 3 is taken from the lm frame 2, BUT
eld 2 of video frame 3 is taken from lm frame 3. Then, eld 1 of video
from 4 is taken from lm frame 3 and eld 2 of video frame 4 is taken from
lm frame 4. Then, both elds of video frame 5 are from lm frame 4.
So, lm frames 1 and 4 are translated directly to video frames 1 and 5,
and lm frames 2 and 3 are combined to make frames 2, 3, and 4 of video,
or 3:2 pulldown.
This is cheating, of course, because lm frames 2 and 3 are different, so
the video frames derived from both of them is garbled a little, depending
on how fast the motion is moving. The actual result is that some artifacts
are present in the video presentation, such as some uttering of motion
and a visible horizontal line once in a while.
Threshold of Hearing The softest sound that a normal human ear can detect
is called the threshold of hearing, and it varies in absolute strength at dif-
ferent frequencies. The threshold for 1,000 hertz is taken as a sound pres-
sure of 20 micropascals, and this is the zero decibel point of the sound
pressure level scale. The threshold for 3,000 Hz is several dB lower than
this, our ears being more sensitive in this frequency range. As the fre-
quency goes down, the threshold gradually rises, until at 30 Hz it is some
40 dB higher than at 1,000 Hz. See also fletcher-munson effect; min-
imum audible field.
THX or THX Ltd. A company originally formed in 1983 as a division of
Lucaslm Ltd. Tomlinson Holman was hired by Lucaslm to design a higher
quality motion picture theater sound system for their dubbing stages. The
THX division was formed to license the design to movie theaters and ver-
ify that the licensed theaters hold to the THX specications. The theory was

392
 Timbre

that if studios used THX dubbing stages to mix their movies, and consumers
heard them in THX-approved theaters, everyone would hear the same high
quality sound. The letters ostensibly stand for Tom Holmans eXperiment,
but note also that one of George Lucass early lms was titled THX-1138.
THX Ltd. was spun off into its own company in 2002, although reportedly
a majority of it is still owned by Lucaslm.
THX has expanded into activities such as certifying home audiovisual
equipment and car audio systems, quality control of commercial laserdiscs,
videotapes, and DVDs, cinema and studio consulting, and studio mixing
equipment. They market no hardware of their own except for the loud-
speaker crossover network in the commercial movie theater system. THX
is not a form of encoding or a delivery system. A THX theater can play ana-
log, digital, mono, stereo, Dolby, dts, etc., to their quality specications.
In order to be licensed as a THX theater, special care must be taken in all
aspects of the sound system design and installation, including such things
as the acoustics of the room and even the building materials used in the
space behind the screen.
Tight The low-frequency performance of a loudspeaker is said to be
tight if it is relatively free of hangover, or in other words, is well
damped. A recording using close miking is described as tight. Con-
versely, sometimes poorly damped systems are said to sound loose.
TIM See transient intermodulation distortion, also called
Dynamic Intermodulation (dim).
Timbre Timbre, pronounced tamber, refers to the subjective quality or tone
color of a sound. It is not related to pitch or loudness. It is timbre that
allows us to tell the difference between musical instruments. The timbre of
a sound depends on many factors, including the strength and number of
partials present.
The character of the beginning and ending transient sounds of a tone
has more to do with the timbre of the tone than its harmonic content. For
instance, if a tape recording of a piano is played backward, it sounds very
much like a harmonium, although the overtone structure is unchanged.
This illustrates the importance of the shape of beginning transients, or
attacks, of the sound. A harmonium recording played backwards does
not sound like a piano, but still sounds like a harmonium. There have been
experiments performed where the attack of one instrument is spliced onto
the steady-state sound of another instrument, and in almost all cases, the
sound is identied as the instrument that supplied the attack. For instance,
an oboe attack spliced onto a violin steady-state sounds much more like
an oboe than a violin.
It is commonly said that the piano gets its characteristic timbre from the
fact that its partials are all out of tune with the fundamentals, an effect
called stretching. If this were the case, it would still sound like a piano
when played backward, for the tuning of the partials is the same each way.
There are many unanswered questions about timbre. For example, the

393
Time Alignment

ute sounds as if it has a simple tone with few if any harmonics, but a spec-
trum analyzer shows the ute to produce a complete array of harmon-
ics. It is the attack which determines the timbre of the ute, for even if a
ute attack is spliced onto a clarinet steady-state, the resulting timbre is
very much like the ute.
Some instruments produce a pitch one would not expect from the fre-
quencies of the sounds produced. For instance, the bass recorder sounds
an octave lower than it should, and the human whistle sounds one or even
two octaves lower that its frequency would indicate. See also pitch; fun-
damental tracking.
Time Alignment In a multiple-driver loudspeaker system, it is impor-
tant that the time delay of each driver and its associated crossover net-
work be the same to preserve accurate transient response. In other
words, the high frequencies and the low frequencies must reach the listeners
ear at the same time. A system that meets this criterion is said to be time
aligned.
Unfortunately, drivers differ in their time delays, woofers generally hav-
ing more than tweeters. Crossover networks also have different delays
depending on the frequency range they cover. One way to correct this is
to place the tweeter farther from the listener than the woofer, and this is
done in some systems. Another way is to specially design the crossover net-
work to add suitable electrical delay to the high-frequency signal.
The term time-align in this regard is copyrighted as a trademark by
Edward Long.
Time Code A special nonaudio signal used to record elapsed time on a
tape track. A time code is useful in allowing later synchronization of the
audio tracks with another tape recording or with a motion picture that was
shot at the same time as the recording was made. See also smpte time code.
Time Constant In an equalizer circuit consisting of a resistor and a
capacitor, the time constant is the product of the two values, or RC. The
time constant determines the frequency response of the equalizer, even
though the values of the R and C may be different in different circuits. It is
only the product of the two that determines the cutoff frequency.
Many pre- and de-emphasis circuits are specied in standards by their
time constants because this is simpler than specifying the frequency
response curve. Thus, a 70-microsecond curve in a tape recorder species
a certain de-emphasis curve, and the manufacturer can choose his Rs and
Cs at will in order to attain it.
Time Delay Spectrometry, or TDS A method for measuring the fre-
quency response of a device. In essence, TDS is the measurement of the
response of a device over a certain time interval after the excitation of the
device occurs. This allows measurement of such transducers as loud-
speakers in real rooms, excluding from the measurement the sound reec-
tions from the rooms surfaces. The basic technique has been around for
many years, but the late Richard Heyser resurrected it and greatly rened
it for use in audio testing.

394
 TOC

TDS of a loudspeaker is accomplished by using a linear sine-wave

sweep as an excitation signal. The response of the device is picked up by
a precision microphone at some distance from the speaker, and this
response signal is passed through a sweeping bandpass filter whose
sweep rate matches that of the excitation sine wave source. The sweep of
the lter is delayed in time, or offset, in order to compensate for the prop-
agation time of the sound from the test speaker to the microphone. The sig-
nal from the speaker will arrive at the microphone just in time to get through
the lter, but other signals, such as background noise and reections from
the room walls, will arrive at the microphone too late to get through because
the lter will be tuned to a new frequency by that time. Thus the signal-
to-noise ratio of a TDS measurement can be very large, and measure-
ments can be obtained in quite noisy environments.
It might seem that TDS is a magical technique, allowing measurements
to be made that could never be done by any other technique. To some extent
this is true, but a careful analysis shows that the increase in signal-to-noise
ratio is accompanied by a decrease in frequency resolution, and that the
size of the room (and the speed of sound) determines the ultimate frequency
resolution attainable.
An important benet of the TDS technique is the ability to generate tef,
or Time-Energy-Frequency, curves.
Tin Ear A Tin Ear is someone who is unable to appreciate the ne points
of quality sound reproduction, sometimes by his or her own admission,
but usually not. The typical Tin Ear probably has no physiological prob-
lems with the hearing mechanism, but simply lacks interest or experience.
The poor fellow, however, is much disdained by the golden ear.
Tip, Ring, and Sleeve, TRS A type of phone plug found on patch cords
used for balanced audio signals. It makes three separate contacts. On the
balanced line, the tip is the hot,or signal +; the ring is the low, or
signal side. The sleeve is connected to the shield of the cord; the tip
and ring are connected to the twisted pair of wires enclosed within the
shield. The plug is designed so it does not cause a momentary short
between the high and low sides of the signal. The most common TRS design
is 14-inch in diameter, but there are other sizes.
The tip, ring, and sleeve plug dates from the early days of telephone
exchanges and is still in wide use today. The telephone or MIL spec design
of the real TRS plug is not quite the same as that of the commonly seen
commercial stereo phone plug, for the tip and ring are smaller in diame-
ter. Some stereo phone jacks will make connections with a standard TRS
plug. All similar versions may be called TRS plugs, and sometimes stereo
plugs since commercial TRS plugs are used on stereo headphones. The sim-
ilar commercial TS plug looks nearly the same, but has only two contacts,
the tip and the sleeve.
TOC Table of contents (TOC) is the inner track of a compact disc, and
contains information about the disc such as number of tracks, their posi-
tion on the disc, the timing, and disc identication number.

395
TOGAD

TOGAD Tone-operated gain-adjusting device. The TOGAD was the con-

trol circuitry in the fantasound system that responded to the tones
recorded on the control track for the purpose of varying the apparent loca-
tion and volume of the stereophonic sound image. It was used for the road-
show performances of the movie Fantasia in 1940 and 1941. A similar
technique was pioneered by Harvey Fletcher and his co-workers several
years earlier in their Stereophonic Sound Film System (ssfs).
Tone Tone refers to a signal that has a particular and usually steady pitch.
The simplest of tones are sinusoids, or sine waves, and they are used a
great deal as test signals for audio equipment. The signals recorded on
tape recorder alignment tapes are called tones and are sine waves of
several frequencies.
Tone also refers to timbre, or quality of a musical sound. Tone can be
described by many different analogies, such as bright, mellow, harsh,
tubby, muddy, etc. See also tone control.
Tonearm The mechanical assembly in a phonograph that contains the car-
tridge and allows it to move in such a manner that the stylus makes con-
tact with the record groove at the proper angle and with the proper forces.
The tonearm is much more important than it would seem at rst glance,
and its design has a great deal to do with the way the record will sound
and the wear it will experience.
The tonearm must be pivoted in both horizontal and vertical directions
in order to allow proper freedom of motion for the stylus. In most tonearms,
the entire mass of the arm and cartridge must be dragged across the record
by a force on the stylus and cantilever as the record is played. This results
in a greater force and consequently greater wear on the outside of the groove
wall than on the inside. (But see also anti-skating device.) Therefore,
the friction in the arm pivots must be very low: friction in the vertical pivot
causes undue vertical forces on the stylus, and friction in the horizontal
pivot causes undue horizontal forces. Sometimes these forces can be so great
that the stylus will skip from one groove to the next.
The tonearm must also shield the signal from the cartridge to the wiring
leading to the preamplifier input, and this must not add signicant fric-
tion or noise. For this reason, the tonearm is almost always made of con-
ductive material, and is grounded to the amplier chassis.
The geometry of the tonearm can be optimized by such design features
as making the cartridge form an angle with the arm itself. This is called
offset, and minimizes the tracking angle error. The vertical pivot should
be placed in the plane of the record so that a warped record will not result
in any back-and-forth motion of the stylus in the groove. This scrubbing
action of the stylus causes audible warp wow.
Although most tonearms are pivoted, i.e., they cause the stylus to swing
in an arc as it plays the record, this does not provide the ideal geometry
as far as the stylus and groove are concerned. The record is cut on a record-
ing lathe where the cutting stylus is carried across the record in a straight

396
 Tone Control

line, so that it makes a constant angle with the surface of the record and
with the groove itself. The pivoted arm causes the playback stylus angle
to vary constantly as the record is played, and this causes a certain amount
of record wear and a small amount of distortion as well. Some tonearms
are designed to move the cartridge in a straight line to match that of the
cutter in order to eliminate these effects. Some such designs work very well,
while others introduce other problems such as mechanical resonance
and/or more friction. See also tonearm resonance.
Tonearm Resonance The stylus of a phono cartridge is fastened to a
springy cantilever, and the tonearm-cartridge assembly has mass. This forms
a spring-mass resonance system similar to a weight hanging on a spring.
The stiffness of the cantilever is directly proportional to the resonant fre-
quency: the greater the stiffness, the higher the resonant frequency. The
mass of the system is inversely proportional to the frequency: the greater
the mass, the lower the resonant frequency. (The stiffness of the cantilever
is seldom specied, but, instead, the compliance is usually specied by the
cartridge manufacturer. The compliance is the reciprocal of stiffness, or 1
divided by the stiffness.)
If the frequency of the resonance in question is in the audible range, it
will be excited by the motion of the stylus as it traces the groove, and an
increase in output at that frequency will occur, resulting in a peak in the
response curve. If the resonance is very strongly excited, it can result in the
stylus leaving the groove entirely, causing skipping. For this reason, the res-
onance is kept below the lowest frequency recorded on the record. But it
should not be so low that it would be excited by other disturbances, such as
eccentricity of the record or warps in the record. The best frequency for the
resonance to have is about 10 hertz. To attain this with high-compliance
cartridges requires arms with very low mass. Unfortunately, the needed
information to determine the resonance frequency is usually not available
to the consumer, complicating the selection of tonearms and cartridges.
In some tonearm systems, attention is paid to the damping of this res-
onance, and this is a good idea. Tonearm damping acts like the shock
absorbers on a car, reducing the amount of motion at the resonance and
making the entire system more stable, especially from oor vibrations, etc.
Tone Burst A test signal composed of several cycles of a sine wave. The
tone burst is a transient signal, with frequency content above and
below the frequency of the sine wave itself.
Tone bursts have been used as test signals for loudspeakers for many
years, where they show the effects of hangover. The frequency of the burst
can be chosen to match that of a mechanical resonance, and the amount of
hangover is a measure of the damping of the resonance.
Tone Control One or more knobs on an audio amplifier or preampli-
fier that modify the relative balance between the treble (high-frequency)
tones and the bass (low-frequency) tones. Tone controls are actually
equalizers, and they change the frequency response curve of the

397
Tonic

Tone Burst

device containing them. Mixing consoles will refer to the tone controls col-
lectively as eq.
Tone controls vary a great deal in their design and in the effects they
have on the signal. The earliest tone controls were simply variable low-
pass filters, designed to reduce the effect of noisy 78-rpm records and
noisy am radio reception.
By far the most common conguration today is the familiar bass and treble
controls. They are designed so that in midrotation they have no effect, yielding
flat frequency response. Leftward rotation reduces the bass or treble,
and rightward (clockwise) rotation increases the level of the applicable fre-
quency range. The ostensible purpose of tone controls is to compensate for
minor irregularities in individual loudspeaker systems, or to compensate for
less than ideal placement of the loudspeakers. Some designs work very well
for this, causing only subtle changes in the sound of the system, while other
designs allow gross distortion of the frequency response to be made. Some
home-type equipment even contains graphic equalizers for tone con-
trols, and they must be used with discretion to avoid such distortion.
Because of the prevalence of the exaggerated effects mentioned above,
some audio acionados have come to believe that no tone controls are any
good, and they refuse to use them at all, thinking any modication of the
tonal balance must be a distortion. This, however, is wishful thinking, for
there is no such thing as a perfectly at recording/reproducing chain, espe-
cially when microphones, records, ampliers, cartridges, compact
disc players, and loudspeakers, etc., are available from such a great num-
ber of different manufacturers, each deciding what the optimum tonal bal-
ance is.
We have rarely heard a sound system that did not benet from at least
the subtle application of tone controls.
Tonic The reference pitch on which a musical scale is built; the do of the
familiar do-re-mi-. . . . It is dened in terms of the musical note rather
than in terms of absolute frequency. If one changes the key of a musical

398
 Total Difference Frequency Distortion

selection from C to A at, this is a change of the tonic from the note C to
the note A at.
Top A term for the uppermost octave or so of the response of a loud-
speaker or other audio device. The term is used more in Britain than in
the United States.
Top Hat A plastic disc about 1012 inches in diameter with a 516-inch hole
and a handle in the center, designed for placing over a pancake of audio
tape on a tape recorder. The top hat takes the place of the upper reel flange
and allows the smooth winding of tape. Sometimes called simply a hat.
Topology Topology is a branch of geometry that deals with the way points
in a plane or in space are connected to each other, without regard to the
actual distances or angles between them. It is sometimes called rubber sheet
geometry. In topology, a circle and an ellipse are equivalent, a sphere and
an ellipsoid are equivalent; but a plane, a Mbius strip, and a torus (donut)
are very different from one another.
In an electronic circuit containing resistors, capacitors, and
transistors, etc., the actual geometry of the circuit is less important than
the topology; the order in which the components are connected is much
more important than their physical layout. (The physical layout can also
be very important in some circuits because of capacitive and magnetic cou-
pling, and this must be taken into account when designing the circuit.)
Circuit designs are thus sometimes called topologies.
Toroidal Sometimes transformers are wound on cores shaped more or
less like donuts (or toroids, as the mathematicians would have us call them).
Toroidal transformers are very efcient because they retain almost all the
magnetic eld inside the toroidal shape. There is little leakage of the eld
away from the assembly, and therefore toroidal power transformers cause
less induced 60-hertz hum in nearby circuits. They are difcult to wind,
however, and are more expensive than conventional E-core transformers.
TOSLINK A consumer ber-optic connector used to convey an optical
S/PDIF digital audio data stream. It is short for Toshiba Link, for the com-
pany that invented it.
Total Difference Frequency Distortion, TDFD A method of measuring
nonlinear distortion proposed by the iec in 1982 and based on a sug-
gestion rst published in JAES by A. N. Thiele in 1975.
The test is a type of intermodulation measurement scheme where
two low-distortion sine tones of frequencies F1 and F2 are passed through
the device under test (dut), and the amount of energy at the intermodu-
lation frequencies is measured. The tones are almost but not exactly a musi-
cal fth apart (a frequency ratio of 2:3). The IEC recommendation is 8
kilohertz for F1 and 11.95 kHz for F2. This gives rise to two intermodu-
lation products, F2 F1 and 2F1 F2, which lie 100 hertz apart in frequency
about an octave below F1 on a piano. These tones are quite easy to iso-
late from the test tones with a filter. The total energy in these tones is
expressed as a percentage of the energy in the parent tones, something like
the way total harmonic distortion (THD) is expressed. One advan-

399
Total Harmonic Distortion

tage of TDFD is that it is much more sensitive than THD and much lower
quantities of distortion can be measured. This is because THD measure-
ments are contaminated by wideband noise in the DUT output; whereas
with TDFD, all the distortion is in the two closely spaced components, and
it is measured through a narrowband lter that discriminates against the
noise. It is said that distortion of less than 0.0001 percent (one part per mil-
lion) can be measured accurately with this technique.
The method is applicable to measurements on preamps, amplifiers,
loudspeakers, and tape recorders.

TDFD

Total Harmonic Distortion, or THD See harmonic distortion.

Totem Pole Gyrator A type of gyrator circuit that uses four resistors,
a capacitor, and two opamps in a conguration that looks on paper some-
thing like a totem pole.
T-Pad A type of attenuator typically consisting of three resistors
arranged schematically in the form of the letter T. It can have a xed amount
of attenuation or can be variable, in which case, the three resistors are
replaced with potentiometers that are ganged together. The variable
T-pad presents constant input and output impedances to its associated
circuitry regardless of the attenuation and was often used as a master gain
control in passive audio consoles. See gure at attenuator for schematic.
Trace A term sometimes used for the copper conductors on etched circuit
boards. See also foil.
Tracing Distortion Records are cut with a chisel-shaped cutting stylus
with a very sharp point, and this point accurately follows the audio wave-
form being recorded. The record is reproduced, however, with a stylus hav-
ing a round cross section with a nite radius. This means that the reproduced
waveform will be distorted because the playback stylus cannot follow pre-
cisely the curve of the groove bottom.
This can be visualized by considering the chisel-shaped cutter moving
side to side as it cuts the groove. The groove will be narrower when the

400
 Tracking Error

cutter is moving sideways the fastest, and will be wider when the cutter is
at the extremes of its motion. The groove thus will become narrower in its
midpoints and wider at its extremes. The playback stylus will be forced
upwards in the narrow parts and will fall downwards at the wider parts.
In other words, it is pinched out of the groove periodically. This is some-
times called the pinch effect.
Because the stereo cartridge is sensitive to vertical as well as horizontal
motion, this pinched motion adds to the signal, and is manifested as har-
monic distortion. The smaller the tip radius, the smaller the effect; and
an elliptical stylus, with smaller side radii, minimizes it. There are other
special stylus shapes designed to have even smaller side radii than an ellip-
tical, and they do reduce tracing distortion.
It is possible to pre-distort the signal fed to the cutter in such a way
as to cancel out tracing distortion, but it can only work for one playback
tip radius. Because these tip radii exist in many different sizes and shapes,
this is only a compromise at best.
Track and Hold See sample and hold.
Track-at-Once, TAO A CD-R production method in which one or more
tracks are recorded at a time and a link is written between the tracks. This
method is often used to create multisession cd-rs. A disadvantage of track-
at-once recording is that gaps between tracks must be at least two seconds
in length. See also disc-at-once.
Tracking When producing a modern multitrack recording, the rst step is
usually to record the separate instruments and vocals on their own audio
tracks, but with less regard to the later mixdown. This is called tracking.
Many types of tape transports using rotary head technology, such as dat,
videotape, and adat, must control the tape motion and synchronization
of the rotating head with the track location in the tape. This is also called
tracking. In most such systems a servo system detects tracking pulses re-
corded on the edge of the tape and automatically makes required small
adjustments in the tape speed. In most video recorders, there is also a knob
labeled tracking that adjusts for differences in the machine geometry of dif-
ferent machines when playing a tape made on another machine. Once
adjusted manually the servo will take over and maintain alignment.
Tracking Error Because the tonearm of a record player is pivoted at the
stationary end, the stylus moves across the record in an arc and meets
the groove at an ever-changing angle. The cutting stylus, however, moves
in a straight line, always 90 degrees to the groove direction. This condition
is called tracking error. If the cartridge is mounted to the end of the arm
at a suitable angle (called the offset angle), the error can be made to be zero
at two places on the record and only a few degrees in the worst case. Sev-
eral schemes for straight-line tracking have been developed over the years,
some with greater success than others. Some of these tangential track-
ing arms are quite complex, with motor-driven servo systems moving the
cartridge across the record. Most people consider the slight improvement
in tracking to be not worth the effort and expense involved.

401
Trank

Trank See transcription.

Transaural Stereo A technique of recording and playback where a bin-
aural head is used to make the recording, but instead of listening to it via
headphones, two loudspeakers are used. In order to simulate binaural
hearing conditions using loudspeakers, the two signals must be specially
processed to cancel the crosstalk between the channels caused by the
fact that each ear hears both loudspeakers. In a true binaural listening sit-
uation, each ear hears only the sound from its corresponding microphone
in the dummy head. In a transaural system, the crosstalk signal heard in
each ear is canceled by inserting an inverted replica of that signal in the
opposite channel. If the inserted signal is an exact replica of the crosstalk
signal, it will completely cancel at the ear of the listener, giving the effect
of true binaural listening. The problem is to determine what the crosstalk
signal actually is and then to generate it. There are several schemes avail-
able for doing this, one of which was patented by Atal and Schroeder in
1966. The schemes enjoy varying degrees of success. It should be noted that
every listener has a unique physiology, and the correct cancellation signal
for one person will not be correct for another. However, exact cancellation
is not required for a convincing effect, and the listener is relieved of the
burden of wearing a headset, contributing to the sense of realism. See also
binaural recording.
Transconductance In vacuum tubes, transconductance is the ratio of a
small change in plate current to a small change in grid voltage. It is
expressed in units of conductance, originally micromhos but now
microsiemens. The transconductance of a tube is closely related to its volt-
age gain.
Transcription Before the advent of the tape recorder, 16-inch acetate
discs were used to record radio broadcasts for later playback at a more suit-
able time, usually in another time zone. Some transcriptions were cut from
the inside out, and some used vertical groove modulation rather than lat-
eral. A typical standard transcription would play for a half-hour per side
in fact, many were only recorded on one side. A smaller version also existed
that was 12 inches in diameter and played for 15 minutes. These transcrip-
tions, originally thought to be needed for one playing only, are the only
records of old radio programs in existence and are quite valuable now as
collectors items.
In the vernacular of radio station personnel, transcriptions were called
tranks or ETs (ET standing for electrical transcription).
Transducer A transducer is a device that converts mechanical, magnetic,
or acoustic energy into electrical energy, or vice versa. Examples of trans-
ducers are microphones, phono cartridges, loudspeakers, tape
heads, and disc cutting heads. In general, the transducers are the weakest
links in an audio chain, and cause most of the distortion. The primary
problem with electromechanical transducers is that they are nonlinear
except for relatively small ranges of signal frequency and amplitude.
In other words, they have a limited dynamic range compared to music,

402
 Transformer

which has energy spread over at least a 90-decibel range. This is a factor
of 1 billion in power!
Transformer A transformer is a device consisting of two or more coils of
wire wound on a common core of soft iron or other magnetically perme-
able material. The number of turns in one coil divided by the number of
turns in the other one is called the turns ratio. An alternating voltage
across one coil will appear across the other coil multiplied by the turns
ratio.
Some transformers are designed to operate at 60 hertz and to handle
a large amount of current. They are called power transformers, and are
found in almost all electronic equipment to change the 110-volt line volt-
age to suitable values. A large version of this type of transformer is found
on utility poles by your house.
Audio transformers are designed to operate at audible frequencies, and
are used to step audio voltages up and down and to send signals between
devices such as microphones and tape recorders, etc., while maintaining
electrical isolation. See also balanced line.
The turns ratio of a transformer always determines the relative voltages
at the input (primary) and output (secondary) terminals, but the imped-
ance ratio of a transformer is equal to the square of the turns ratio because
the inductance of a coil is proportional to the square of the number of
turns. Because the power in the signal is conserved between the input and
output of a transformer, it can be shown that a 2:1 step-up transformer will
have a 4:1 impedance ratio. A 100-ohm resistor across the secondary of a
2:1 transformer will be reected at the primary as 25 ohms. Usually, audio
transformers are rated by their impedance ratios rather than their turns
ratios. A microphone input transformer might be rated as 150 ohms to 15,000
ohms. Such a transformer will have a turns ratio of the square root of 100:1,
or 10:1. One millivolt generated by the microphone will be transformed into
10 millivolts at the secondary, and a 15,000-ohm impedance connected to
the secondary will look like 150 ohms to the microphone.
Transformers are nice because they can provide voltage gain without

Transformer

403
Transient

signicant noise increases, but they do add a certain, although small,

amount of distortion. In recent years, much effort has been expended in
rening transformer design, and much progress has been made. Today, one
can nd audio transformers with very wide frequency response and
uniform phase response, along with very low distortion.
Transient A transient is a nonperiodic, or nonrepeating, sound waveform
or electrical signal. Music contains many transients, such as the attacks
of musical instruments, especially percussion instruments. The consonants
of speech are transients.
Transients are more difcult to record and reproduce without distor-
tion than are steady-state sounds. A transducer or amplifier with per-
fect transient response must have uniform frequency response, but also
must have uniform phase response. Transients in music are often of much
higher amplitude than the average amplitude level, and they can cause
overload of devices with too little headroom. How an amplier or other
device recovers from a momentary overload has a great deal to do with its
sound quality.
The achievement of adequate transient response in all parts of an audio
recording and reproducing system is an ongoing quest.
Transient Intermodulation Distortion, TIM A type of dynamic inter-
modulation distortion that occurs in feedback ampliers and is
caused by nonlinear operation of the gain stages within the feedback
loop. It occurs when the input signal voltage contains a very rapid
change. In essence, the cause of TIM is that an amplifier has a nite tran-
sit time for a signal going through it, and this means the feedback is a lit-
tle late in its ability to reduce distortion. A rapidly changing signal gets
through the amplier before the feedback can control it. TIM is also called
slew-induced distortion. See slew rate. There is no standard for mea-
suring TIM, and several methods are used by different investigators. TIM
affects the sound of very high frequency, high-level sounds, such as that
of a cymbal crash at short range, and the result is the addition of low fre-
quencies, causing a muddying or thickening of the sound. To reduce TIM,
ampliers are used that have a very wide open loop bandwidth,
which is another way of saying that their transit time is very short. Early
forms of integrated circuit opamps are notorious for large amounts
of TIM.
digital audio systems are also capable of a type of slew-induced dis-
tortion. In dac theory, it is assumed that the DAC output moves instanta-
neously from one level to another, creating a staircase waveform with
vertical transitions. The actual steepness at the output of a practical DAC-
deglitcher combination will be determined by the slew rate of the out-
put amplier. If the slew rate is too low, small triangular pieces will be cut
out of the staircase, causing an error or distortion that is a nonlinear func-
tion of the input signal. This is slew-induced distortion, and the problem
does not stop at the output of the deglitcher. The signal at this point still
contains voltage transitions that are too fast for many audio processing

404
 Trigger

devices to handle. The anti-imaging filter must be carefully designed

to see that it is not driven into slew limiting by these transitions.
Transient Response See transient.
Transistor The transistor is the basic solid-state amplifying element used
in almost all audio equipment today, having almost completely replaced
the vacuum tube.
The transistor is nonlinear in its amplication of electric currents,
and to make it usable in ampliers, great pains must be taken to make it
behave in a linear fashion and so to reduce its distortion.
The transistor is a three-element device consisting of two terminals to
pass the signal current and a third terminal for controlling the signal cur-
rent. In this it resembles the triode tube, although the triode is much more
linear and has other desirable characteristics such as a very high input
impedance.
Conventional transistors pass current in one direction only and are
called bipolar. (Logic would dictate that they should be called unipolar.)
There are two types, npn and pnp, and each carries current in the oppo-
site direction.
In addition to bipolar transistors there exist field effect transistors
(FETs), which behave much more like triode tubes. They also are of two
types, called P-channel and N-channel. FETs are suitable for high-quality
audio devices, but they are only recently available with sufciently high
current-carrying capacity to be used in high-power amplier outputs. See
also semiconductor.
Transondent Possessing the ability to freely transmit sound, analogous to
transparent in reference to light. The grill cloth covering a loudspeaker
must be transondent for obvious reasons. The term is believed to have been
coined by the late Paul Veneklasen, a respected acoustician in the Los Ange-
les area.
Trap A band reject filter designed to eliminate a particular frequency from
a desired signal. See also bias trap; bass trap.
Treble That portion of the audio spectrum that encompasses the frequen-
cies from about 2,000 hertz to 20 kilohertz, as opposed to bass and
mid-range.
TRF Acronym for Tuned Radio Frequency. A type of radio receiver invented
and patented by the U.S. Navy in the mid-1920s that used resonant circuits
in the grid or plate circuits of one or more RF amplifier tubes. The res-
onant circuits were tuned by variable capacitors or inductors, that were con-
nected to the tuning dials, one for each stage. Thus several dials had to be
tuned to the station frequency for reception. TRF receivers were superseded
by the superheterodyne sets.
Tri-amplication A system used in some loudspeaker systems where a
three-way crossover network is used before three power amplifiers,
each amplier handling only its limited part of the total frequency range.
It is an extension of biamplification.
Trigger An instantaneous voltage pulse generated from a synthesizer key-

405
Trim

board when any key is depressed. The trigger is used to initiate the action
of some other device, such as an adsr.
In analog CRT oscilloscopes, the horizontal sweep is initiated at a
particular instantaneous signal level, and this is called a triggered sweep.
Trim Generally refers to a small adjustment, analogous to tweak. Small con-
trols such as gain controls operated by screwdrivers are called trimpots
or trimmers, and are used for such things as adjusting pre-emphasis
and de-emphasis in tape recorders, etc. Some recording consoles label the
ne-adjust microphone preamplier gain control a mic trim.
Trimmer A variable capacitor whose value can be adjusted by means
of a small screwdriver. Trimmers are meant to be adjusted only rarely and
are commonly found in equalization circuits, where they are nely
tuned to achieve a particular frequency response characteristic. See also
trim.
Also, a control knob that has a relatively subtle effect, such as the ne
volume control on a recording console input module, is called a trimmer
or a gain trimmer.
Trimpot A small potentiometer whose setting is usually adjusted by a
small screwdriver. Trimpots are designed to be adjusted only rarely and
are used in sensitive circuits such as equalizers, where they can be nely
adjusted and left for long periods. See also trim.
Triode The three-element vacuum tube in which the elements are the cath-
ode, the grid, and the plate. The triode is probably the most common
type of tube and is used primarily as an amplifier stage and as a cath-
ode follower. The triode was invented and patented by Lee DeForest in
1906, and this is one of the fundamental patents in the history of electron-
ics. It is interesting to note that DeForest thought of the triode as a detec-
tor of radio signals, not an amplier. As a detector, it was inferior to other
detectors of the time, including the fleming valve, which was the rst
diode detector. It was the quick-witted engineers at the Bell Telephone Lab-
oratories who rst hit on the idea of using the triode as an amplier, and
they obtained a license to build it from DeForest very early and immedi-
ately started using it as an amplier in long-distance telephone circuits. They
also started manufacturing triodes for commercial sale under the Western
Electric brand.
TRS See tip, ring, and sleeve.
T-Section A lter, or part of a lter, that has two series and one shunt reac-
tive components. Its name derives from its schematic resemblance to the
letter T. The T-section has a at response in its passband and an 18 dB per
octave slope in its stopband.
TTS See temporary threshold shift.
Tube Short for vacuum tube. The tube is the forerunner of the transis-
tor. The rst tube used for amplication of signals was the audion
invented by Lee DeForest. In England, tubes are called valves, a carry-
over from the fleming valve, which was the rst two-element tube, or

406
 Tube

diode. In actuality, Edison invented the diode but did not exploit it for any
practical use. Fleming, an Englishman, used it as a detector of radio sig-
nals, and patented it for this use.
Tubes generally contain the cathode and anode of the diode, with
the addition of one or more grids that control the current between cath-
ode and anode, and are relatively linear in their amplication of signals.
The primary disadvantages of tubes compared to transistors is that they
are much larger and they generate much more heat. They are thus much
less efcient. Some applications require tubes, such as X-ray equipment and
high-power radio and television transmitters.
The natural linearity of tubes, as compared to transistors, is much ad-
mired by some audio people, who insist that tube equipment sounds more
musical than equipment using transistors. Although a great deal of non-
sense has been written about this subject, sometimes by people who ought

T-Section

407
Tuchel Connector

to know better, it is true that there is a difference in the sound produced by

the two systems, especially when signal levels are high and likely to drive
the system into overload. Generally, tube equipment produces more even-
order harmonic distortion, while transistors tend to produce more odd-
order harmonics. Even-order harmonics are associated with loud,
brassy, etc., timbres, while odd-order harmonics produce a more cov-
ered or closed sound. Tube equipment usually overloads more gradu-
ally, causing the signal to begin to sound loud and bright before clipping
occurs. With transistor equipment, hard clipping starts at the instant of
overload, and it generates large quantities of high-order distortion.
There is a viable market for very expensive, hand-built tube-type audio
equipment, and most of this market is, maybe surprisingly, in Japan. The
Japanese are masters at large-volume production of high-quality electron-
ics, and they lead the world in the manufacturing of audio components. But
they are not so much interested in the limited production of high-quality
handmade components, and most of the worlds production of this eso-
teric equipment is in the United States. Curiously, tube manufacturing is
now mainly in China and eastern Europe.
Tuchel Connector Some European and, to a lesser extent, American audio
equipment uses cable connectors that conform to the German din stan-
dard for pin locations. These DIN-standard connectors are made in large
quantities by the German manufacturer Tuchel, and this name is commonly
used for all of them, regardless of manufacture. DIN connectors are rela-
tively inexpensive, and the standard dictates not only the locations of the
pins, but also what the pins are connected to. For instance, input and out-
put signals might use the same connector, but would use different pins. Alas,
not all manufacturers conform to the letter of the standard, causing some
confusion.
Tuning Condenser A variable capacitor, usually consisting of two
groups of aluminum plates that gradually intermesh as a shaft is turned.
It is the component used to change the frequency of maximum sensitiv-
ity of conventional radios, allowing different stations to be tuned in. In some
small radios, the tuning knob is connected directly to the shaft of the tun-
ing condenser.
Turnover In the recording of phonograph records, the turnover frequency
is that frequency below which the recording is done at constant dis-
placement and above which it is constant velocity. Before the riaa
standardized the playback equalization curve for stereo records, different
record manufacturers used different turnover frequencies, and preampli-
fiers had adjustable controls to select various curves.
Turntablism A musical form based on the real-time manipulation of musi-
cal sounds recorded on vinyl LP stereo discs. The Berklee Press, a subsidiary
of the Berklee College of Music in Boston, publishes books and LP records
for turntablism.
Turntablist A musical performer who uses two or more phonograph turn-
tables and records (usually stereo LPs) along with a mixing console to con-

408
 UDF

trol the resultant musical effect; a highly specialized Disk Jockey. One tool
used by turntablists is the cross-fader, a type of mixer that fades from one
stereo source to a different one.27
TVI See television interference.
Tweak, Tweak Up Tweak is technicians slang for a delicate adjustment
procedure. Often the performance of audio components can be improved
over the original manufacturers specications by judicious tweaking up
of the adjustments, or by making small changes in certain component values.
Some audio buffs spend a good deal more time tweaking up than lis-
tening to their sound systems. See also alignment.
Tweek A trade name of Sumiko for an electrical contact enhancer uid sold
to the home audio market, much praised by some audiophiles. It is other-
wise known as Stabilant 22 from D. W. Electrochemicals, Ltd., of Canada.
Tweeter In a multiway loudspeaker system, the small speaker that
emits the high frequencies is sometimes called a tweeter.
Twin-Tone Intermodulation Distortion A method of measuring inter-
modulation distortion by passing two sinusoidal tones of equal
amplitude through the dut and examining the output signal to see the
amplitude of the difference frequency. The two tones, which differ in fre-
quency by a xed amount, usually 50 hertz, are swept through the audi-
ble range and often into the ultrasonic range. Even though the ultrasonic
signals themselves are not audible, the difference frequency is in the audi-
ble range.
Twist-Lock A trademark of the Hubbell Corporation for a family of lock-
ing AC power connectors. For years, many sound companies used them
for loudspeaker to amplifier connectors, but since they were also often
used for mains power distribution, the obvious disaster often occurred.
Recently, the Neutrik Speak-On connector has become popular for loud-
speakers.

U
U Abbreviation of unit, as in modular unit or rack unit, which is the stan-
dard unit of height measurement of rack-mounted equipment. One mod-
ular unit is 134 inches, and all rack-mounted components measure an
integral number of modular units high. Thus, a device might be described
as occupying 3U of rack space, or 5 14 inches. The standard width of 19
inches is assumed. The Modular Unit standard was agreed upon by the
eia and ansi.
UDF, Universal Disk Format A trademarked name for an optical disc

27. Thanks to Dennis Bohn, Rane Corp.

409
UHJ

le format dened by OSTA, the Optical Storage Technology Association.

It complies with the ISO 13346 standard (volume and le structure of write-
once and rewritable media using nonsequential recording for information
interchange), and its use facilitates compatibility among computers of dif-
ferent operating systems.
UHJ Abbreviation for Universal H J matrix, a method of encoding three or
four channels of audio onto two channels. H and J are simply code let-
ters adopted by the BBC, and have no other meaning. See also ambison-
ics.
UL Short for Underwriters Laboratories, which is a standards and testing
laboratory nanced by the insurance industry. The purpose of UL testing
is to ensure re safety of electrical equipment. UL approval has nothing to
do with the audio quality of any apparatus.
Ultrasonic Having frequencies above the normal range of human hear-
ing. Not to be confused with supersonic, which means faster than the
speed of sound. Ultrasonic signals can cause distortion in audio compo-
nents that is in the audible range. See also transient intermodulation
distortion.
U-Matic A 34 helical scan video cassette tape recorder format developed
by Sony for professional use. It was the world standard for industrial and
semipro videotape productions for a long time. This format was also used
in conjunction with a Sony PCM 1610 or PCM 1630 encoder to record dig-
ital audio destined for release on CDs.
Una Corda Italian for one string. Notation used in printed piano music
to indicate the use of the soft pedal, which shifts the keyboard so the ham-
mers only hit one or two of the three strings per note.
Unbalanced Line Any transmission line in which the ground conductor
completes the electrical circuit and serves as a shield. See also Appendix 6.
Underhung Coil One way to increase the linearity and reduce the dis-
tortion of a dynamic loudspeaker is to design the voice coil so it is rel-
atively short compared with the length of the magnetic gap in the pole
pieces. This means the coil sees a uniform magnetic eld strength regard-
less of its position in the gap, reducing the nonlinear distortion. This does
not reduce the efficiency of the loudspeaker, but it does increase the cost,
because a larger magnet must be used. See also overhung coil.
Uni-axial Microphone A unidirectional microphone that has max-
imum sensitivity in the axial direction, or along the main axis of the
microphone.
Unidirectional Microphone A microphone that is more sensitive to sounds
coming from one particular direction than those coming from other direc-
tions; sometimes called simply directional microphone. There are many
degrees of directionality, and there is no such thing as a perfect unidirec-
tional microphone. See cardioid; super-cardioid; shotgun; line
microphone; and hyper-cardioid.
Unidyne The rst directional microphone to use a single active trans-
ducing element was the Unidyne, invented by Benjamin Bauer in 1937. The

410
 Variable Area

name comes from unidirectional dynamic, and it was a dynamic microphone

that was modied to allow sound to reach the rear of the diaphragm through
a time-delay network. This caused the phase cancellation of sounds com-
ing from the rear. Today the principle is used in many dynamic microphones
having cardioid or similar polar patterns. Unidyne is a trademark of
Shure.
Prior to the Unidyne, directional microphones were made with two ele-
ments whose outputs were added. Typically, a bidirectional velocity
microphone element and a dynamic pressure microphone were used.
The rear lobe of the ribbon element was canceled by the rear response of
the pressure microphone, as the two were of opposite polarity.
Unity Gain A device that does not attenuate or amplify a signal is said to
be a unity gain device. In other words, the gain is 1. Most signal pro-
cessing devices have unity gain, which means they can be inserted into
an audio system at various places without upsetting the overall gain of the
system.
USB Universal Serial Bus, a medium bandwidth serial digital data inter-
connection standard, commonly used for keyboards, mice, and other periph-
erals, including some digital audio gear. Even though the newer USB 2 may
appear fast enough, firewire/ieee 1394 is usually preferred for digital
audio and video work.
UUT Acronym for Unit Under Test. See dut.

V
VA Short for Volt-Ampere, or the product of the voltage times the current.
In a resistive circuit, the VA will be equal to the wattage, for a watt is 1 volt
x 1 ampere. In an inductive circuit, or device such as an electric motor, how-
ever, the current will not be precisely in phase with the voltage, and the
average current times the average voltage will be greater than the actual
power absorbed by the device. Also short for variable area.
Vacuum Tube See tube.
Valve British terminology for vacuum tube. The rst tube was a diode
introduced by Ambrose Fleming in 1904, and he called it the fleming
valve. It was used as a detector of radio signals. See also edison effect.
Van den Hul A specially shaped stylus tip for reduced distortion and
wear and increased high-frequency performance in playing phonograph
records. See also stylus.
Varactor A special semiconductor diode used as a variable capacitor.
The capacitance varies as a function of an applied DC potential. Varactors
are sometimes used as tuning condensers in certain radio receivers.
They are also sometimes called varicaps.
Variable Area The most common type of analog optical sound track
used in motion pictures. The variable area track is a transparent line in a

411
Variable Area

Variable Area

black background. The relative width of the transparent part (strictly

speaking, not the area) is varied in accordance with the sound waveform.
To produce the sound in the projector, a thin at beam of light is passed
through the lm and onto a photocell, which generates an electric current
in response to the amount of light transmission.
Most analog movie sound tracks today consist of two parallel variable
area tracks, and this makes it possible to record stereophonic sound. See
sva. The track is recorded on the unexposed lm by a special optical device
known as a light valve.
Another type of early variable area lm recorder was a Dutch invention
called Philips-Miller (see illustration below). It had a special mechanical
cutter, and was used in the 1930s for recording radio transcriptions in Europe
and in the 1950s for a few television productions made on 16-mm lm. The
lm had a thin opaque coating over the sound track area, and a small pointed
plow-shaped stylus was used to cut a furrow through the coating into the
lm. As the depth of cut was varied according to the audio waveform being
recorded, the furrow varied in width in like manner, resulting in a sound
track that could be read by conventional optical sound heads.

412
 Variable Pattern Microphone

The Philips-Miller cutter

Variable D A trademark of Electro-Voice for a series of dynamic direc-

tional microphones introduced in 1954. The term stands for variable delay,
and refers to three sound entry ports on the side of the microphone. Each
port incorporates a different time delay for sound entering it and then reach-
ing the back side of the diaphragm. The time delays for rear entering sounds
effectively shift the phase of the sound by 180 degrees so they cancel the
sound entering the front of the microphone. The longest delay extends the
phase cancellation to lower frequencies reducing the proximity effect.
Variable Density A type of optical sound track used for motion pic-
ture sound for many years until the advent of the variable area sound
track. The variable density sound track is a narrow gray strip between the
picture frames and the sprocket holes of the lm. In recording the track,
the audio waveform is used to amplitude modulate a thin, narrow
light beam that exposes the lm. The resulting lm density, or light trans-
missibility, varies in accordance with the variations in the waveform. To
reproduce the sound in the projector, a tiny beam of light is passed through
the lm and onto a photocell, which generates an electric current propor-
tional to the amount of light transmitted, thus reproducing the waveform.
The track resembles a modern bar code that uses shades of gray.
Variable density recording is plagued with problems, one of the worst
of which is a high noise level caused by the inherent graininess in the lm.
The photographic process is nonlinear and this causes a relatively high
distortion. These problems led to a great deal of research in the 1930s,
much of it done by the Bell Telephone Laboratories, and the result was the
introduction of the much-improved variable area system.
Variable Pattern Microphone A microphone that can be adjusted to achieve

413
Variable Pitch

different directional characteristics, or polar patterns. Some of the ear-

liest variable pattern microphones were the 77 series of ribbon micro-
phones introduced in the 1930s by Harry F. Olson of RCA. Most variable
pattern microphones made today consist of two condenser micro-
phone elements placed back to back with their output voltages combined
in different ratios and polarities. This conguration was originally devel-
oped by H. J. Von Braunmuhl and Walter Weber in Germany and has been
gradually rened over the years.
Variable Pitch The pitch of a phonograph record has nothing to do with
musical pitch; it is dened as the groove density in grooves per inch.
If a record is cut with low pitch, i.e., with widely spaced grooves, the
recorded volume can be large with no danger of the grooves inter-cutting
one another. Low pitch, however, means less recorded time on the record,
and we see there is a trade-off between recording time and volume or
amplitude of the signal. Of course, music is not uniformly loud, and
recording time can be increased if the grooves are packed together more
closely during soft passages and spread wider apart on louder passages.
This is variable pitch recording.
In order to record a disc with variable pitch, the recording lathe must
sense what the signal level is in order to adjust the spacing accordingly.
This is done with a preview head, which is another playback head on the
tape machine that plays the master tape into the record cutter. The preview
head is placed so it hears the signal before the main playback head does
so, and this negative time delay allows the pitch control system to start spac-
ing the grooves wider apart before a loud passage comes along. A digital
delay can also be used to delay the signal feeding into the cutter.
Variometer A type of rf transformer where one coil is wound on a spher-
ical form and placed inside the other coil. By rotating the inner coil, the
amount of coupling between the coils is varied. Variometers were used in
the 1900s to add positive feedback to the RF stages of radio receivers. See
also regeneration.
Vari-Speed Vari-speed is a means of changing the speed at which an ana-
log tape recorder runs in order to change the pitch of a recorded tape or
to change the time duration of the recording. In most modern tape machines,
a servo-controlled capstan is used, and frequently the speed may be var-
ied by changing the reference frequency of the servo.
Vari-speed historically consisted of an oscillator coupled to a power
amplifier that had enough power output to drive the synchronous cap-
stan motor of the tape recorder. As the oscillator frequency is varied from
the nominal 60-hertz line frequency, the capstan motor will change speed
accordingly, changing the tape speed. The rst commercial tape recorder
speed control system used a 60-hertz tuning fork as a standard and was
introduced by the Ampex company in the 1950s. It was not meant to con-
trol the speed over a broad range, but only to provide a steady 60 Hz ref-
erence for use when the tape recorder had to be operated from power lines

414
 VF-14

other than the normal 60 Hz. The exact term vari-speed was trademarked
by Superscope (Superscope distributed Sony recorders in the U.S. at the
time) in the 1970s.
Varistor A semiconductor circuit component whose resistance is a function
of the voltage applied across it. Its resistance drops nonlinearly with
increasing voltage, with a steep decrease at its rated voltage. Varistors are
used as surge protectors in equipment sensitive to overvoltage conditions.
VCA See voltage-controlled amplifier.
VCF See voltage-controlled filter.
VCO See voltage-controlled oscillator.
Velocity Microphone A type of microphone that has a polar pattern
shaped like a figure 8. This pattern is also called a cosine pattern
because the cosine function gives this curve when plotted in polar coor-
dinates. The rst velocity microphone was the ribbon microphone, invented
about 1931 by Harry F. Olson of RCA Research Laboratories.
The ribbon microphone uses as an active element a small corrugated strip
of very thin aluminum ribbon hanging loosely in a strong magnetic eld.
The ribbon is moved by the action of air molecules, which are set in motion
by the sound wave. The resonant frequency of the ribbon is very low,
below the audible range, so the motion of the ribbon is mass controlled,
or is proportional to the velocity of the air particles. For this reason, it is
called a velocity microphone.
The ribbon is most sensitive to sound approaching it either at normal inci-
dence or perpendicular to the plane of the ribbon because this causes it to
execute maximum motion. As the angle of incidence of the sound changes,
the motion of the ribbon is reduced and the output voltage from the micro-
phone is reduced, until it is zero when the angle of incidence is 90 degrees.
This is like a window shade, which is easily moved by air currents perpen-
dicular to it, but is not moved by wind blowing across its surface.
The ribbon was the rst commercially successful directional microphone,
and found instant acceptance in the motion picture and radio broadcast-
ing industries. It is still in use for many purposes, although it has numer-
ous disadvantages compared to modern microphone types, especially
condenser microphones. It is characterized by very smooth sound,
much appreciated by brass players for recording. It suffers from proxim-
ity effect, however, making it less than ideal as a vocal microphone. It is
also quite sensitive to air currents and wind, making it useless outdoors,
unless a good windscreen is used.
In recent years, special condenser microphones have been designed with
the cosine pattern. They are compound microphones, consisting of two
cardioid microphones placed back-to-back and connected in phase
opposition.
VF-14 A vacuum tube made in the 1950s by Telefunken in Germany. The
VF-14 was used as a preamplifier in the famous U-47 condenser
microphone. The U-47 has a pronounced high-frequency peak, and

415
VFO

Velocity (Ribbon) Microphone

recordings made with it have a characteristic sound much praised by many

recording engineers.
The VF-14 has been obsolete and out of production for many years, and
most U-47s have been converted to fet preampliers. However, many
people believe the tube sounded better, and the price of used VF-14s has
reached astronomical values. There is now a company making replica Tele-
funken microphones that has the rights to the Telefunken name in the U.S.
They will supply various tubes including nos VF-14s.
VFO Acronym for Variable Frequency Oscillator. Audio-frequency VFOs
are used in the generation of electronic music and for generating test sig-
nals for audio equipment. Radio-frequency VFOs, also called local oscilla-
tors, are used in superheterodyne AM and FM tuners and television sets
to generate signals for mixing with the received signal to produce the
intermediate frequency.
VHS Hi-Fi A method of very high-quality audio recording using a VHS
video cassette as the recording medium. The audio recording is a com-
panded frequency modulation process, and the carrier is recorded onto
the tape at a very high level so the recording penetrates into the magnetic
coating. The recording still exists after the video track is recorded over it,
a trick called depth multiplexing. VHS Hi-Fi was introduced after
Sonys beta hi-fi, and has similar performance specications, but is not
compatible with it.
Vibration Switch See shock switch.

416
 Vitasound

Vibrato The frequency modulation of the sound of a musical instru-

ment, especially the singing voice. The vibrato also includes amplitude
modulation, but the frequency modulation is more apparent to the
listener.
Most solo instruments are able to produce the vibrato, although some,
most notably the piano, cannot. The combined vibratos in a string section
help give rise to the chorus effect.
Victrola Victrola was the trade name from 1906 for phonographs made by
the Victor Talking Machine company, of Camden, New Jersey. It soon
became a generic name for the record player. Victor also introduced the
famous little fox terrier named Nipper sitting in front of the Victrola lis-
tening to His Masters Voice as a trademark in the United States.
Video Literally I see in Latin. Video is to television what audio is to radio.
Virtual Pitch See residue pitch.
Viscous Damping The damping of a mechanical resonance by the viscosity
of a thick, oily uid. The uid used today is usually a type of silicone oil.
Some early cutterheads were viscous damped, and many record-playing
tonearms have been viscous damped. One advantage of viscous damping
is that the amount of damping can be adjusted by the amount of the damp-
ing uid used or by the viscosity of the uid. Disadvantages of viscous
damping are the fact that the amount of damping is temperature depen-
dent, and it can be messy to deal with the uid.
Vitaphone A corporation formed in 1926 by Western Electric and Warner
Bros. for the purpose of manufacturing and installing sound motion pic-
ture equipment. The Vitaphone system was designed by Maxeld and Har-
rison of the Bell Telephone Laboratories and used a 3313-rpm, 16-inch
diameter disc record coupled to the movie projector mechanism. This was
the rst recording medium to use a speed of 3313 rpm and was the rst
successful commercial sound system for movies in the United States. The
Jazz Singer, starring Al Jolson and released in 1927, was the rst to movie
use it.
One problem with the Vitaphone was the inconvenience of handling the
large records, one for each reel of lm. There were also synchronization prob-
lems, especially if the lm had broken and had been spliced. The Vitaphone
system was supplanted in the 1930s by optical sound tracks recorded on
the lm.
Vitasound A short-lived type of motion picture optical sound track
system developed by Warner Bros. in the late 1930s. It used a single con-
ventional variable area sound track in the usual place on the lm and
two more, smaller variable area tracks along the sprocket holes of the other
side of the lm. These two control tracks, as they were called, controlled
the gain of variable gain amplifiers connected to loudspeakers strate-
gically located around the theater. By properly encoding the control tracks,
the source of the sound could be effectively moved around the room. Of
course, this is no substitute for stereophonic sound, but it was capable of
spectacular sonic effects. The system was compatible in that a standard

417
Vitruvius

projector would sense only the standard sound track and ignore the con-
trol tracks.
Vitruvius A Roman architect who lived about 50 b.c. He is the earliest
known writer about acoustics, having described in some detail the
acoustical characteristics of Greek and Roman theaters. Among other
things, he said that theater acoustics would be improved if sounding ves-
sels tuned to the pitches of the musical scale were placed around the seat-
ing area. His Ten Books on Architecture has been translated and is fascinating
to read today.
Vocoder An electronic device similar to a music synthesizer that is used to
synthesize the human voice and modify a voice spoken or sung into it. It
is used for special effects like changing the pitch of a voice, adding
vibrato, replacing the voiced pitch with a noise signal, etc.
Voice Coil In a dynamic loudspeaker, the cone is caused to move by a
coil of wire which is wound around a cylindrical former attached to its
center. The coil is called the voice coil, a quaint reminder of the early days
of audio.
The voice coil is immersed in a strong magnetic eld emanating from a
permanent magnet. current in the coil causes another magnetic eld to
be developed, and the two elds interact to cause the force on the coil. Cur-
rent in one direction moves the coil one way, and current in the other direc-
tion moves it the other way. Thus, the coil (and attached cone) will move
in a way analogous to the waveform of the signal applied to it.
Voice coils may be wound with relatively more or less wire of any of
several diameters. In this way, the impedance of the coil is determined.
The impedance is a combination of the resistance of the wire and the
inductance of the coil. The motion of the coil in the magnetic eld cre-
ates another component of the impedance, called the motional impedance.
This reects the dynamic characteristics of the moving system, including
the mass of the cone assembly, the springiness of the suspension of the cone,
and the acoustic load on the speaker. It is possible to determine a great
deal about the loudspeaker itself by examining and testing the impedance
versus frequency.
It is desirable to minimize the resistance of the coil because it wastes
energy, simply turning it into heat. Therefore, the more copper that can be
placed in the magnetic gap, the more efcient the unit will be, with less
power lost to heat. Sometimes voice coils are wound with at wire placed
edgewise on the former in order to get more copper in the gap.
The Voice of the Theatre, VOTT A trademark of the Altec Lansing com-
pany for their family of horn-loaded loudspeakers used in movie theatres.
Their popularity was unmatched for many decades, and they were widely
used for music and sound reinforcement.
Voice-Over A voice sometimes present in motion pictures and video pre-
sentations without the person speaking being visible, i.e., a narration.
Voice Processor, Vocal Processor A device that includes several audio
processors designed to economically replace a battery of outboard proces-

418
 Voltage Gain

sors commonly used to record a human voice or vocal signal, although any
signal can be used. It may include a mic preamp with phantom power,
a compressor/limiter, a de-esser, a multiband equalizer, and other
processors in one small package.
Volt, Voltage The voltage is the electrical potential difference between two
points; it is dened in terms of the work required to move a unit charge
(1 coulomb) from one point to the other. The unit of voltage is the volt,
named after Alessandro Volta, and is numerically equal to 1 joule per
coulomb.
The voltage of an audio signal is usually measured in terms of the rms
value of the signal, but sometimes the peak or average voltages are mea-
sured. The rms voltage squared is proportional to the amount of power
the signal carries.
Voltage-Controlled Amplier, VCA An amplifier whose gain is
adjusted, or controlled, by the application of an external direct voltage.
One of the most common uses of VCAs is in electronic music synthe-
sizers, where they are used to generate the envelope of generated sig-
nals. They are also used in compressors, limiters, and companders.
Voltage-Controlled Filter, VCF A filter whose frequency is controlled
by the application of an external direct voltage. VCFs exist in many forms,
including low pass, high pass, bandpass, and band reject. Not only the
tuning of the lter but also its q can usually be voltage controlled. VCFs
are very important components of electronic music synthesizers, where
they offer amazing control over the timbre of signals passed through them.
Voltage-Controlled Oscillator, VCO An electronic oscillator whose out-
put frequency is controlled by the application of an external direct volt-
age. VCOs are used extensively to generate musical signals in electronic
music synthesizers. The ease with which their frequency can be controlled
makes them very suitable for frequency modulation and the genera-
tion of very complex sounds. See also phase-locked loop.
Voltage Gain The output voltage of a device divided by its input voltage.
Most passive devices have a negative voltage gain, and most active
devices, especially amplifiers, have a positive voltage gain. A negative
voltage gain is sometimes called a loss or an insertion loss. Voltage gain,
by denition, is a ratio of two voltages and is therefore a dimensionless
number.
It is common practice to express voltage gains, especially of ampliers,
in terms of decibels, using the following formula:
Vout put
Gain = 20 log10




Vinput

However, this formula is valid only if the input and output impedances
are the same, and this is seldom true in practice. The problem arises
because the decibel is dened as 10 times the logarithm of a power ratio.
The square of a voltage ratio is a power ratio if the impedances across which
the voltages of the numerator and denominator are measured are the same,

419
Voltaic Cell

and this allows us to express power ratios from voltage measurements by

multiplying the logarithm by 20 instead of 10.
It is our contention that voltage gains should be simply stated as pure
ratios, with no reference to decibels. An amplier with 60 dB of gain then
would be said to have a gain of 1,000, which would be true regardless of
the loading conditions on the input and output.
Voltaic Cell Any type of electric cell, which normally cannot be recharged,
such as the leclanche cell.
Volume Expander A device for increasing the dynamic range and reduc-
ing the apparent noise of a signal. See also compander. A volume
expander decreases the system gain as the signal level decreases, making
soft signals softer still. This results in an apparent noise decrease because
the relative level between the softest and loudest sounds is greater. If the
noise level is already low enough that the signal will mask it in the loud
passages, the expansion will put the low end of the dynamic range at a point
where the ear has reduced sensitivity, making the noise less audible. See
also peak expander.
VOM Short for Volt-Ohm-Milliammeter, which is a portable device for mea-
suring AC and DC voltage, resistance, and electric current. The VOM
was the basic and venerable instrument for the electronics technician for
many years, but it has essentially been replaced by the ubiquitous digi-
tal multi meter (dmm) even though the DMM is almost useless when
trying to measure a changing voltage or current. See also vtvm.
VST, Virtual Studio Technology A proprietary digital computing sys-
tem to perform the various tasks performed in a recording studio, such
as limiting, compression, equalization, reverberation, etc. The vir-
tual studio can also act as the recording machine by digitizing the audio
signals and writing them directly to a hard disc where they can be
recalled and manipulated as desired. VST is a registered trademark
owned by Steinberg Media Technologies GmbH, a member of Pinnacle Sys-
tems, since December 2002,
VTVM Short for Vacuum Tube Volt Meter, which was an electronic, or
active, version of the passive VOM. This allowed a very high input
impedance, so it did not disturb the circuit being measured nearly as
much as did the vom, and it allowed higher sensitivity. The VTVM was
eventually made with transistors and integrated circuits rather than vac-
uum tubes as amplifiers, but the acronym still hangs on. The VTVM has
been made obsolete by the digital multi meter. See also vom.
VU VU is the abbreviation for Volume Unit, which is a measure of signal
level on a meter with a decibel scale and accurately controlled rate of
response to a signal. The reference power for the volume unit is delivered
to a 600-ohm load when the voltage is 0.775 volts rms. This is zero on
the meter, or 0 VU. Signal levels are generally less than this value, and are
read as negative decibel values.
The VU was rst used as a measure of signal level in radio broadcast-
ing, but it has been used for many years in other areas, especially tape

420
 Wavelength

recorders. It is interesting to note that a true VU meter is somewhat com-

plex in that its speed of operation (ballistics) is closely controlled. Most
meters on home-type tape recorders are calibrated in decibels, but are not
true VU meters.

W
Wall Wart Slang for a power transformer/power supply that plugs directly
into the power outlet on the wall, usually covering both outlets.
Warp Wow If the vertical pivot of a tonearm is above the surface plane
of the record being played, a warp in the record that causes the stylus to
move up and down will also cause it to move a little forward and back-
ward along the groove. Sometimes this motion along the groove is called
scrubbing. This motion is added to the relative velocity of the groove
with respect to the stylus, and it causes a small variation in frequency
that is called warp wow. Longer tonearms have less of a problem than
shorter ones, and the closer the vertical pivot to the plane of the record,
the better.
Watermarking The addition of unique digital data to an existing digital le,
in order to identify it. The user need not be aware at all of the extra data,
but it can identify les such as photos or music and would be used to detect
piracy. The term comes from the paper industry technique of embossing a
unique design into moist paper during manufacture.
Watt The watt is the metric unit of power and is dened as 1 joule per sec-
ond. The joule is a unit of energy, so power is the rate of energy transfer,
or the rate of doing work. In electrical circuits, power can be calculated
in three ways: current squared times the real part of impedance, volt-
age squared divided by the real part of impedance, or voltage multiplied
by current.
The unit is named for James Watt, the developer of a practical steam
engine and inventor of the speed governor. Incidentally, it is commonly
thought that Watts speed governor was the rst feedback control system,
but this is not true. Organ builders have been using a feedback control mech-
anism to regulate wind pressure since the thirteenth century.
WAV The Windows le extension that indicates Microsofts audio le for-
mat. .WAV les can include mono or multichannel audio at 8-bit or 16-bit
resolution at several sampling rates up to 48 kHz.
Waveform The waveform of a signal is a graph of the instantaneous volt-
age versus time. The familiar sine wave is an example. The picture seen
on an oscilloscope is a waveform.
Wavelength In a sound wave in air, the distance between two successive
pressure maxima or minima is called the wavelength, and it is equal to the
speed of sound divided by the frequency. In air at standard conditions
of temperature and pressure (STP), sound travels at about 340 meters per

421
Wavelet Transform

second, so the wavelength of a 10,000-hertz sound is 340/10,000 meters,

or about 3.4 centimeters. The wavelength in air of a 34-Hz sound (low C
sharp of a 16-foot organ pipe) is thus 340/34 = 10 meters. See also partials.
Audio signals travel in cables and wires much faster than sound travels in
air. This means the wavelengths of the electrical signals is far longer than
their equivalent sounds in air. The actual speed of transmission depends
on the characteristics of the transmission line.
Wavelet Transform A type of mathematical transform from the time
domain to the frequency domain, somewhat analogous to the fourier
transform. Whereas the Fourier transform decomposes a time wave-
form into a series of sinusoidal components that are innite in extent, the
wavelet transform decomposes the time signal into a series of tone bursts.
One disadvantage of the Fourier transform is that discrete events in the time
domain are smeared over a wide frequency range when transformed, and
discrete frequencies are spread throughout innite time when transformed
back. For this reason, nonstationary signals, which contain rapid changes
and transients, are not efciently handled. The frequency domain repre-
sentation of them does not encode their placement in time. The wavelet
transform, on the other hand, also encodes the placement of changes in the
waveform along the time axis. For this reason, it is thought to be more suit-
able for analysis of speech and music waveforms.
There are many possible variations of the wavelet transform, and they
are being investigated at present by many workers to determine their most
suitable forms. Today we have available a few wavelet analyzers to com-
plement our overworked frequency analyzers.
Weighting Weighting is a standardized frequency-variable discrimination
against a measured quantity of something, usually to make it more nearly
descriptive of the subjective perception of the quantity. An example is the
a-weighting curve applied to many measurements of sound pressure
level in order to correlate the measurements better with our loudness per-
ception. Another example is the weighting applied to measurements of
flutter and wow. The weighting curve takes into consideration which
frequencies are the most annoying to human listeners, and attenuates other
frequencies. In this way the measured level of the utter will correlate well
with the annoyance of the utter.
Wet Wet refers to a recording or a sound that has signicant reverbera-
tion present, as opposed to dry sound. The output of an effects device is
100 percent wet when only the output of the processor itself is being heard,
with none of the unprocessed signal.
Whistle Filter A low-pass filter with a cutoff frequency of 10 kilo-
hertz found in am tuners and receivers. Its purpose is to eliminate the 10-
kHz tone that would otherwise be audible if there were another station
broadcasting on an adjacent channel. The fcc has stipulated that the
frequencies for AM broadcasting be 10 kHz apart, and if a receiver has
response above this frequency, the adjacent channel carrier frequency will
be detected. When the FCC made the frequency allocations many years ago,

422
 Williamson Amplier

the maximum frequency response of receivers was considered to be 5 kHz,

and many of them then did not exceed 3 kHz.
White Noise White noise is a special type of random noise where the energy
content is the same at each frequency. Strictly speaking, true white noise
would have energy extending from dc, or zero frequency, to innitely high
frequency. In practice, we see only band-limited white noise. The sound
heard when an fm receiver is tuned between stations is quite close to white
noise over the audible frequency range.
Because of our ears peculiar method of determining loudness of
sounds, white noise sounds as if it has more energy at high frequencies than
at low. See also pink noise.
White Van Speaker Perhaps an urban myth, such loudspeakers look very
impressive, may have a brand name on them and sell for an unbelievably
low price from a street vendor in an unmarked automobile. However, when
later examined, the product turns out to be very low quality and the ven-
dor has disappeared.
Whizzer Cone A small supplementary cone attached to the voice coil of
a dynamic loudspeaker for the purpose of radiating high frequencies more
efciently than the regular cone. The whizzer cone is unsupported at the
outer edge, being simply glued to the voice coil at the point where the reg-
ular cone is attached. If a whizzer cone is to work, there must be some com-
pliance between the coil and the regular cone so at high frequencies the
regular cone is decoupled from the coil so as not to restrain the motion of
the whizzer cone.
Whizzer cones were popular in the 1950s but have fallen into disfavor
because it has been determined that they add coloration due to the non-
linearity of the added compliance. Some other loudspeaker units featured
a decoupling compliance partway up the cone. One of these was the Biex
of Altec Lansing. These have receded into oblivion for the same reasons.
Whole Step The musical interval of a major second in a diatonic scale.
The frequency ratio between the notes of a major second in just intona-
tion is 89, and in equal temperament it is the 6th root of 2, or about 12%.
See Appendix 8.
Wideband See broadband.
Wild Sound A sound recorded for a motion picture where there is no
attempt to synchronize the sound recorder to the camera, generally because
no visual synchronization is needed. An example might be the sound of
crickets for a night scene. See also double-system sound; single-system
sound.
Williamson Amplier A famous power amplifier designed in Britain
by D. T. N. Williamson in the 1950s. It used all triode tubes, with push-
pull class a output circuitry and a specially designed output trans-
former. It had a single feedback loop from the output transformer
secondary to the input tube cathode. The original Williamson put out only
about 8 watts of power, but it had a smooth and musical sound much
appreciated at the time. There were many variations of the basic circuit

423
Winding

devised by other designers, probably the most famous of which was David
Haers Ultra Linear conguration, which used a different output trans-
former with taps on the primary winding being returned to the output tube
screen grids. This increased the power output and efciency with not too
much penalty in added distortion.
Winding The wire coils that compose the primary and secondary of a
transformer are called windings.
Windscreen Although most microphone diaphragms are protected by
perforated metal or a mesh of some sort, these do little to prevent a blast
or wind or breath from overloading the diaphragm. Therefore, various
devices can be added to a microphone to prevent this. The rst level of pro-
tection may be built into the structure of the microphone, especially a hand-
held vocal microphone, as a pop or breath filter, and is usually a piece
of open-cell foam. A ball of such foam can be tted over most mics if needed.
Another tactic, especially with studio mics, is to place a ring with gauze-
like material stretched over it between the vocalist and the microphone.
Recording outdoors is especially tricky from a wind standpoint, and a device
resembling, and called, a blimp or zeppelin is tted over the entire micro-
phone, with several inches of airspace between the mic and the blimp shell.
The blimp may be lined with acoustic foam and mesh, and for very seri-
ous air movements, the blimp may be covered with a porous synthetic fur.
See also pop filter.
Wiper The movable contact in a potentiometer is called the wiper or,
sometimes, the arm.
Wireless Microphone A microphone with a miniature radio transmitter
built in. The wireless is used with a remote receiver that picks up the sig-
nal. Wireless microphones are a good choice when a cable would be trou-
blesome, such as when performers must move over a considerable area, or
when the microphone must be concealed.
The quality of wireless systems is improving, but there are still problems
with their use. Interference with other broadcast signals, such as citizens
band equipment and radio-dispatched taxicabs, can be an embarrassment.
Also, the signal is subject to multipath distortion in some cases where
the radio frequency can be reected around inside a building. Many of
these problems are much improved by the use of a diversity receiver.
Woofer The woofer is the low-frequency loudspeaker, or driver, of a
multiway loudspeaker system. Sometimes if frequencies below 30 or 40
hertz are to be produced, it is called a subwoofer. Woofers generally are
quite large, usually 12 or 15 inches in diameter, or sometimes 18 or even 24
inches. At one time, Electro-Voice made a giant woofer of 40-inch diameter.
Sheer size, however, is not the most important attribute of a woofer. A
loudspeaker cone does not vibrate as a simple piston at all frequencies;
it breaks up into several modes of vibration, each one associated with
a different frequency. Normally, the cone would not be operated at fre-
quencies above the second mode of vibration, because its frequency
response and directional characteristics are quite irregular and unpre-

424
 X-Copy

dictable in this region. The larger the cone, the lower the frequency of the
second mode and the lower the upper cutoff frequency must be to avoid
these problems.
One of the most important developments in low-frequency reproduc-
tion with low distortion was the acoustic suspension principle invented
by Harry Olson of RCA Research Laboratories and rst commercialized in
1954 by the Acoustic Research company. This system uses a relatively small
woofer cone (8 to 12 inches in diameter) with the capability of very large
amplitude of vibration. The cone is made quite massive and very loosely
suspended. The box behind the cone is sealed, and the springiness of the
air in the box is the major contributor to the force that returns the cone to
the rest position. This air spring is linear, and allows the cone to move
relatively great distances without distortion. The box is lled with a
damping material such as berglass or mineral wool, partly to damp inter-
nal reections of sound and partly to further linearize the air spring.
This invention was the rst to permit excellent low-frequency perfor-
mance from relatively small enclosures. The compromise in this type of sys-
tem is that the efficiency is reduced from that of larger systems, meaning
that more amplifier power is needed for the same output sound level.
See also bass reflex.
Word Clock The digital timing reference signal that is the same frequency
as the sampling rate being used. In digital audio and video facilities with
many digital devices, a single master word clock generator may be used
as the timing reference for all devices, minimizing timebase jitter that might
occur if the separate digital devices all used their own clocks for timing ref-
erence. There are also digital system clocks that are multiples of word clock,
with claims for higher accuracy. See also house synch.
Wow Wow is a relatively slow variation in frequency of reproduced sound
caused by slow speed variations in records, tape recorders, etc. pitch uctu-
ations of one or two per second or fewer are classed as wow, while faster
variations are called flutter.
The term was rst used during the early days of sound motion pictures.
When threading a movie projector with the sound head activated, mov-
ing the lm by hand makes a sound something like the word wow.
Wrap In an analog magnetic tape recorder, the tape wrap angle is the angle
made by the tape as it approaches and leaves the tape head. Wrap is actu-
ally a measure of the distance along the face of the head where the tape
makes contact. This distance has an effect on the low-frequency response
of the playback head. See also head bump.

X
X-Copy An identical copy of pre-existing audio or video material used in
movie editing.

425
X curve

X curve Extended curve. In the motion picture sound industry the X curve,
also known as the wide-range curve, is a specication for the acoustic
response curve of a movie theater. It is dened by ISO Bulletin 2969, which
species at the listening position in a dubbing situation or two-thirds of
the way back in a theater that the acoustic frequency response be at
to 2 kHz, rolling off at 3-dB/octave above that. The input signal is
specied as pink noise. The small-room X curve is designed to be used
in rooms with less than 150 cubic meters, or 5,300 cubic feet. This standard
species at response to 2 kHz, and then rolling off at a 1.5 dB/octave rate
of roll off. Some people use a modied small-room curve, starting the
roll off at 4 kHz, with a 3-dB/octave roll-off rate. The X curve replaced the
old Academy Curve, and the X curve is meant to be used for monaural and
Dolby Stereo analog sound tracks. The X Curve must be met in all theaters
that are certied to the thx specication.
Xformer Abbreviation for transformer.
XLR Connector In its three-conductor form (XLR-3), a common audio con-
nector, widely used in balanced professional audio applications and for the
aes3 (AES/EBU) digital interface. The XLR connector was a trademark of
the Cannon company, but usage has made it a generic term and an inter-
national standard (as IEC 268-12), and many manufacturers make such con-
nectors. Pin number 1 in the XLR plug is always connected to the shield.
The reason is that the connectors are so designed that pin 1 makes contact
rst, ensuring that the ground connection is made before the signal con-
nection. This greatly reduces the transient thumps and pops that can
occur when a circuit is patched with the power turned on. IEC standards
now specify that pin 2 is wired to the high or plus (+) wire of a balanced
pair, and pin 3 to the low or minus () wire. Unfortunately, there was no
standard several years ago, and one still nds items wired with pin 3 hot.
The XLR connector is called a QG connector by one manufacturer, for Quick
Ground.
XM Satellite Radio See satellite radio.
Xmtr Short for transmitter.
Xophonic An articial reverberation device for the home made by Radio
Craftsmen in the 1950s. The Xophonic was a box that looked about like a
bookshelf loudspeaker. It contained an analog time delay device in the
form of a small loudspeaker connected to a coil of tubing about 50 feet
long with a microphone in the other end, producing a time delay of about
50 milliseconds. The incoming signal from the amplifier of the sound
system was fed to the time delay unit and to a mixer, which combined it
with the delayed signal from the tube microphone. A power amplier and
loudspeaker were included to amplify and radiate this combined signal
into the listening room.
The Xophonic was probably the rst signal processing device intended
for home use. It enjoyed a brief popularity, and then quietly fell into obliv-
ion, aided by the advent of the stereophonic record.

426
 Zenith

X-over Short for crossover. See crossover frequency; crossover

network.
Xtal Abbreviation for crystal.
XYL Amateur radio operator slang for wife. The term was popular in the
1950s and earlier, when ham radio was dominated by males. See also yl.
X-Y Stereo See intensity stereo.

Y
Yagi An early type of directional receiving antenna, mostly used for tele-
vision reception, invented by Hidetsugu Yagi in 1928.
Y-Cord A Y-cord consists of two short audio cables with similar connec-
tors on one end that are joined to a single connector, usually of similar type
but different gender. The Y-cord, or Y-connector as it is sometimes called,
is used to create split feeds, where the same signal is sent to two differ-
ent places at the same time. Sometimes a Y-cord is used as a mixer to short
the two channels of a stereo signal together to make a monaural sig-
nal. This can lead to problems in low-impedance circuits, because each
output circuit loads the other one. The solution is to place a series resis-
tor in each signal path.
Also, a Y-cord can never be used to connect two channels of digital
data together!
Yellow Book See red book.
YL Amateur radio operator slang for girlfriend; YL stands for young
lady.

Z
Z The symbol for impedance.
Zener Diode The zener diode is a solid-state silicon diode that is
reverse biased to such an extent that it conducts current backward. It
is then said to be operating in the avalanche mode, and is sometimes called
an avalanche diode. Under these conditions, the voltage across it will be
very nearly constant over a range of current values. This constant voltage
property is exploited by using it as a voltage regulator or voltage reference
in power supply circuits. Zener diodes are available in a wide range of volt-
age values.
Zenith The angle that the face of the tape head in an analog tape recorder
makes with the top plate of the machine. The zenith should be 90 degrees,
and if it is not, the head will not wear uniformly. Zenith angle is adjustable
on some machines.

427
Zero Reference

Zero Reference A standard signal level, usually recorded on an align-

ment tape, used to set the reproduce gain control on a tape recorder. This
level may or may not result in a zero vu reading on the tape recorder meter,
depending on the philosophy espoused by the recording engineer. See also
reference level.
Zipper Noise Audible artifacts or loudness steps heard in the audio when
being changed by a digital system. Digital controllers have a nite num-
ber of discrete steps when controlling a signal parameter such as loudness,
and under certain circumstances they can be heard. Some companies claim
to use smoothing techniques to eliminate the problem. It is compared to
the sound of the clothing fastener of the same name.
Zobel Network An R-C network consisting of two resistor-capacitor series
circuits connected in parallel. The network is connected across the input
terminals of a loudspeaker, where the impedance of the network is
designed to cancel the inductance of the speaker voice coil. The result
is a more nearly resistive load on the crossover, smoothing the high-
frequency amplitude and phase response of the crossover-speaker com-
bination. In a two-way system, the woofer and tweeter can be corrected
with a Zobel network associated with each driver.
Zoom Microphone A type of microphone system consisting of three car-
dioid microphone elements and a special phase correction and equal-
ization circuit. By varying the position of a control knob, the microphone
outputs are combined in such a way that the directivity of the array
changes from omnidirectional through cardioid to super-cardioid.
The control can be synchronized with the control of a zoom lens on a video
or motion picture camera so that the auditory perspective changes with the
visual perspective.

428
Appendix 1
The Art and Science
of Good Acoustics

The systematic study of musical acoustics existed as early as the rst century
b.c., as documented by Vitruvius in his Ten Books on Architecture. Great strides
in understanding were achieved in the nineteenth century, especially with
the work of Hermann Helmholtz. The twentieth century has spawned a near
explosion in the eld, as witnessed by the great number of recent publica-
tions on all elds of acoustics.
Research in physical acoustics has produced theories leading to improve-
ments in the design of musical instruments and has greatly inuenced the
design of concert halls built in the last 100 years. Research in psychoacoustics
has likewise inuenced instrument design and auditorium acoustics.
One might expect that all this activity would have produced better instru-
ments, better music, and happier musicians. This appendix attempts to show
that this has not always happened. Some observations on the reasons for this
are presented, along with some suggestions for further research. The areas
considered are (1) auditorium acoustics, (2) psychoacoustics, and (3) musical
instrument design.

Fads in the Design of Auditoriums

The rst truly scientic architectural acoustician was Wallace Clement Sabine,
professor of physics at Harvard University. He is credited with the acousti-
cal design of Symphony Hall, built in Boston in 1900. Sabine investigated the
measurement of reverberation time and devised the rst formula to predict
reverberation time from the dimensions of a room and the amount of sound
absorption in it. He invented the concepts of acoustical absorption coefcients
and absorption units, now called sabins. (One sabin is dened as one square
foot of total sound absorption. The absorption of a given area of a material
may be predicted by multiplying its absorption coefcient by its area in square
feet.) Sabine also did much work on the measurement of absorption coefcients
of many materials. He was probably the rst to suggest that materials be man-
ufactured for the express purpose of sound absorption, in effect making him
the father of acoustical tile.
Symphony Hall has been universally claimed by musicians and audiences
to possess superlative acoustics. Many would agree with the 1950 opinion of
Rudolf Elie, music critic of the Boston Herald, It is very clear to me now that
Symphony Hall is the most acoustically beautiful hall in the United States. It

429
Appendix 1

is to the orchestra what a Stradivarius is to the great violinist in providing a

sound box of the utmost brilliance and sensitivity.
While Sabine deserves credit for many of the halls acoustic features, he
was also very lucky that even though all of the halls design features were
not planned specically for acoustic reasons, they contributed positively to
the halls qualities. Examples are the basic shape of the space, the large amount
of sound-diffusing surfaces that are relatively broadband in action, and the
well-placed natural frequencies of the wall and ceiling surfaces.
Later in the century, as a sequel to Sabines work, the measurement of rever-
beration time in halls became fashionable, and along with this came the abil-
ity to control the reverberation time by the use of special sound-absorbing
materials. The control of reverberation rapidly attained the status of an archi-
tectural fad. Existing auditoriums and all types of public buildings were cor-
rected by liberal applications of acoustical tile, or acoustical plaster. Many
architects, acousticians, and especially acoustic tile salesmen were convinced
that all manner of acoustical defects could be cured simply by the liberal appli-
cation of sound-absorbing material. Books on architecture and musical
acoustics began to appear, nearly all citing reverberation time as the most
signicant aspect of the acoustics of a room, in spite of such an authority as
Vern Knudsen writing in 1963, A felicitous shape is a requirement of the high-
est priority. Unfortunately, many architects believe that faulty shapes can be
corrected by covering the offending surfaces with highly absorptive materi-
als and by adjusting the reverberation time. Thus deluded, they adopt a fash-
ionable construction method, such as a concrete shell, and produce a building
that is an acoustical perversion. A bad shape is a permanent liability.
It became apparent very early that reverberation time could be made to
have different values at different frequencies. It was noted that many ex-
isting buildings had much longer reverberation at low frequencies than at
high frequencies. The reaction to this observation was to introduce selec-
tive absorption so that the reverberation would be nearly constant with fre-
quency, the rationale being that the overall tonal quality would be more uni-
form over the musical pitch range. It was thought that longer low-frequency
reverberation caused the music to sound muddy and ill-dened and that
uniform reverberation vs. frequency would improve musical clarity and
denition.
The result of this thinking was the construction of many new music audi-
toriums with very hard, crisp, or tinny sound with very little warmth and
solidity. In a few cases the trend was carried to such an extreme that the high-
frequency reverberation time actually exceeded that of low frequencies. Of
course the musicians knew better all the time and acousticians and architects
began to be held in contempt.
In the November 5, 1955, issue of the New Yorker, Joseph Wechsberg wrote,
Most of the people who have set themselves up as consultants on matters
of acoustics contend, not unnaturally, that by applying certain laws of physics
and using certain testing devices they can determine in advance how hos-
pitable to sound a new auditorium will be. The fact is, however, that several

430
 Appendix 1

auditoriums built in Europe recently under the guidance of consultants who

presumably applied the laws of physics and using the testing devices have
turned out to have dreadful acoustics.
The eminent American acoustician Leo Beranek read these words and was
forced to admit that most modern auditoriums were not free of criticism.
Beranek then undertook a systematic study of fty-four concert halls located
in various parts of the world. He identied several acoustical characteristics
other than reverberation that contribute in large measure to the acoustics of
halls. One of these measures is the Initial Time Delay Gap, or ITDG as Beranek
called it. The ITDG is dened as the time interval between a listeners hear-
ing of the direct sound from a source and the earliest reected sound from
the walls or ceiling. It is a measure of the subjective intimacy of a music
hall, a shorter ITDG corresponding to the subjective impression of smaller
rooms and vice versa. Beranek reasoned that a large concert hall must have
a high ceiling to attain sufcient reverberation, but should have a low ceil-
ing to provide a sufciently short ITDG for subjective intimacy. The two
requirements being mutually exclusive, led Beranek to the invention of rela-
tively small sound-reecting panels (called acoustical clouds) suspended from
the true ceiling and forming a second lower but partially acoustically-
transparent sound ceiling. The reasoning was that sounds reecting from the
clouds would provide a short ITDG, while sounds passing between the clouds
would excite the reverberation of the entire room volume.
Leo Beranek was the acoustician for Philharmonic Hall at Lincoln Center
in New York, completed in 1962. His well-written and researched book, Music,
Acoustics, and Architecture, published in 1962, tells the story of the halls design
in painstaking detail. The book is fascinating, both in regard to its technical
completeness and in respect to the complex fate of Philharmonic Hall since
its opening. As is well known, the hall was subjected to criticism from all areas
of music from its opening, and several expensive attempts to correct it were
made, ending in a $750,000 remodeling in 1969. This remodeling purportedly
eliminated the clouds but retained the same shape. Even after this, musicians
and music lovers were not happy with the acoustics and nally the building
was gutted and a completely new concert hall constructed within the four
walls at a cost of $17,000,000. The new auditorium was named Avery Fisher
Hall in honor of the man who donated most of the needed funds. The new hall,
designed by the acoustician Cyril Harris, is similar in shape to Symphony Hall
in Boston, the hall the New York Philharmonic Society had asked Beranek to
emulate in the rst place (letter from George Judd, Jr., to Max Abramovich,
dated April 20, 1959, quoted in Beranek, op. cit.). At last the controversy over
Philharmonic Hall seems to have ended, but the obvious question is, What
went wrong? Beranek is in every sense a respected and accomplished
acoustician and it is obvious from his book that he did his homework. The
answer must come from psychoacoustics: the study of our perception of what
we hear.
There is an excellent, but relatively unknown, book by Georg von Bksy
called Sensory Inhibition (Princeton University Press, 1967), that summarizes

431
Appendix 1

much of Bksys research spanning many years. There are some strong clues
to why the clouds at Lincoln Center did not behave as expected. Bksy points
out that an array of acoustical clouds will reect high frequencies while
permitting low frequencies to pass through and be reected from the true
ceiling above. This means the high frequency portion of complex musical
sounds reach the listener earlier than the lower frequency components of
those sounds. Bksy had established earlier that sounds arriving in quick
succession at the ear result in complex nervous system processing whereby
a part of the response is inhibited. He called this phenomenon auditory inhi-
bition. He states, In the concert hall the low frequencies will lose their effec-
tiveness in brief pulses of sound heard by a listener. These tones are physically
present but are inhibited because they are delayed in arrival at the listeners
ears. This type of distortion may be called room acoustics phase distortion
and represents an interesting new eld. The construction of auditoriums and
the subsequent discovery of these distortions represent very expensive
experiments on inhibition. Bksy then points out that this type of phase
distortion is not noticed if it is slight, but when it is increased, a threshold is
reached where suddenly the unexpected happens. Such thresholds exist also
for other sensory phenomena. These sensory thresholds present barriers
that often prevent technical extensions into what otherwise might be unlim-
ited elds.
The lesson from this is that architectural acousticians must be mindful of
psychoacoustics and sensory inhibition and must be wary of extrapolating
known designs into larger scales. Also, perhaps a different point of view should
be sought. Siegmund Levarie and Ernst Levy bemoan the fact that architec-
tural acoustics is still thought of in terms of reverberation and absorption of
sound. In other words geometrical acoustics, or ray acoustics, has been used
because of its ease of theoretical simplication and measurements. When
addressed from the standpoint of volume resonance and wave theory, room
acoustics measurements seem much better correlated to what a person hears
with his complex auditory mechanism. There has been recent work, much of
it instigated by Manfred Schroeder of Gottingen, on the importance of strong
side wall reections and diffused ceiling reections in concert halls; this work
appears promising. New auditoriums designed to provide strong reection
from the side walls across the audience have been designed recently by some
acousticians, notably Paul Veneklasen of Los Angeles; these rooms are praised
by musicians and audiences alike. The Japanese acoustician Yoichi Ando, in
collaboration with Manfred Schroeder, has done much work to place a rm
theoretical formulation under this concept. In this work the somewhat sim-
plistic ITDG is being replaced by the IACC, or Interaural Cross Correlation
function, which is a mathematical measure of the sameness of the sounds
reaching the two eardrums. Ando has been able to establish a connection
between the IACC function and the degree of preference expressed by listeners
to different sound elds. Perhaps this will allow innovative concert halls to
be designed and still attain acceptance by musicians.

432
 Appendix 1

Psychoacoustics, Hearing Threshold,

Critical Bands, and Earphones
In efforts to determine how our auditory system works, a great many exper-
iments have been performed that effectively isolate various hearing phe-
nomena. One of the best known of these was the measurement of equal
loudness contours by Harvey Fletcher and W. A. Munson in the early 1930s.
By presenting pure tones of many frequencies and intensities to various sub-
jects, it was shown that the ears sensitivity varies with the frequency of the
test tone. Most notably, tones of low frequencies must be many times more
intense than high frequency tones to be judged equally loud. Also the audi-
tory threshold, or least intensive sounds that can be heard at a given frequency,
is drastically higher at low frequencies and at very high frequencies compared
to the most sensitive region of about 3000 Hz. Any sound must be above the
threshold in intensity at its frequency to be heard at all and presumably any
sounds below threshold will not be heard.
This observation holds true, of course, for single pure tones, but when the
ear is subjected to more than one tone at a time, things get complicated quickly.
It has been shown that the presence of any tone that is above threshold will
raise the auditory threshold for other frequencies. In other words the pres-
ence of a single relatively intense tone can cause a less intense tone of another
frequency to be inaudible. This phenomenon is called masking; masking con-
tours have been plotted that show under what conditions of frequency and
intensity tones can be heard. These masking contours have different shapes
for masking tones of different frequencies and intensities, making the situa-
tion somewhat complicated.
Examination of this masking data leads one to believe that many instances
occur in common experience when sounds cannot be heard even though they
are known to exist. This has been the case in the eld of electronic reproduc-
tion of music, known popularly as high delity. Music reproducing equip-
ment always adds some spurious signals to the music in the form of distortion
of various kinds. Distortion has been reduced as equipment has become more
sophisticated, especially in the case of ampliers. Some amplier manufac-
turers claim the distortion added by their equipment is inaudible because of
its extremely low intensity and its being masked by the music itself. Many
people have been skeptical of these claims, believing that they could easily
tell differences between various ampliers all having very low distortion.
An interesting study was performed in Finland and reported in the Jour-
nal of the Audio Engineering Society in 1978 (cf. Petri-Larmi, Otala, and Lamuras-
niemi). A special machine was built to introduce controlled amounts of
Transient Intermodulation Distortion to recorded musical examples. One sur-
prising result was that the detection threshold was .003% rms distortion for
piano recordings. The authors acknowledge that this is difcult to explain
because the distortion components are so low in intensity as to be below the

433
Appendix 1

hearing thresholds of the subjects! In other words, adding certain sounds that
are so low in intensity that they are inaudible by themselves to a piano record-
ing will result in a detectable difference in the piano sound. Clearly more work
needs to be done in audibility of low intensity sounds, and the Fletcher-
Munson curves do not predict the audibility of some complex sounds.
Generally, psychoacoustic testing is performed with earphones rather than
loudspeakers in order to facilitate exact control of the signals reaching the ears.
Audiometric testing is always done with earphones, presumably for the same
reason. But earphones present formidable problems in standardizing sound
levels at the eardrum and in attaining standardized frequency response with
different listeners. The seal between the phone and the head greatly affects
the low frequency response and the shape of the listeners pinna greatly affects
the middle and high frequency response. Greater than 10 dB frequency
response discrepancies between subjects using the same earphones have been
reported (Sank, op. cit.). Also the presence of the earphone on the ear increases
the audibility of the listeners own physiological processes. This is somewhat
akin to hearing the ocean roar when holding a seashell to ones ear. Those
extraneous noises can act as masking signals and could, it would seem, inval-
idate audiometric testing. Is it possible that critical bands and loudness con-
tours are artifacts caused by using earphones? In any case it is obvious that
earphone listening is an unnatural condition for our hearing mechanism, and
it seems that insofar as possible, psychoacoustic testing should be done under
free eld conditions using both ears.
Problems can occur when trying to apply the results of psychoacoustic
testing to musical experiences. Most experiments aimed at investigating our
hearing ability have used sine waves (pure tones) as independent variables.
One phenomenon that was noticed early is that the perceived pitch of a tone
varies with intensity and is not directly correlated with its frequency (Stevens,
op. cit.).
Low pitches are made subjectively lower with increasing intensity and high
pitches are made subjectively higher with greater intensity. One might expect
this to explain out-of-tune playing among orchestra members, but it happens
that according to some researchers the effect disappears with complex tones!
However, this has not been shown to be true under all conditions and more
studies need to be done. In summary we think it would be desirable if psy-
choacoustic research could be conducted with sounds that more clearly imi-
tate natural sounds, and the use of earphones should be approached with
caution.

Instrument Design Piano and Stringed Instruments

It has been known for many years that the overtones present in the sound of
the piano are not accurately tuned to the frequencies of the harmonic series,
but are in fact higher in frequency. This sharpening of the overtones increases
with the order of the overtone. Because the piano is tuned in octaves up and

434
 Appendix 1

down from a central temperament octave, the piano will be tuned progres-
sively sharp in the treble octaves and progressively at in the bass. On the
surface it would appear that this is a defect in the piano. As long ago as 1949,
F. Miller proposed that if small masses were applied to the strings near their
ends, the inharmonicity could be cancelled. He proposed that the piano could
be improved by this method, but there seemed to be essentially no interest
among piano makers to adopt the improvement. Then in 1962, Harvey
Fletcher, et al., reported that synthesized piano tones without inharmonicity
were not preferred by listeners over synthetic tones that did contain inhar-
monicity. The non-stretched tones were described as dull, uninteresting, and
not piano-like. After a series of similar listening experiments, the Conn com-
pany came to a similar conclusion. Thus it turns out that a defect in the piano
tone, as discovered by physical measurement of the sound, is actually an
important factor in the quality of the instrument.
Stringed instruments such as the violin, viola, cello, and bass viol have been
built for so many years that one would expect their design to have been opti-
mized by now. Many people have studied most aspects of stringed instru-
ments and the consensus has been that the violin has been optimized insofar
as its size and shape in relation to its pitch are concerned. Acoustically the
violin has its resonances properly placed to ensure a uniform tonal output
and its size allows relative ease of playing. The cello is also close to an ideal
size for its pitch range. The viola and string bass, however, are too small phys-
ically to radiate optimally in their pitch range. Carleen Hutchins of the Cat-
gut Acoustical Society has studied this situation and has proposed new designs
for the viola and bass that nearly optimize their size and shape to their respec-
tive ranges. Because of its larger size, the Hutchins viola must be played ver-
tically in the manner of the cello. According to one violinist who has mastered
the instrument, this is not an important advantage. The so-called Great Bass
is signicantly larger than its double bass cousin, making it somewhat awk-
ward to play, at least by persons of normal size. These instruments are char-
acterized by a greater production of the fundamental frequencies, especially
in the lower register. The difference is quite striking.
Hutchins envisioned the use of the new instruments in orchestras where
their greater strength in the low register would be welcome. Perhaps at last
the viola section could compete on a more equal basis with the violins, and
perhaps the Great Basses would nally provide the rich, strong support in
the low frequencies that has been lacking in symphony orchestras. However,
sufcient time has passed that these instruments should have made denite
inroads in orchestra makeup if they are going to do so. The Great Bass in par-
ticular was praised by Leopold Stokowski; this alone should have encour-
aged its introduction and use, but no such revolution has occurred. The initial
interest and excitement has seemingly dissipated.
Can this fact be attributed to a simple inertia and unwillingness on the part
of musicians to try something new? Is it simply a matter of time before we
see vertical violas and Great Basses in all our symphony orchestras? We think
not. We believe that even though the new instruments may be optimized from

435
Appendix 1

the standpoint of their acoustical performance in some sort of absolute sense,

they are not necessarily well suited to performance of the existing body of
symphonic music. It must be remembered that composers have had to work
around the so-called inadequacy of existing instrument design, and the
structure of the music has been designed to exploit the tonal characteristics
of existing instruments. The introduction of a relatively new and different
sound to the performance of this music must be expected to meet with
signicant resistance. Innovations in instrument design probably should not
exhibit great differences in tone quality if they are to be accepted by the musi-
cal community. An example is the transverse ute designed by Theobald
Boehm in the nineteenth century. Boehms work resulted in greater ease of
playing and improved tuning in both the ute and clarinet, and musicians
readily accepted the changes. The basic sound of the instruments was changed
very little. Perhaps our research efforts should be aimed at greater ease of per-
formance and tuning stability of instruments rather than at the creation of
new sounding instruments that correct the deciencies of existing instru-
ments as measured by scientic instrumentation and techniques.

References

Beranek, Leo. Music, Acoustics, and Architecture. New York: John Wiley & Sons,
1962.

Fletcher, Harvey, et al. Quality of Piano Tones. Journal of the Acoustical Soci-
ety of America, Vol. 36.

Hutchins, Carleen. Founding a Family of Fiddles. Physics Today, Vol. 20, Feb-
ruary 1967.

Knudsen, Vern O. Architectural Acoustics. Scientic American, Vol. 209, No.5,

November 1963.

Levarie, Siegmund, and Ernst Levy. Tone: A Study in Musical Acoustics. Kent,
OH: Kent State University Press, 1968.

Miller, F., Jr. Proposed Loading of Piano Strings for Improved Tone. Jour-
nal of the Acoustical Society of America, Vol. 21, 1949.

Petri-Larmi, M., M. Otala, and J. Lamurasniemi. Psychoacoustic Detection

Threshold of Transient Intermodulation Distortion. Journal of the Audio
Engineering Society, Vol. 28, No. 3, February 1980.

Roederer, Juan. Introduction to the Physics and Psychophysics of Music. Heidel-

berg, Germany: Heidelberg Science Library, 1975.

436
 Appendix 1

Sank, J. R. Improved Real-Ear Tests for Stereophones. Journal of the Audio

Engineering Society, Vol. 28, No.4, April 1980.

Stevens, S. S. The Relation of Pitch to Intensity. Journal of the Acoustical Soci-

ety of America, 1935.

Terhardt, E. Inuence of Intensity of the Pitch of Complex Tones. Acustica,

Vol. 33, 1975.

437
Appendix 2
Some Frequently Used
Symbols and Units

Table 1

Prexes

Prex Multiple Symbol

deka ten times d, or da

hecto hundred times h
kilo thousand times k
mega million times M
giga billion times G
tera trillion times T

deci tenth part d

centi hundredth part c
milli thousandth part m
micro millionth part
nano billionth part n
pico trillionth part p

438
 Appendix 2

Table 2

Symbols and Names of SI Units

Name Quantity Symbol Unit

kilogram mass kg
meter length, distance m
second time s
ampere electric current A coulombs/s
coulomb electric charge C
farad capacitance F
henry inductance H
hertz frequency Hz cycles/s
joule work, energy J newton-meters
newton force N
pascal pressure Pa newtons/m2
siemens electric conductance S
tesla magnetic ux density T
volt electric potential V
watt power W joules/s
weber magnetic ux Wb
ohm electric resistance volts/ampere

439
Appendix 2

Table 3

Decibel Values

Power Ratio Voltage Ratio Decibel Value

1 1 0 dB
2 1.4 3 dB
4 2 6 dB
10 3.16 10 dB
100 10 20 dB
1,000 31.6 30 dB
10,000 100 40 dB
100,000 316 50 dB
1,000,000 1000 60 dB

By memorizing Table 3, one can easily approximate calculations involving

decibels and power and voltage ratios.
Adding decibels is equivalent to multiplying power ratios. For instance,
adding 3 dB to a power level multiplies it by 2, and adding 6 dB multiplies
it by 4. Therefore, 36 dB (30 dB + 6 dB) is equivalent to a power ratio of 4,000
(1,000 x 4); and 23 dB (20 dB + 3 dB) equals a power ratio of 200 (100 x 2), etc.
Adding decibels is also equivalent to multiplying voltage squared ratios.
Thus, 26 dB is equivalent to a voltage ratio of 20 (10 x 2), and 43 dB equals a
voltage ratio of 140 (100 x 1.4), etc.

440
Appendix 3
How to Subdue a Hi-Fi Salesperson

Who has not innocently walked into a stereo store to listen to certain com-
ponents in a leisurely manner, only to be inundated with technical jargon (or
pseudo-technical jargon) by a salesperson who is seemingly intent on mini-
mizing the contents of your pocketbook, with apparent disregard for your
sonic requirements and desires? What is the best way to handle such a situa-
tion? How can one make meaningful comparisons of loudspeakers, ampliers,
record players, etc., in the presence of such a confusing array of equipment,
almost each component of which represents a breakthrough in technology?
Well, we believe it is possible to make some sense out of a seemingly hope-
less situation.

Loudspeaker Evaluation. The rst thing your salesperson is likely

to say is that you must compare the sounds of various loudspeak-
ers switching between them while playing a recording of some sort.
Surely you will be told that everybody hears differently and what
sounds accurate to you may seem inaccurate to others, and sup-
posedly you will be able to select the most natural sounding unit
easily. However, this has been shown many times over to be untrue.
Comparing the sounds of loudspeakers might allow you to select
the most pleasant-sounding one with that particular CD, but this has
almost no relation to absolute accuracy.
Accuracy in reproduction is not a subjective effect, but is objec-
tivenot a characteristic of perception, but an actual characteristic
of sound itself. How, then, can you select the most accurate loud-
speaker out of a group? You must compare the sounds of the speak-
ers to a live sound with which you are familiar. A good test sound
is a male voice. If possible, arrange to record the voice of a friend
whose voice you know well, preferably a male friend. The record-
ing should be done either in a very large space or outdoors, and the
microphone must be of as high a quality as possible. Be sure the
microphone is at least three feet from the talker to avoid recording
proximity effect on the tape. A good quality cassette recorder is ade-
quate for this. Then, accompany your friend to the stereo store, hand
the salesman the recording and ask to hear it through various loud-
speakers. Have your friend stand near the loudspeaker in question,
have him speak, and then adjust the loudness of the reproduced

441
Appendix 3

sound of his voice to match the loudness of the live voice. This is
important. You will very quickly nd that almost none of the speak-
ers sound natural.
Listen particularly to the low-frequency tones and note the res-
onant boominess projected by many systems. Then listen to the
sibilant sounds, and note the great differences in the way they are
handled by the various speakers. At the start of the test, the tone con-
trols in the amplier should be set for at response, and you can
experiment with different settings to increase the naturalness of the
sound. You will nd, however, that many speaker systems will not
sound natural, no matter what the tone control settings are.
While it may seem that a voice will not exercise the full frequency
range of a loudspeaker system, and this is true in the strict sense,
the male voice does cover the part of the frequency range most impor-
tant to music. It is interesting to note that this is a much more severe
test than simply to tune in a radio announcer you may be used to
hearing and use this voice as a test signal. The latter is not good for
two reasons: you are familiar with the sound of the voice heard over
the radio and not in person; and radio broadcasters nearly always
use directional microphones at short range to pick up announcers,
causing very large increase in bass response due to proximity effect.
Some have called this the Jolly Green Giant effect, and it makes
this a hopeless test signal.
If you are not able to do this test, another simple experiment is
enlightening. Simply tune an FM tuner in the store to a point
between stations and listen to the random noise thus produced. (Be
sure the tuner used does not have interstation muting, or turn the
muting off.) This random noise is devoid of any musical pitch, being
made up of a collection of all frequencies mixed in a completely
chaotic manner. It exercises the loudspeaker in all frequency ranges
simultaneously, allowing you to hear anomalies over the whole audio
range. The noise should have no perceived musical pitch at any pitch
level. Switching between speakers will immediately show up those
that have peaks, or resonances, in the audible range. Listen especially
at very high frequencies and very low frequencies. It takes practice
to learn how to listen to random noise. (It always must be reproduced
by some sort of loudspeaker to be heard, and none are perfect.) Nev-
ertheless, this is a very easy task to perform, and it will tell you a
great deal about the frequency responses of loudspeakers. Also, it
is fun to watch the expression on the salespersons face.

Phono cartridge evaluation. Probably the most objective way to eval-

uate a cartridge is to use a special test record that contains a record-
ing of pink noise, which is random noise that has been equalized
so it contains equal energy in each octave band of frequency. (The
noise heard between FM stations is not pink noise, but it is an approx-

442
 Appendix 3

imation of it. True pink noise is somewhat difcult to generate.) Some

dealers may have such a test record. Simply play the record with
different cartridges, using the most neutral-sounding speaker sys-
tem, selected from the previous test. Listen for any signs of denite
pitch, especially in the very high frequencies. Also, play various
records and listen in particular to the timbre of the music. In gen-
eral, classical music is best for this type of evaluation because pop-
ular music is usually highly compressed in dynamic range and also
greatly equalized.
Listen to the tone color of percussion, especially cymbals. Note
if the tone color is similar when playing various records, or whether
each record sounds individual, as it should. Similarities in tone color
are due to the cartridge, which has characteristic timbre of its own
that it impresses on the reproduced music. Find the cartridge which
seems to have the least inuence on the tone color of various types
of music. One type of music which will show up high-frequency
anomalies is the massed strings of a symphony orchestra. These com-
ments also apply to loudspeaker evaluation, so when listening to
cartridges, use only one speaker.
These tests are aimed at the transducers, the devices which alter
the sound more than other audio components. Evaluation of CD
players, ampliers, tape recorders, etc., is more difcult but also less
important, most such units sounding quite similar. Perceptual encod-
ing recorders and players can be a different matter.
There are some ways to evaluate CD players, whose digital to ana-
log converters can vary in quality, and whose error correction and
concealment systems differ. Try the pink noise test described above
for phono cartridges, using a test CD. In addition pay particular atten-
tion to the very soft sounds, and sounds fading into silence, using
headphones if necessary. Avoid a player that makes fading sounds
just above silence seem to splatter, or sound like static.
The error correction and concealment characteristics of players
differ widely. The sounds players emit when confronted with increas-
ing data error rates can be remarkable, with one player sounding
ne and another playing heavy static. This can be a problem not only
with dirty and damaged CDs, but with poor CD-R burns. You can
approximate an error test CD by using a ne point alcohol-based
marker, such as a Sharpie, and putting about a 1-mm dot on the
underside of the CD. CDs play from the inner diameter to the out-
side, so putting the dot about the 4 mm from the beginning of the
data area avoids the TOC and should affect the music a few min-
utes from the start. Some players will skip, and some may sound
ne. The dot cleans off easily with a issue moistened with alcohol.
Devices using perceptual coding such as MP3 devices and Mini-
Disc are carefully designed to disguise their performance limitations
when playing music. They should be considered a step down in
delity from the CD, and it is usually relatively easy to get poor

443
Appendix 3

sound from them. Be sure to compare the same data rates or disc
play times. Like most digital systems, it is usually in the softest
sounds where these systems show weakness.

By the time these tests are completed, the salesperson may be stunned into
silence, or possibly will have understood what you are doing and have helped
you through the process, thereby learning something as well.

444
Appendix 4
Good Acoustics in
Small Rooms and Auditoria

1.Small Room Acoustics

In this section, we will try to demystify the subject of the acoustical proper-
ties of a small room, and we will offer suggestions on how to optimize the
performance of the loudspeakers in your listening environment. Information
presented here will be useful to recording engineers, but should be of inter-
est to serious audiophiles as well.
We will study some of the acoustical principles involved and the behav-
ior of sounds of different frequencies in small rooms in order to get a better
understanding of what we hear. By a small room, we mean a room the size
of a typical recording studio control room or a home living room. (By this
denition, a shower stall is a very small room, albeit one with interesting
acoustical properties. We will return to the shower stall a little later.)
It is interesting that the acoustics of a small room are quite different from
the acoustics of a large room such as an auditorium. It is also quite well known
that a given loudspeaker will sound very different in a small room compared
to how it would sound in an auditorium.
We will look into the reasons for these differences, and will explain the
principles at work in such a way that you can understand them, even if you
do not happen to have a Ph.D. in acoustical engineering. With these general
facts in mind, you will be able to make valid decisions about the placement
of the speakers, the placement of the listeners, and the adjustment of the speak-
ers in order to achieve the most agreeable listening experience. We will try to
inform you what to do if your room seems to be impossible to congure for
good sound, even if, in the extreme case, only a stick of dynamite will serve
to correct the situation. Fortunately for all involved, this extreme case is sel-
dom encountered in practice.
It will be interesting to think about the philosophy of loudspeakers record-
ing engineers use in recording studio control rooms for evaluating microphone
types and placement and the monitoring of the nal mix of recordings. After
all, the recording engineer makes his living with his ears, and they must not
deceive him. (The serious but nonprofessional music listener will also benet
from this.)
There is a long history of the various types of speakers used in monitor-
ing recordings. In former times, very large theater speakers were often used,
the idea being that everything should be accentuated and reproduced bigger
than life so the slightest defect in the recording would be painfully obvious.

445
Appendix 4

The sound levels used in monitoring often came close to deafening the engi-
neer, making him insensitive to the subtleties in the reproduced sounds.
Today, we know better. We understand the effect of very high-intensity
sounds on human hearing, such as threshold shifts and auditory fatigue. We
believe the monitoring levels should not be much greater than the levels that
will be used in the nal reproduction in the users environment. We believe
the monitor speakers should be as neutral sounding as possible, that they
should have no discernable acoustic character or tone color. Simply put, they
should just convert the electrical signal fed to them into sound without adding
or subtracting anything. All the sounds that emanate from the monitor speak-
ers have their origin elsewhere in the recording chain. If the engineer hears
something he does not like, he knows that it is not a distortion or coloration
introduced by the speakers. He is therefore able to track it down and correct
it and then be able to hear the difference.
However, speaker systems are not used in a vacuum. They must be
located in rooms that will modify their sounds in complex ways, as we shall
see shortly. Therefore, we must nd the best way to see to it that the place-
ment in the room and the characteristics of the room itself do not destroy the
purity and linearity of the sounds the loudspeakers are capable of producing.
The typical small room is unfortunately not ideal for listening to live music,
and is also not ideal for listening to reproduced music. There are many rea-
sons for this, and most of them boil down to the fact that the wavelengths of
the sounds produced by music vary from about 40 feet for the lowest bass
notes to about 12 inch for the highest notes in the treble. The wavelengths of
the sounds of the fundamental pitches of a piano are shown in gure 1, below.
(The harmonics of these pitches extend to much higher frequencies and much
smaller wavelengths.) Large auditoriums that are designed for listening to
music can handle this huge range of sound wavelengths well, mostly because
even the longest waves are short compared to the dimensions of the room.
The large room swallows up these long waves, and allows them to bounce

Frequency Pitch Wavelength

27.5 Hz Bottom A 40 feet

55 Hz Low A 20 feet
110 Hz Tenor A 10 feet
220 Hz Middle A 5 feet
440 Hz A 440 2.5 feet
880 Hz 6th A 1.25 feet
1760 Hz 7th A 7.5 inches
3520 Hz 8th A 3.6 inches

Figure 1

446
 Appendix 4

around freely and to be diffused by the reections from the distant walls so
they can attain a proper balance with the higher pitched tones. The small room,
however, is not able to cope with these long sound waves in a manner that
satises the musical ear.
The typical small room does not have enough space to permit this diffu-
sion of the longer waves, and they bunch up, interfere with one another, and
generally raise havoc with the musical balance. This is bad enough when real
musicians with conventional instruments are performing in a small room, but
the situation is worse when you try to reproduce the music with loudspeak-
ers. This is because the speakers are the sources for all the musical sounds,
rather than having the music come from many different locations in the room.
Sound waves are actually pressure disturbances that travel at a nearly con-
stant speed (the speed of sound, about 1,100 feet per second) through the air.
They carry energy as they travel, or in other words they convey power from
the sound source to the sound receiver. The molecules of the air do not travel
from the source to the receiver but simply vibrate, transmitting energy from
one to another, somewhat analogous to a series of upright dominoes each one
of which will fall due to being pushed by its neighbor. Each domino does not
move very far, but the energy it took to topple the rst one is transmitted all
along the path of the domino wave. See gure 2.

Figure 2

When a source of sound is radiating into three-dimensional space with no

barriers present, the energy emitted by the source will ow outwards in all
directions in the form of compression waves in the air. The waves will
expand in spheres ever increasing in size, and this will go on ad innitum if
the waves do not encounter any barriers or objects. This is called a free-eld
condition, and the sound level you would sense at any location in the sound
eld would be dependent on the strength of the source and the distance
between you and the source. See gure 3.
Since the waves are expanding in area as they move out from the source,
each doubling of the distance results in a reduction of the sound energy by a
factor of four. (The area of a sphere is proportional to the square of its diam-
eter.) In more precise terms, the sound level is said to obey the inverse square
law. For instance, the sound power you would sense at 1 foot will be four
times the power at 2 feet, and nine times the power at 3 feet. A factor of four
in sound energy level is a 6-decibel (6 dB) change, so we say the sound level
falls off at 6 dB for each doubling of distance.

447
Appendix 4

Figure 3

True free-eld conditions are almost never met in practice since there are
always barriers and objects in the path of the sound waves, but under certain
conditions, there can exist a good approximation of a free eld. We will return
to this concept a little later.
When a sound wave encounters a large barrier, such as a wall, part of it
will be reected from the wall and the wall will absorb part of it. The amount
of the energy in the wave that is absorbed is dependent on how much the
wall moves in response to the pressure exerted by the wave. If the wall is very
solid, i.e., made of concrete or solid uranium a foot thick, the wall will not
move appreciably and almost all of the energy in the wave will be reected
back into the room. The amount of energy reected by the wall also depends
on the surface condition of the wall, and on the wavelength of the sound. If
the wall is covered with a soft material, such as egg cartons, berglass, cloth
tapestries, or old undershirts, sounds with short wavelengths will be absorbed
more efciently than sounds with long wavelengths. This is the reason that
most rooms will have a longer reverberation time at low frequencies than at
high frequencies. Of course most rooms are not made of uranium and they
do ex when they are struck by sound waves, and this absorbs energy, reduc-
ing the strength of the reected sound.
In large rooms where the smallest dimension is about 50 feet, the waves
from the sound sources expand in spheres as if in a free eld until they meet
the surfaces. Then they are reected in such a way that they continue to spread
out after they are reected. This allows the reected sounds to become ran-
domized on multiple reections, and the interference patterns are so com-
plex and overlapped that they are scarcely noticeable.
In a large room, the sounds of all wavelengths get reected many times
and in all directions as they meet the surfaces of the room. They nally reach
a condition called diffusion, which means the waves are moving in unpre-
dictable directions as they gradually die away from being partially absorbed

448
 Appendix 4

at each reection. In a diffused sound eld, you are not able to tell where a
sound is coming from. You are immersed in an ocean of uniform sound com-
ing from all directions at once. Complete diffusion is not possible in practice,
but in some very reverberant spaces, it is approached. For listening to music,
however, you do not want the sound eld to be too diffuse, since the musi-
cal line would be blurred, and the structure of the melody and harmony would
be difcult to discern at best. For this reason, concert hall designers strive to
achieve a pleasing balance between the diffuse reverberation in the room and
the direct sounds coming from the musical instruments.
The acoustics of large rooms can be wonderful for live music, but they often
present problems for amplied sound. The sound system designer must be
careful to see that the speakers direct the amplied sounds to the listeners
ears before they reect from the walls or other surfaces. Otherwise, the sound
heard by the listener would be too reverberant and muddy. Also, the micro-
phones used in sound reinforcement systems must be close to the perform-
ers so they dont pick up the reverberant sound of the room.
In a small room, the situation is different. The small room has reecting
surfaces that are close enough to the source of sound that there is not sufcient
space for the long-wavelength sound waves to diverge, and they strike the
surfaces almost as plane waves rather than spherical waves. Therefore the
long wavelengths of the lower frequencies prevent them from being able to
diffuse smoothly throughout the room. The main reason for this is the phe-
nomenon called interference. Interference is a condition where two sound
waves occupy the same space; their sound pressures add together when they
are in phase with each other and they subtract (partially cancel) when they
are out of phase with each other. This is called constructive and destructive
interference.
A special case of constructive and destructive interference is the dreaded
standing wave, sometimes called the stationary wave. A standing wave
occurs when two large parallel surfaces have a sound source between them
operating at certain specic frequencies. To form a standing wave, the sound
waves must be of a frequency where one half a wavelength is a submultiple
of the spacing between the surfaces. In other words, an integral number of 12
wavelengths of the sound must t between the two surfaces.
In the example shown in gure 4, there are ve 12 wavelengths between
the wall surfaces. As the sound energy bounces back and forth between the
walls, the sound pressures successively add together and cancel, causing a
series of locations of maximum loudness and minimum loudness to exist. (This
example shows complete cancellation and zero sound pressure at the min-
ima, but this never happens in practice.) Note that there is always a pressure
maximum at the wall surfaces. The locations of maximum sound pressure
are called antinodes, and the locations of minimum sound pressure are called
nodes. Of course, the locations of the nodes and antinodes depend on the
wavelengths involved and the spacing between the walls. The two surfaces
will generate standing waves at every frequency at which the wall spacing
is divisible by 12 of the wavelength. Each of these standing wave patterns is

449
Appendix 4

Figure 4

known as a room mode or simply a mode, not to be confused with a node.

A room mode is similar in many ways to a mode of vibration of a mechani-
cal structure, such as a violin string. It is seen that there are theoretically an
innite number of possible modes between two surfaces; however the wave-
lengths of sounds above about 500 Hz are short enough that they form spher-
ical waves that tend to become diffused very quickly.
A standing wave pattern like the one shown above obviously creates a very
nonuniform sound level as you move around between the surfaces. For
instance, if you are standing at the location of a pressure minimum, the sound
will be very soft, even inaudible, but when you are near a pressure maximum,
the loudness can be very great. Needless to say, this is not a desirable state of
affairs in which to listen to music!
It gets worse. The problem is that in any room that has parallel walls, oor,
and ceiling, there will be a whole slew of standing wave patterns that exist
between the two front-rear walls, the two side walls, and the oor and ceil-
ing. The worst possible case would be a room that is shaped like a cube, with
all the linear dimensions the same. This kind of a room would have three sets
of equal standing wave frequencies, which would add together to produce
extremely nonuniform distribution of sound pressure in the room. There are
also other families of standing waves that occur in rectangular rooms. One such
is caused by reections back and forth between the diagonal corners of the
room. These diagonal modes are the lowest frequency modes possible in a room,
since the distance between the diagonal corners is greater than any wall spacing.
For an illustration of a prominent standing wave phenomenon, consider
a shower stall for a moment. It is generally about 7 feet high, and has hard,
reective walls. The lowest frequency standing wave that such an enclosure
can produce is at that frequency where 7 feet is 12 wavelength. This turns out
to be about 157 Hz, which is not far from the fundamental frequency of the
average male speaking voice. The voice thus excites this room mode, and the

450
 Appendix 4

persons voice and ears are near the upper end of the enclosure, which is close
to the upper antinode, so the room mode is heard as a strong reinforcement
of the voices fundamental tone. This is no doubt the reason that men, espe-
cially basses, often are pleased by their own performance of singing in the
shower!
Since parallel surfaces are a major source of standing waves, designers have
learned that it is not a good idea to design music rooms with parallel walls,
oors, and ceilings. This is not as easy as it sounds, because you still have to
have corners (unless the room is round), and they will generate their own
standing wave pattern. Not only that, but the walls have to be very signi-
cantly out of parallel to prevent standing waves to build up between them.
The walls would have to be at fairly large angles to appear nonparallel to a
sound wave with a 10- or 20-foot wavelength. Nonparallel walls do prevent
standing waves from being much of a problem at higher frequencies, i.e., at
frequencies above 200 Hz or so.
Another thing that has been tried is to treat the wall surfaces with suitable
materials that absorb the sound rather than reect it. This is also easier said
than done, for long wavelengths are notoriously difcult to absorb. It takes
very thick layers of absorptive material (about 14 of the wavelength you want
to absorb) in order to be effective. This is not only unsightly and expensive,
but it reduces the useful size of the room. Another way to reduce the strength
of the room modes is to make the walls exible rather than massive. A room
lined with 12-inch sheet rock on 2x4 studs spaced at 16 inches is quite exi-
ble at low frequencies, and it absorbs these frequencies much better than a
solid wall of brick or concrete. However, thin exible walls will also let more
sound pass through them, possibly waking up Aunt Minnie, or worse.

Resonant Absorbers

There do exist so-called resonant absorbers (sometimes called bass traps)

of fairly small size that will absorb energy at specic frequencies. They can
take the form of rectangular Helmholtz resonators, which are resonant cavi-
ties than can be built behind the walls with holes in the walls connecting to
them. Another form of resonant absorber looks like a cylinder several feet
long with a hole in the end about a foot in diameter. These devices have been
used to help in reducing the effects of troublesome low-frequency standing
waves. Again, they are expensive, unsightly, and they have to be tuned fairly
accurately to the frequency being absorbed. See gure 5.
Instead of, or in addition to, placing sound absorbent material on the walls
of the room, it is a good idea to cover them with fairly large irregular sur-
faces. There are many different forms that diffusing surfaces can take, and to
describe them all is beyond the scope of this appendix. In general, the irreg-
ularities in diffusers must be of sizes comparable to the wavelengths they are
to diffuse. This means they must be of varying size in order to attain diffu-
sion over a frequency range from 40 Hz to over 500 Hz or so.

451
Appendix 4

Figure 5

One thing that is commonly overlooked in the design of music listening

rooms is the need for good diffusion of sounds reected from the ceiling. Dif-
fused ceiling reections are very important if you desire good stereo imag-
ing. This is because a specular reection from the ceiling produces an audible
image above the speaker, and the sound from the image interferes with the
direct sound from the speaker. Figure 6-A shows the effect of a at ceiling,
and gure 6-B represents a diffusing ceiling. Note that here the listener is not
hit with the rst reection
Interestingly enough, specular reections from the walls are much less annoy-
ing than the ones from the ceiling and provide a certain amount of widening
of the stereo image. This does not sound unnatural, while widening of the image
in a vertical direction does not correlate well with our everyday experience.
All in all, experience over many years has taught us that no matter how
carefully a room is designed, it will still contain a collection of room modes
that will produce nonuniformity in its distribution of sound energy. It is hope-
less to try to design a space that will have no room modes. We must live with
this fact, and we must seek other ways to reduce the effect of the troublesome
room modes on the quality of the musical sound that we experience in a room.
That is the subject we will consider in the next section.
Before you attempt to set up a loudspeaker system in a room, you should
rst determine what types of problems the room has, and how severe the prob-
lems are. Only then can you meaningfully attack the situation and realize a
satisfactory result.
As we have seen, there is no such thing as a room without room modes,
so forget trying to nd one. But there are huge variations in the frequency
and amplitude distribution of the modes that exist in different rooms. As a

452
 Appendix 4

Figure 6-A

Figure 6-B

general guideline, you do not want different modes to produce nodes and
antinodes that are in the same location in the room at the same frequencies.
These are called overlapping modes. Therefore, the room should not have
dimensions that are simple multiples of each other. For instance, the room
height and width should not be multiples of each other, or multiples of any
common factors. Ideally, the ratios of height, length, and width should be irra-
tional, like the square root of three (1.73 . . . ), or the Golden Mean.1 It is impor-

1. The Golden Mean, also called the golden ratio, is a relationship discovered by the
ancients that occurs frequently in nature, and has a numerical value of about 0.618. It was
intensely studied by many people involved in number theory and artistic endeavors. It has
been said that a rectangle with length and width proportional to the golden mean is more
pleasing to the eye than any other rectangular shape. An ordinary 3 x 5-inch card conforms
very closely to the golden mean.

453
Appendix 4

tant that the ceiling height of the listening room not be too low, as discussed
in the section on stereo imaging.
If at all possible, the walls should not be simple plane surfaces, but should
be broken up with furniture, windows, tapestries, paintings, etc. to reduce
the efciency of the standing wave patterns. The walls should not be paral-
lel if possible for the same reason, and they should not be extremely rigid,
such as concrete or brick. A good compromise construction is 12-inch gyp-
sum wallboard (sheet rock) on 2x4 wooden studs. The ceiling should not be
parallel to the oor, or else should be deeply coffered or have irregular sur-
faces on it.
Carpeted oors are in general good for listening rooms because they reduce
reverberation and also tend to break up oor-to-ceiling standing waves.
The room where you will install your monitor speakers will have a series
of room modes that will be more or less severe depending on the shape and
construction of the room. Placing a loudspeaker in the room will not affect
the pattern of room modes, and moving the loudspeaker around to different
locations in the room will not change the room modes either. However, dif-
ferent locations of the speaker will affect how strongly the speaker will excite
the various modes. This means the room will affect the perceived frequency
response of the loudspeakers, as discussed next.
While frequency response problems may not generally be considered dis-
tortion in the strict sense of the word, the frequency response produced by
a sound system is extremely important in the perceived tone quality of the
reproduced sound. Poor frequency response is a linear type of distortion rather
than a nonlinear distortion, but it still is distortion and is annoying like any
other type of distortion.
When different frequency components of a reproduced sound are produced
at sound pressure levels different from the levels that existed in the record-
ing venue, the result is frequency response distortion. If all frequencies are
reproduced at their original relative levels, the frequency response of the sys-
tem is said to be at. Irregularities in frequency response cause the tone
color, or timbre, of the sound to be different from the original sound, and this
is probably the most common and noticeable problem heard in music repro-
ducing systems.
The frequency response of a music system is a curve that plots frequency
versus the sound pressure level, or response, produced by the system. The
response is normally scaled in dB, and the frequency is normally plotted on
a logarithmic scale, where octave bands all span the same distance on the hor-
izontal axis. The ideal frequency response curve is a straight horizontal line,
indicating no accentuation of sound pressure level at any frequency, nor any
attenuation at any frequency. This is why it is called a at response.
It is possible to have a great many different types of variations in frequency
response, and they have quite different effects on the perceived sound. Prob-
ably the most disagreeable type of irregular frequency response is the exis-
tence of peaks. High frequency peaks sound harsh and bright, while
low-frequency peaks will sound boomy or honky, depending on the fre-

454
 Appendix 4

quency where the peaks occur. Your loudspeakers are generally designed to
produce the most at and uniform frequency response possible when radi-
ating sound into a free eld, i.e., into a eld where there are no standing waves.
However, the monitors are at a loss to correct for frequency response anom-
alies contributed by the room.
For instance, if there were a prominent room mode between the front and
rear walls at 50 Hz, this mode would result in a maximum in sound pressure
at both wall surfaces when a 50 Hz tone was sounded in the room. It will also
produce a minimum in sound pressure at 14 wavelength away from the walls.
Now, if you place the speaker against the front wall where the sound pres-
sure from that mode is a maximum, the speaker will excite that mode to its
maximum degree. Only a small amount of acoustic energy at 50 Hz will pro-
duce very strong sound levels at all the antinodes. If, on the other hand, you
move the speaker out from the wall and place it in the location of a node (min-
imum) of the 50 Hz mode, that mode will not be excited to any great degree.
It is as if the room will not accept any energy at this frequency at that partic-
ular location. Of course, there may be other modes near 50 Hz that exist
between the side walls or the oor and ceiling that would not have minima
at that same location, and these modes would be excited by the speaker, pro-
ducing signicant sound pressure at that point.
The whole process of determining the best location for the speakers and
the listeners in a room is simply to nd the location where the speakers are
placed so they do not excessively accentuate any room modes that are in the
musical range. Also, the listeners must not be located at minima of these
modes. In other words, the most prominent room modes must not be excited
to a large degree, and the other modes should be excited to the same mod-
erate degree.
If the room is quite reverberant (live), the listener should be fairly close to
the speakers for increased clarity of the sound. For critical monitoring of music
sources in most typical medium-sized rooms, a good speaker to listener dis-
tance would be somewhere around 4 to 6 feet. For dead rooms this could be
a little farther, and for very live rooms, a little closer.
It might seem intuitive to most people that the loudspeakers in a stereo
listening set up should be symmetrically disposed with respect to the center-
line of the room. This, however, is frequently not the case. If you are listen-
ing to stereo in a small room, and if the room is symmetrical and the speakers
equidistant from the side walls your two ears essentially hear similar sounds.
A sound that comes from both speakers equally, i.e., from an instrument cen-
tered in the stereo eld, will be reected from the sidewalls, and the reections
will arrive at your ears with the same delay time. It has been shown that under
these conditions, the stereo balance is less pleasant to the listener than if there
is some asymmetry in the acoustic eld. This increases the complexity and
subjective spaciousness of the sound eld. So, dont worry about exact sym-
metry in the setup. You may be happier being a little to the left or right of
center. (Of course, for proper stereo imaging you must be located at the same
distance from each speaker.)

455
Appendix 4

The proper location of a speaker in a room involves two essentially dif-

ferent tasks; optimizing the low-frequency performance and optimizing the
high-frequency performance. The low-frequency performance is almost exclu-
sively determined by the interaction between the speaker and the room modes,
while the high-frequency performance is essentially immune to the mode
structure of the room. High-frequency characteristics are affected by reect-
ing surfaces near the speakers and by the relative liveness of the room at the
higher frequencies.
Low-frequency behavior is the most difcult to control, so we will tackle
it rst. One fairly easy, but somewhat tedious, way to minimize the effect of
the room modes on the response of the speakers in a room uses an omnidi-
rectional microphone of good quality.2 The microphone must have smooth fre-
quency response at low frequencies, and it can be connected to each monitor
(one at a time) through a recording console or tape recorder set to monitor mode.
The microphone should be placed very near a corner of the room, and fairly
close to the oor or ceiling. This will assure that the microphone will respond
to all the room modes with equal sensitivity.3 A convenient way to mount it
is to simply tape it to the wall. Do not change the microphone location dur-
ing these tests; you want to have as few variables as possible. See gure 7.
Start with the volume at minimum with the speaker located in the room
at the height at which it will be used. This is usually ear height. Then very
slowly bring up the volume control and listen carefully to the sound just as
the system begins to go into feedback. Do not turn up the volume to produce
a roar or shriek; just listen to the rst onset of a feedback tone. You will nd
the feedback starts at a specic frequency and this will usually be the frequency
of one of the most prominent of the room modes.
Then, with the volume turned down, move the speaker to another posi-
tion and repeat the process. Listen to nd out if the same frequency tone is
activated. Keep repeating this until the onset of feedback occurs at several
different frequencies at the same time. If, in repeating this process, the feed-
back always occurs at high frequencies, roll off the treble response of the
speaker, and/or roll off the microphone sensitivity at high frequencies with
an equalizer.
You will nd that in some cases, moving the speaker only a few inches
will make a big difference in the tones that are activated at feedback. If at all
possible, place the speaker where it excites the maximum number of differ-
ent frequencies at the same time. Sometimes in good rooms this will turn out

2. If you do not have an omnidirectional microphone, a suitable one can be purchased

from most audio electronic parts houses, such as Radio Shack, for a few dollars. The smaller
the physical size of the microphone the better. You are only concerned about its low-frequency
response, so you do not have to have an expensive microphone.
3. If the microphone is placed away from the walls, its response will depend on its loca-
tion relative to the room modes. If it is near the corner, it will be at a maximum (antinode)
for all the room modes.

456
 Appendix 4

Figure 7

to be as many as four or ve different tones. When you have achieved this

situation, the speaker is exciting several room modes equally, with none stand-
ing out over the others. This will be close to the best you can do in that room,
and you should bolt the speaker in place. The process must be repeated with
the other speaker, and hopefully, its ideal location will be such that the two
of them make a good stereo pair.
When you discover this magic location, you will have located the speaker
so that it excites the most prominent of the room modes to a nearly equal
degree. You will nd that music heard over the speakers will sound as neu-
tral as is possible in that particular room. Listen critically to the speakers with
several different types of music. The room modes will still be there, and if
you walk around the room, you will hear different degrees of low frequency
response, but the variations will be minimized.
Incidentally, this same process can be used to nd the location for a micro-
phone in a recording studio to minimize the effects of the room modes. Place
the speaker in the corner to excite all the modes, then move the microphone
around for the greatest number of tones at feedback. You will discover that
a recording made with this microphone location will sound more balanced
and neutral than if you had placed the microphone at random. Of course, you
will have to position the instruments to be recorded to suit the microphone
rather than vice versa, but this is generally a small price to pay.
The control of high frequencies in a room is usually just a matter of equal-
ization to establish the correct balance between the mid-range and the treble.
The high-frequency balance must be adjusted for the listening position. It will

457
Appendix 4

be found that reduced high-frequency levels will be required for closer lis-
tening than for more distant listening. This is because most rooms have a
greater amount of high-frequency absorption than low-frequency absorption.
The level of the highs will also be affected by the presence of reecting
surfaces near the speakers. In general, there should be no hard at surfaces
near the speakers (within several feet or so) to prevent sounds reected from
them from reaching the ear. This not only would change the balance of the
high-frequency energy, but also would compromise the stereo imaging pro-
duced by the speakers.
In recording studios, it is not uncommon to nd that high-frequency
reections off the mixing console adversely affect the tonal balance and stereo
imaging heard by the mixing engineer. To avoid, or at least minimize this effect,
the speakers should be placed so the working surface of the console is at least
a few feet from the tweeter, and that the console work surface does not pro-
duce a reected image of the speaker that can be heard by the listener. The
angle of reection of the sound is equal to the angle of incidence. See gure
8 for more detail.

Figure 8

It has been known for a long time that precise imaging of the location of
the musicians in a two-channel stereo sound system is difcult at best and
even impossible under some conditions. For good imaging, the phase and
amplitude vs. frequency characteristic of each channel must be very closely
matched, and this includes the entire recording chain from microphones
through recording consoles and speakers. At the playback end of this chain,
the loudspeakers must be matched in frequency response, sensitivity, and
phase. These characteristics are accurately controlled in the high-quality stu-
dio monitors, so you shouldnt worry about that.
Stereo imaging is almost totally controlled by high-frequency cues that the
ear interprets as source location information. The high-frequency amplitude
and phase response must, at all costs, be identical in the two speakers, and
the acoustic paths between the speakers and the ears must also have identi-

458
 Appendix 4

cal characteristics. To attain acceptable imaging with two-channel stereo, you

must be very careful to eliminate short time-delayed reected sounds from
reaching your ears from the loudspeakers. This means the speakers should
be placed at least ve feet or so from any reecting surfaces such as walls or
large objects. If this is unavoidable, then the surfaces or objects must be irreg-
ular enough that the reected sound is diffuse. A smooth at surface acts like
an acoustical mirror and sound reected from such a surface is called a spec-
ular reection. A specular reection will create an image of the sound source
(the speaker) that appears to be behind the offending surface, and this image
behaves like a delayed source of the sounds the speaker produces. This added
delayed sound would confuse the ear and lead to poor imaging, as well as
poor frequency response due to interference. See gure 8 for an example of
bad reections from at surfaces.
For optimum performance, the vertical and horizontal angles the moni-
tors make with respect to the listeners ears need to be considered. If the mon-
itors are to be placed very much above ear level (a practice that is not rec-
ommended for best performance), then the cabinets should be tipped toward
the listener at the top so the line of sight to the center of the speakers is at a
vertical right angle to the faces of the cabinets. See gure 9.
The horizontal dispersion of sounds of different frequencies is very well
controlled by the speakers, and while the speakers should each be toed in
to face directly to the listener, a horizontal angular error of up to 10 degrees
should not be audible under normal operation.
As was mentioned earlier, the ceiling can be a troublesome source of spec-
ular reections. If the ceiling above the monitors is closer than about 5 or 6
feet from the speakers, then it should have a convoluted shape to avoid spec-
ular reections. Alternatively, the ceiling can be treated with acoustical
absorbent material, but adequate absorption is difcult to achieve at long
wavelengths.

Figure 9

459
Appendix 4

2. Auditorium Acoustics

Among the denitions presented in this book are several references to good
acoustics. Although many people seem to be able to tell good acoustics from
bad acoustics, it is quite another matter to adequately describe what actu-
ally constitutes acceptable acoustics, especially in any quantitative way. It is
always risky to try to relate subjective impressions to objectively measured
data, as has been abundantly proved by some acousticians who thought they
designed the perfect concert hall only to have it condemned wholesale by
the public and critics.
Charles Garnier, the architect for the Paris Opera, wrote a book in 1871 on
theater design titled Le Theatre. Here are some excerpts from the chapter on
acoustics:

For two years at least, I kept this question of the elements of acoustics
before me. I read all the books dealing with it; I looked over almost all
the theaters built, hoping to discover for myself some clear rules which
the literature failed to reveal. In short, I left no stone unturned to nd
a way, a means, a bit of factual, helpful information. . . .
But what can we conclude? Halls almost identical in shape, arrange-
ment, and dimensions vary a great deal. This one is good; that one is
bad. One hall quivers at the slightest sound from a stringed instrument.
This is an excitable, nervous hall. Another doesnt even vibrate under
the inuence of the entire orchestra. That is a lifeless, soulless hall. Then,
an echo may be heard during the day which disappears in the evening,
or one is heard in the evening but disappears during the day when the
hall is empty. And the notes one takes on these observations, the com-
parisons, the thorough researchall these data tend to merge into an
endless mass, an inextricable maze where the end of the string cannot
even be located.
Dissatised, one then inquires of others. Buildings are examined and
other works are consulted with the hope of at least nding someone to
guide, but none turns up. The terrain is unknown, the guides do not
know the way, and each one goes on blindly in his own way. This one
wants a long hall; that one a wide hall. This one a low ceiling; and some-
one else a lofty one. . . . The end result is simply confusion with noth-
ing converging toward the adoption of any xed rules. . . .
After sound quality, the matter of resonance again emerges to com-
plicate matters. The walls of the hall for some writers must be con-
structed of stone and have rigid partition walls. Following the direction
of others, it should be made of wood or of a light elastic material. One
German wants a dome of masonry; another German might want it
padded with wool.
And Lord knows what else! In the nal analysis everybody wants
something, but nobody agrees on the choice. Reconciling all these par-

460
 Appendix 4

ticulars is impossible. The science of theater acoustics is still in its infancy

and its outcome uncertain.

In view of the spotty track record of certain modern halls, it seems that Gar-
niers tristesse might be echoed by contemporary theater architects and that
his conclusions are basically as valid today as they were in 1871.
But all this notwithstanding, let us at least take a cursory look at some of
the acoustical measurements we can perform and try to correlate them with
our aesthetic senses.
Reverberation time is probably the most discussed acoustical parameter
and, as we shall see, is also probably the most misunderstood. The concept
of reverberation time was developed by Wallace Clement Sabine, of Harvard
University, around 1900. He called it the duration of audibility of residual
sound. Reverberation is easily heard and its concept is easy to grasp. Sabine
quantized the measurement and prediction of reverberation time by dening
it as the time in seconds it takes for a sound to decay in level by 60 decibels
after the source of the sound is stopped. He also developed a formula for pre-
dicting what this value will be based on: the volume of the room, the surface
area of all the boundaries, and the amount of sound absorption these bound-
aries provide. It should be clear that for most materials the absorption of
sounds of different frequencies will not be the same, leading to a reverbera-
tion time that varies with frequency.
It is generally recognized that long reverberation times cause musical
phrases to overlap and become indistinct; in other words, clarity, or intelli-
gibility, suffers. On the other hand, short reverberation times allow excellent
intelligibility and very little blurring of musical passages while at the same
time sounding dry and lifeless. Thus one might think that it would be easy
to dene an optimum reverberation time that would be acceptable for all
rooms. This, however, is a gross oversimplication.
Consider the subjective impression of being in a very large reverberant
space, such as Chartres cathedral in France, which has a reverberation time
at midfrequencies which exceeds 10 seconds. Two people engaged in con-
versation in such a building are scarcely aware of any reverberation at all.
Only if they suddenly raise their voices to a shout or loudly clap their hands
do they consciously hear the sound die away after what seems a very long
time. Intelligibility between them is excellent so long as they are relatively
close together. This is because the reverberation, although very long-lasting
at 10 seconds, is very low in level compared with the direct sounds of the
voices, and the individual sound reections that make up the reverberation
are quite late compared with the direct sound because of the large distances
to the walls and ceiling.
Now, consider a small room, say 20 feet square, which has very hard walls
and a 10-second reverberation time. The reverberation versus frequency curve
of this room could be made to be the same as the large cathedral if desired.
Would the two spaces sound the same? Decidedly not! The small room would

461
Appendix 4

have nearly zero intelligibility because the reverberation would start out at
almost the same level as the direct sound, and the reections would be much
more densely spaced in time. The result would be extensive masking of the
direct sound.
From this we see that a large space can have much more reverberation than
a small space for the same perceived acoustical clarity. In fact the apparent
reverberation time will depend on the level of the reverberant sound imme-
diately after the arrival of the direct sound, and it decreases with room size
for a given reverberation time.
The variation in decay time for different frequencies is also very impor-
tant. Most intelligibility is transmitted by the high-frequency content of a
sound, the lower frequencies contributing to the timbre, or tone color, of the
sound. Therefore, one can tolerate much longer reverberation time at low
frequencies than at high. Real rooms naturally provide this, for high frequen-
cies are much more readily absorbed than low, especially in large rooms. If
a room is constructed so it does not have this reduction of decay time as
frequency rises, it will denitely sound strange and will be shunned by
musicians.
Another acoustical parameter related to reverberation is diffusion, which
is a measure of the randomness of the directions in which reverberant energy
is moving. In a perfectly diffuse space, the sound at any point is equally likely
to come from any direction; in other words, sound is arriving from all direc-
tions at once, and a listener cannot localize a sound source. (The degree of
diffusion is independent of the reverberation time and cannot be inferred from
reverberation measurements alone.) The opposite extreme is no diffusion at
all, which is the condition in a free eld, such as out of doors or in an ane-
choic chamber. In a free eld, all the energy is moving in the same direction,
and source localization is easiest. A free sound eld would be described by
a listener as completely dead and lifeless.
How much sound diffusion is desirable in a room? This is a very good
question. The Hungarian physiologist and acoustician Georg Von Bksy has
shown that the human hearing mechanism is somewhat directional and that
sounds arriving from the frontal angles of less than 60 degrees from straight
ahead attract much more attention from us than sounds coming from the sides
and rear, and we are much more able to ignore sounds coming from the sides
and rear. The Japanese acoustician Yoichi Ando has conducted experiments
to show that listeners prefer reected sounds to come from within this plus
and minus 60-degree angle. In other words, our preference is for reected
sounds to come from the direction in which we pay the most attention. Ando
also found that our preference is for directions that maximize the difference
in the sounds heard by the two ears. In other words we prefer to listen to
sounds that have a maximum of stereophonic information. The objective mea-
sure of this difference in sounds is the so-called interaural cross-correlation
(IACC), which can be found by using a dummy binaural head and suitably
analyzing the signals from the two microphones.
From all this, it can be determined that in a music-listening room, the rever-

462
 Appendix 4

beration decay curve must be steep enough and the reverberant level must
be low enough so as not to obscure the clarity or intelligibility of the music.
Also, and even more important, the reections that make up the early rever-
beration should come from fairly wide horizontal angles, preferably about
60 degrees off axis, and there should not be strong reections from the ceil-
ing. Ceiling reections result in increased IACC and are not preferred. This
means the reection from the ceiling should be diffuse, and reections from
the sidewalls should be discrete and should be aimed laterally across the room.
A room meeting these criteria is said to have a large degree of envelopment.
An excess of diffusion of the early sound is undesirable because it reduces
our ability to localize sources. For this reason, it is not desirable to place
strongly diffusing surfaces near the performers in an auditorium.
It is important that the early reections reaching the ear contain the full
frequency range, which means reecting surfaces must be quite large. Other-
wise only high frequencies will be reected and the tone quality will be per-
ceived as thin. Some music auditoriums have been designed with small
reecting surfaces spaced out from the walls or hung some distance from the
ceiling, and they distort the tonal balance by their frequency-selective
reections.
In a music room, it is important to avoid echoes, which are discrete strong
reections occurring at times later than about 60 milliseconds or so after the
direct sound reaches the listener. Echoes can be difcult to avoid in large rooms.
A similar problem, but one that occurs more often in small rooms, is that of
standing waves, which are the result of sound bouncing back and forth
between two parallel surfaces. Standing waves accentuate certain frequen-
cies and distort the tonal balance of music as well as alter the uniformity of
decay of the reverberant energy in a room.
Most of these comments apply to rooms for listening to recorded music
as well as to rooms for live music, although such rooms will be relatively small
and consequently less reverberant.

463
Appendix 5
Some Notes on Audio Measurements

In order to improve and rene the performance of audio equipment, the engi-
neer must undertake considerable research and development and, in so
doing, has a vast array of measuring instruments available. This instrumen-
tation for audio measurements has become very sophisticated in recent
years, drawing on the discoveries and techniques of many scientic disciplines.
In a way, it is surprising that the 1920s through 1940s brought forth so many
audio inventions when one considers how primitive most of the measuring
techniques and instruments were. For instance, measurements of distortion
and frequency response were difcult and time-consuming and were of neces-
sity of limited range. But this preWorld War II era saw the invention of all
the microphone types in use today, as well as the exponential horn with com-
pression driver, the introduction of negative feedback in ampliers, and the
development of stereophonic recording and reproduction. Most of the impe-
tus for all this invention came from the motion picture industry. In Alan Blum-
leins famous 1933 patent disclosure, where he puts forth the ideas of intensity
stereo and the stereophonic record, he never mentions commercial products
for home use, but discusses in detail the applications of his invention to motion
picture sound. The same may be said of the many inventions by the Bell Tele-
phone Laboratories of the period.
In this section, we would like to review the measurement techniques in
use today in an attempt to demythologize them, if our editor will let us get
away with this word. But rst, a word about the philosophy of audio testing:
Our friend Henning Mller likes to make an analogy between listening to a
sound system and looking at an attractive work of art such as a statue. When
listening to reproduced music, the mind integrates the sensory input from
the ears into a global experience; the sum of the parts is appreciated at once
without reference to the various local parts such as frequency response,
stereo imaging, distortion, etc. Similarly, when rst looking at the statue, it
takes only a few seconds to sense its overall beautywe do not consciously
(at least at rst) evaluate its individual features and textures in order to reach
the conclusion.
Now, suppose we wish to describe the beautiful statue to someone
objectively, and do so in such a way that the person could sense the global
effect we experienced on seeing it. The key word is objectively, for it is
easy to fall into subjective descriptions of local details, but they would not
help if the person were trying to build up analytically a total image of the
work of art.
The audio designers plight is similar, for he must assemble his transis-
tors, voice coils, styli, capacitors, etc., to optimize their performance in repro-

464
 Appendix 5

ducing music in an attempt to produce a subjective global experience for the

listener. To do this, he must make measurements and tests, always trying to
see that they are measuring and testing the correct local parametersthe ones
that make a difference. The making of a recording of a musical performance
is something like taking a photograph. A picture of our statue will convey a
large amount of its character to the viewer, even if the color is off, the lens not
sharp, or the perspective incorrect. The viewer is able to see through the
photograph and imagine what its true appearance is. The same is true of the
listener to the recording. He is able to listen through the medium and to
build in his imagination a replica of the performance. Incidentally, this abil-
ity of listeners to overlook the medium is a great stroke of luck for audio
designers, for there is often precious little in common between the original
performance and the reproduction of it! It behooves the audio person to train
himself to listen to the medium rather than to the music.
For many years there has been an antipathy between certain audio critics
and audio designers. Many times an equipment reviewer believes the audio
device designer is measuring the wrong parameters, for even though there
may be minuscule distortion and near zero noise, etc., the device may sound
very bad when reproducing music. The designer, on the other hand, is sure
the reviewer is crazy because the measurements prove the component is
nearly perfect.
For instance, harmonic distortion is a local parameter that can be reduced
to almost unmeasurable amounts. If this is done by the indiscriminate use of
negative feedback, the result can be the introduction of large quantities of tran-
sient intermodulation distortion, which does not show up in steady-state dis-
tortion measurements at all. The device, however, will sound bad with music.
On the other hand, analog tape recorders typically produce several percent
of harmonic distortion when signal levels are relatively high. But the tape
recorder does not sound bad from a musical standpointthe distortion it pro-
duces is more musical in character.
This is only one example of measurements being at odds with what our
ears tell us. The ear is after all the nal arbiter, and when properly trained, it
can be a very sensitive instrument indeed. It is interesting that the ear is at
once extremely forgiving and tolerant of gross distortions (witness the pop-
ularity of the miniature transistor radio) and also is the most sophisticated
measurement instrument available when properly educated and experi-
enced. The noncritical ear nances the audio industry, and the critical ear keeps
the designers forever busy and frustrated.
Now let us dene some relevant measurements and then look at some tech-
niques for performing them. The mechanical engineer has a certain advan-
tage over his electrical colleague in that he can see and feel the various parts
of a machine and easily visualize the way in which they interact. The elec-
tronic designer cannot directly sense the workings of his circuit, that is, he is
unable to see the charges moving around being resisted and impeded. There-
fore, he must use instruments to transform the operating parameters of the
circuit into a form he can sense. This necessary use of instruments places the

465
Appendix 5

observer another level of abstraction away from the actual device under obser-
vation, and this contributes to the sense of mystery that the uninitiated feel.
A true understanding of instrumentation and what it can and cannot do is
probably the most important body of knowledge the electronics person can
have.
The most fundamental of audio measurements is frequency response, or
more accurately, amplitude response versus frequency. Electronic device fre-
quency response is measured in a straightforward manner to be sure that all
frequencies are amplied by the same amount. The frequency response of
transducers is quite another matter, for here the environment in which the
measurements are made inuences the result, requiring much more sophis-
ticated techniques.
The next important measurement to be made is phase response, for fre-
quency response alone does not tell us much about the handling of transients.
Frequency and phase response together allow us to predict the transient
response of any device or system.
The third class of measurements is of distortions and noise, of which there
are a great many. Some are far more detrimental to reproduced sound than
others, and some are very difcult to measure.
The fourth class of measurements is of those applicable to the evaluation
of digital audio systems and devices. Digital systems produce different types
of distortions than analog systems do, and their measurement can be tricky.
Many audio measurements use a sine wave as a test signal, but the sine
wave does not occur naturally in music, and it does not resemble music in
any statistical or spectral way. Statistically, the sine wave spends much more
time at the upper and lower extremes of its voltage swing than does music,
which spends most of the time nearer to zero voltage. The sine wave thus does
not exercise the dynamic range of an audio device the same way a musical
signal does. Spectrally, the sine wave consists of a single frequency with zero
bandwidth. Music is always a broadband signal, consisting of many frequency
components at once. Moreover, it is almost always characterized by a con-
tinuous spectrum, where energy is distributed over a frequency band, rather
than by a discrete spectrum, where energy is concentrated at specic fre-
quencies. Pink noise has a much closer resemblance to music from a techni-
cal standpoint (and maybe to some music from the aesthetic standpoint also)
than does a sine wave, and therefore it is in some cases a more appropriate
test signal for audio devices.

Electrical Frequency Response and Signal-to-Noise Ratio

A straightforward method for testing frequency response is to apply a con-

stant amplitude sine-wave signal to the input of a device, measure the out-
put of the device with a voltmeter or a recorder that plots the amplitude on
a chart, and then vary the input frequency over the applicable range. The audi-

466
 Appendix 5

ble range may be from 20 hertz to 20 kHz, but response curves are usually
made over a much wider frequency range, sometimes up to 200 kHz or so.
This is because the out-of-band response affects other parameters such as
distortion.
Figure 1 shows a typical classical frequency response measurement setup.
Measurements of this type are time-consuming to perform because the fre-
quency must be slowly swept over a wide range for each measurement. The
measurements are repeated at various levels to evaluate the amplitude lin-
earity of the device under test (DUT).

Figure 1

Another way to do a similar test is to input a random noise signal (pink

noise or white noise) to the DUT and examine the output signal with a spec-
trum analyzer. Random noise is a combination of all frequencies at once, and
it exercises the entire bandwidth of the device at the same time, unlike sine-
wave testing, which presents only a single frequency at a time. Random noise
is a better simulation of music than a sine wave for this reason.
The spectrum analyzer can be an FFT analyzer, which can provide
extremely high resolution over the frequency range of interest. It effectively
divides the audio band into 400 or 800 spectral lines, giving amplitude
information every 50 or 25 Hz. It is important in this type of test that the
input spectrum have uniform output level at all frequencies, and this is not
particularly easy to achieve. In any case, the test signal being random, mean-
ing its long-term average is at with frequency and amplitude, requires that

467
Appendix 5

the output spectrum be averaged over a time period. This allows the uc-
tuations in the signal to be smoothed out. The FFT analyzer does this time
averaging.
A newer faster method for doing the same test is to use a dual-channel
FFT analyzer (gure 2). This unit looks at the input and output signals at the
same time and calculates the frequency response by dividing the output spec-
trum by the input spectrum. This allows much shorter averaging times to
achieve accurate results. An added advantage is that this analyzer also cal-
culates the phase angle of the output versus the input at all frequencies and
displays a phase versus frequency plot. Also, the noise generator is built into
the analyzer, making the setup quite simple.

Figure 2

Signal-to-noise ratio is also measured with the setup of gure 1. First, the
output level is measured with the voltmeter at a mid-frequency, and the input
level is adjusted until the output begins to show an overload. This is the max-
imum output level, and it is noted. Then, the input signal is removed and the
residual noise at the output is measured with the voltmeter. The ratio of the
two measurements, expressed in decibels, is the signal-to-noise ratio. The FFT
analyzer also can measure S/N ratio and has the advantage of showing the
frequency distribution of the noise.

Acoustical Frequency Response

To measure frequency response of a loudspeaker requires a setup similar to
that of gure 1, with the addition of a power amplier and a microphone.
Figure 3 illustrates the setup. The loudspeaker and the microphone must be
in an anechoic room to eliminate the inuence of sound reections. Alterna-
tively, the testing can be performed outdoors if the background noise is low.
The microphone and power amplier are also in the signal path and are mea-
sured as well, so it is necessary to ensure that they have at frequency response
so as not to distort the measurement. By making several measurements with
the loudspeaker at different angles, it is possible to plot its polar response pat-

468
 Appendix 5

Figure 3

tern, or its response versus angle. This measurement is very important in

assessing the subjective effect of a given loudspeaker, for off-axis sounds excite
the reverberation of the listening room, coloring the perceived sound adversely
if they are not of uniform strength.
This same measurement is easily made with the dual-channel FFT ana-
lyzer. One channel looks at the signal at the loudspeaker terminals and the
other channel looks at the microphone output. See gure 4. If the time delay
between the two signals (due to the nite speed of sound) is taken into
account, the phase versus frequency of the speaker can be measured as well.
The analyzer will do the time compensation. The dual-channel FFT approach
is a great timesaver, and polar plots and response plots at various levels are
quickly made.
It is possible to make anechoic or free-eld measurements in real time
by using a time-gating technique, or time delay spectrometry (TDS). The sig-
nal from the microphone is passed through a sweeping bandpass lter whose
sweep rate is the same as that of the signal generator. The instantaneous fre-
quency of the lter is offset from that of the generator by a constant amount.
This frequency offset is equivalent to a time offset because of the propaga-
tion time between the speaker and the microphone. By the time the micro-
phone receives the signal, the generator is emitting a different frequency. Thus,
the lter in the microphone circuit is effectively looking at a different time,

469
Appendix 5

Figure 4

and it ignores signals arriving later. The length of this time window is a
function of the lter bandwidth. The room reections reaching the micro-
phone are at different frequencies because of their time delays, and so are
not measured.
TDS is a powerful and complex technique that can be used to separate sig-
nals closely spaced in time. For example, the sound reected from the rear of
a loudspeaker cabinet and coming through the speaker cone a little later can
be measured. The dual-channel FFT analyzer is able to do a similar mea-
surement by gating the signal to the loudspeaker; in other words by using
short bursts of random noise and opening the time window of the analyzer
so as to hear the direct sound and ignore the reected sounds. The impulse
response of the speaker is also measured with this instrument. This mea-
surement will show that sounds from the loudspeaker do not all arrive at the
listener at the same time, there being various reections from parts of the cab-
inet, and there being frequency-dependent time delays in the loudspeaker
drivers and crossover.

Harmonic Distortion
In the classic technique for measuring harmonic distortion, a single-frequency
signal is input to a device and the output signal is passed through a notch
lter tuned to that test frequency. The notch eliminates the input frequency
from the measurement, and all that is left are the harmonics and any resid-
ual noise. This residue is measured with a voltmeter and expressed as a per-
centage of the input signal level.

470
 Appendix 5

Sometimes a tunable bandpass lter is used to look at the residual output

signal in order to determine which harmonics are present and to reduce the
contribution due to the noise. A faster and more meaningful way to measure
harmonic distortion is to use an FFT analyzer. The spectrum of the output
will show the input frequency and all the harmonics, which can each be accu-
rately noted in terms of level. Also, any noise in the output is effectively
excluded from the measurement. See gure 5.
Harmonic distortion of a phono cartridge is measured using a test record
that has a sine-wave sweep recorded on it. The signal is replayed through a
sweeping lter that is offset to track the second or third harmonic of the test
signal rather than the fundamental. If the fundamental is tracked, the frequency
response of the cartridge is measured. If the test frequency on the record goes
to very low frequencies (typically 1 or 2 Hz), the effect of the tonearm reso-
nance can be measured.

Figure 5

471
Appendix 5

Intermodulation Distortion

Intermodulation can be measured by inputting two simultaneous sinusoidal

signals to the DUT. The amount of residual energy at the difference frequency
between the two signals is then measured. See gure 6. A relatively new inter-
modulation test sweeps the two input signals over a wide frequency range
while keeping the analyzer at the output tuned to the difference frequency,
which is held constant. Many audio devices are found to have easily audible
intermodulation even when the two test frequencies are outside the audible
range. This causes trouble, for instance in phono preamps where ultrasonic
noise from the record causes intermodulation products in the audible range.

Figure 6

Transient Intermodulation Distortion (TIM)

From Fourier analysis we know that the performance of any linear system is
completely described by its frequency and phase responses, provided they
are measured over a wide enough frequency range. Unfortunately, this is only
true for linear systems, and it is nonlinearities that create distortion. Transients
can drive an otherwise linear system into a nonlinear range, and the behav-
ior of the system is not predicted by classic theory.
The cause of TIM is the nite time delay inherent in an electronic circuit
that has feedback around it. This time delay means the feedback arrives at
the input too late to reduce the distortion of very high frequencies. The slew
rate of an amplier, or how fast its output will move from one voltage to
another, is related to its susceptibility to TIM.
There is no generally agreed-upon method for quantitatively measuring
TIM, although several schemes have been proposed. One of these is to use
a test signal consisting of a band of pink noise with a one-third octave wide
notch in it. The notch is placed in the mid-frequency range, where our hear-

472
 Appendix 5

ing is most acute. When the signal is passed through the DUT, any residual
signal in the notch is due to TIM. The problem with this method is that inter-
modulation and harmonic distortion will also result in energy in the miss-
ing band.
Another suggested approach is to use a square wave mixed with a high-
frequency sine wave as a test signal. TIM will cause the high-frequency part
of the signal to be amplitude modulated at twice the square wave frequency
because of ineffective feedback while the device is slewing from one side of
the square wave to the other. A spectrum analysis of the signal will reveal
sidebands around the high-frequency signal caused by the modulation. There
is no standard that species the frequencies or their levels or how to state the
result in numbers.
A third method of measurement is to use the previously discussed twin-
tone intermodulation test at very high frequencies, above the audible range.
Again, there is no standard procedure.
We do not know the complete answer to TIM problems, but we do know
that listening tests conrm its audibility. It seems safe to say that measure-
ments in the past were conned to too narrow a frequency range to adequately
investigate TIM.

Frequency Response of the Listening Room

As was mentioned before, the acoustics of the listening room will inuence
the frequency response of loudspeakers. If a playback system is to be opti-
mized, then this inuence must be measured. One of the best ways to do this
is to play a test recording that has one-third octave bands of pink noise
recorded on it. The recording is played and the sound pressure level (SPL) is
measured with a sound level meter at the listening position. Ideally, if the over-
all frequency response is at, all the bands of noise will be at the same SPL.
This is a simple test to conduct and involves a minimum of instrumentation.
It might seem that third octaves are too broad in frequency to give adequate
resolution to the measurement. It has been shown, however, that the ear is not
very sensitive to deviations in frequency response as long as they are less than
one-third octave apart and if they are not too extreme in amplitude.
This is a good test to perform to determine the best loudspeaker position
in a room. Large changes in low-frequency performance will be seen for dif-
ferent locations, and the one giving the most uniform response can be cho-
sen (consistent with domestic harmony, of course). Most listening rooms will
be plagued with standing waves, which result in uneven distribution of low
frequency level around the room. Moving the loudspeaker to various places
will not change this, contrary to some opinions; but moving the loudspeaker
will change the relative level at which the standing wave pattern will be excited,
and can result in more uniform response in at least one listening position.
This test will also allow one to adjust the tone controls of the sound sys-
tem for optimum performance, but some care must be exercised. It is not true

473
Appendix 5

that the acoustic response of the listening room should be equalized to be at

to 20 kHz. There should be a gradual roll-off above 2,000 Hz or so, amount-
ing to about 10 dB at 15 or 20 kHz. The reason for this is that the measuring
meter is sensitive to the sound from the reverberant energy of the room, as
well as to the level of the direct sound from the loudspeaker. Our ears, how-
ever, are much more sensitive to the direct sound than to the reverberation.
Also, the loudspeaker will always be more directional at the high end than
at mid and low frequencies. This means the reverberant eld of the room is
less excited at high frequencies than at low. The room also has much greater
sound absorption at high frequencies compared to low. These effects mean
that if the reverberant level is equalized to be at, the direct sound will be
grossly accentuated.
In recent years, it was in vogue to equalize playback systems by placing
one-third octave graphic equalizers in them, performing the above-described
test, and adjusting the response to be at. This is a questionable practice for
two reasons: the rst is the reverberant effect just discussed, and the second
is the fact that the equalizer will introduce phase anomalies and ripple in the
response. Poor phase response is especially bad in the upper and middle
ranges, where it degrades stereo imaging. At least the phase response in the
two channels should be the same for proper stereo. The upper frequency ranges
are best modied with conventional tone controls, which have very gentle
slopes and relatively smooth phase curves. At the low-frequency end, where
much of the irregularity in response will be found, it is easier to justify nar-
rowband equalization.
Theoretically, it makes more sense to equalize the sound power output of
the loudspeakers to be at than to equalize the reverberant eld. This can be
approximated by measuring the speaker output at close range rather than at
the listening position. In this way, the actual signal input to the room is made
more uniform as a function of frequency. The room will modify the response,
of course, but it will do so in the same way it modies any other sound source
in the room, such as a live musical instrument. If a violin were to be played
in a room, there would be no attempt made to equalize its response at the
listeners position! It is true that some rooms are very bright and others very
dull. Subtle adjustments of the tone controls can assist in these cases.
In any case, it is important to remember that the ear is the nal measuring
instrument, and it must be satised in the end. We were once asked to listen
to a newly installed sound system and to comment. The sound was extremely
bright, harsh, and unpleasant. When we pointed this out to the owner, he said,
Yes, but I have equalized the system and it is absolutely at to 18 kHz just
where you are standing.

Measurements of Digital Audio Systems

All the techniques discussed above can be and are used to evaluate digital
audio systems. However, digital systems frequently produce noise and dis-

474
 Appendix 5

tortion in such small amounts that conventional instruments may not be able
to measure them. For instance, to measure harmonic distortion of a device
down to .001 percent requires a sine-wave test generator with signicantly
less distortion than that. Such generators are rare.
In the measurement of digital systems, sometimes the frequency of the test
signal will inuence the measured result. If the test signal is a submultiple of
the sampling frequency, or is related to the sampling frequency by a rational
factor, quantization noise (or quantization error) will be added to it in exactly
the same phase each cycle of the signal. The output waveform will be peri-
odic, and the error will show up as harmonic distortion. If, on the other hand,
the test frequency is not related to the sampling frequency, the quantization
error will show up as broadband noise because the individual cycles of the
signal will be sampled at random times. These two types of quantization error
will sound very different, even though the only difference in the input sig-
nal is a slightly different frequency.
For measurement of noise, harmonic, and intermodulation distortion of
digital systems, some sort of DSP (digital signal processing) device is proba-
bly best. An example of such a test device is the high-resolution FFT analyzer.
But the analyzer itself, being also a digital device, suffers from the same types
of limitation that the devices under test exhibit. An example is phase shift in
the anti-aliasing lters. It so happens that these lters are not needed if the
input signals are band-limited to below the Nyquist frequency of the analyzer
and so they may be switched out. This is normally possible when measuring
digital devices, for the anti-imaging lter in the device under test will effec-
tively band-limit the output signal.
The FFT analyzer, however, analyzes signals of different frequencies in dif-
ferent ways. If an integral number of periods of the test signal t into the time
window of the analyzer, the analysis will be correct. But if the period of the
test signal is such that the analyzer time window truncates part of the signal
waveform, the phenomenon of leakage will occur, and the frequency res-
olution of the analyzer will be reduced. Therefore, test signal frequencies, such
as are found on compact discs designed for testing CD players, should be cho-
sen carefully and the type of analyzer to be used kept in mind.
A test signal suitable for measuring the response and phase characteris-
tics of the anti-aliasing and anti-imaging lters of CD players is the impulse.
An impulse can be made by setting all the bits of the PCM code to 1 (maxi-
mum amplitude) for one sample, and thereafter setting all the bits to 0 (no
signal amplitude). The spectrum of such an impulse is perfectly at in ampli-
tude and phase over the entire bandwidth of the CD player, so a measurement
of the output spectrum of the player will be the response of the system, includ-
ing the anti-imaging lter. The impulses on the test disc should be spaced far
enough apart so that the lter will stop ringing from one impulse before the
next one arrives. This allows evaluation of the time response of the lter.
When digital audio signals are recorded on videotape recorders or com-
pact discs, the PCM code is arranged into frames occurring at 30 per sec-
ond in order to simulate a video signal, and 1 every 136 microseconds in the

475
Appendix 5

case of the compact disc. On playback, the frames of data must be put back
together into a continuous code so the audio signal will be continuous. Because
of uncertainty in the timing of the beginning of each frame, there will be a
time jitter introduced onto the signal. This is like a high-speed utter, or phase
modulation, and cannot be easily measured with conventional utter meters.
But it is still a type of distortion of the signal, and is thought to be audible
under certain conditions. (It could be the cause of the roughness or grain-
iness some people say they hear in digital systems.) It should be measured,
although no standard method is available.

476
Appendix 6
The Balanced Line

In order to transmit power from one place to another by electrical means, two
conductors are required; e.g., a light bulb needs two wires, one for the cur-
rent to enter and one for the current to leave. Likewise, in order to send sig-
nals, which are alternating electric currents, two conductors are also required.
It was found in the early days of the telegraph industry that the earth could
be used as one of the conductors, meaning that only a single wire was needed
to send telegraph signals. See gure 1.

Figure 1

The use of the earth as a conductor is only useful if small amounts of power
are to be transmitted, because the high resistance of the ground results in large
power losses and low efciency. This is the reason your electric utility runs
two wires to your house rather than using the earth as a return path.
In electronic equipment, such as audio ampliers, it is convenient to con-
nect one terminal of the power supply to the chassis of the device. This sim-
plies internal wiring somewhat. The chassis is called the ground side of
the power supply, in keeping with the telegraph tradition.
As can be seen from gure 2, one terminal of the signal is common to one
terminal of the power supply. By connecting this terminal to the chassis, the
chassis is made to act as a shield around the circuits and thus reduce the inter-
ference due to external elds, such as 60-Hz hum from the power lines. In nearly
all audio circuits, this chassis connection is continued in the form of a shield
in the cables which interconnect devices. The shield isolates the inner conductor
from electrostatic hum elds and also serves as one of the conductors for the
audio current. This type of interconnection, where the shield carries signal cur-
rent, is called single-ended or unbalanced. Single-ended circuits are ne
for almost all short cable runs that carry relatively high signal voltages.

477
Appendix 6

Figure 2

However, if a situation arises where the shield has induced in it a noise

current, such as a 60-Hz current due to a ground loop or magnetic induction
from a power transformer, the noise will be added to the signal, causing an
audible hum. If any two devices whose chassis are not at exactly the same
potential are connected, the shield will carry 60-Hz current as well as signal
current, and the signal-to-noise ratio will be degraded.
The way to avoid this possibility is to arrange things so that the shield does
not carry any signal current, and any currents induced in the shield will not
be added to the signal. One way to do this is to place a transformer at the
input of the receiving device. See gure 3. One secondary terminal is con-
nected to the chassis, but neither primary terminal is so connected because
the two windings are insulated from one another. This conguration is called
an isolated or oating input. The transformer input exhibits another very
important characteristic called common-mode rejection. The transformer is
only sensitive to current passing through its primary winding, and the only
way to establish such a current is to apply a voltage difference to the input
terminals. If the two terminals are at the same voltage at the same time, even
though it may be a varying voltage, there will be no current in the winding
and no resultant output from the transformer. This arrangement is also called
a differential input or balanced input.
A long audio cable, such as a microphone cable, will have noise induced in
it by stray magnetic and electrostatic elds. The same voltage will be induced
in both conductors of the two-wire cable, and if these two wires are connected
to the transformer primary, the induced noise will not be sensed by the trans-
former because it has the same voltage in both conductors. The noise is said
to be common-mode voltage, and the transformer rejects it. This is common-
mode rejection. The microphone, on the other hand, generates a differential
voltage in the wires, causing currents that are equal and of the opposite direc-
tion in each wire, and the input transformer is sensitive to this. Such an
arrangement is called a balanced line.
Further Renements: This long balanced line is still susceptible to hum

478
 Appendix 6

Balanced Amplier Circuit

Figure 3

induced differentially in the conductors if there is any appreciable area

between them through which a magnetic ux can pass. Therefore, the two
wires must be kept as close together as possible. The best way to do this is to
twist them together, forming a twisted pair. All long audio lines are of this
type. The twisting also helps in another way because the current induced in
each half twist is in the opposite direction to that induced in the next half twist.
Over the length of the cable, the induced currents tend to cancel. Multiple
twisted pairs can be run alongside each other for long distances without
crosstalk.
Another renement is to place a shield around the twisted pair and con-
nect it to the system ground. This prevents very large common-mode volt-
ages from building up, reducing the noise even further. This shield is less
important than the shield on a single-ended circuit, which is absolutely essen-
tial because it carries signal current. Telephone lines are an example of bal-
anced twisted pairs that are not shielded.
It is very important that the shield on a balanced line not be connected to
either of the signal wires. To do so will throw the line out of balance. It is also
important that the shield not carry any current, especially any 60-Hz power-
line-induced current. For this reason, sometimes the shield will be connected
to ground on one end only, the chassis being connected by a separate wire.
See also pin 1 problem. There is one exception to this rule, and that is the
case of the phantom powering of some condenser microphones, where the
shield carries the preamplier DC power current. This is a small current, how-
ever, and it is direct current, so it normally causes no problems. In long micro-
phone lines, if the ground connectors on the cables are not in good condition,
a poor shield connection can cause intermittent phantom current, resulting
in noise.
Nowadays, transformers used for balanced lines are falling into disuse,
and are being replaced by electronic circuits that mimic the transformer. Such
circuits have the advantages of lower cost, lighter weight, and superior
performance.

479
Appendix 7
Some Notes on Digital Audio

Digital audio is a very wide-ranging subject that can be described in general

terms as the application of digital computer-based technology to the pro-
cessing, recording, and reproduction of music. For details on how this works,
see the many dictionary entries on digital audio devices.
In all digital systems, the audio signal is reduced to a series of numbers
that are then manipulated with computer-like techniques to effect time
delays, reverberation, equalization, editing, etc., and/or storage.
The rst commercially available digital audio equipment was the digital
time delay system, used for anging, phasing, and articial reverberation. The
next was the digital tape recorder, which stores the signal in digital form on
magnetic tape. One advantage of digital storage is that copies can be made
with theoretically no degradation in quality. A copy of a digital tape is an exact
replica of the original digital tape (provided no bits are lost and no noise
is added in the copying process). In practice some bits will be lost in copy-
ing, and any of several error correction schemes can be used to restore the
signal to a reasonable facsimile of what it was. Generally, these error correc-
tion processes work well and the corrected signal usually cannot be distin-
guished by ear from the original, but if multiple copies of a signal are made,
each requiring error correction, obvious signal degradation will be the result.
Some advantages of digital storage of music signals are extremely wide
dynamic range and freedom from noise, very wide and at frequency re-
sponse, and very low distortion. The rst practical truly digital medium avail-
able to the consumer was the compact disc, which uses an optical, rather than
magnetic, coding of the signal. In this, it was extremely innovative. The com-
pact disc had the highest density of information storage of any available
medium when it was introduced and is replayed by shining a tiny laser beam
onto the disc surface, which contains microscopic pits that alternately reect
and cancel the beam. Because there is no physical contact involved, there is
no wear on the disc, and no degradation of sound with repeated playings,
assuming the disc is not scratched, contaminated with foreign material, or
otherwise deformed.
At the time of its introduction, the compact disc was probably the highest
quality medium for reproducing sound ever offered to the public, and this
fact was much ballyhooed by the press and the manufacturers. The SACD
and the DVD are signicant techical advances over the CD. The SACD pro-
vides an improvement in signal-to-noise ratio and distortion, and the DVD
is used for either audio or video or both. It offers much longer playing time
than the CD or SACD and the capability of multichannel audio.
Digital audio, and in particular, compact discs, have been heralded as the

480
 Appendix 7

last word in quality, the ultimate state of the art in sound reproduction beyond
which there can be no improvement. This, of course, is nonsense. While dig-
ital systems do offer much reduced noise and distortion compared to ana-
log systems, they still do have noise and distortion. The types of distortion
are quite different from those of analog systems, and their objective mea-
surement and subjective evaluation are difcult. For instance, certain digi-
tal systems usually had large amounts of high-frequency phase shift and
group delay distortion, but this has been much improved in much of todays
equipment.
The nonlinear distortion in digital systems tends to increase as the signal
level decreases, meaning soft passages are more distorted than loud ones.
As signals become very small, the energy in the fundamental components
of the sounds becomes distributed into the harmonics, and this is a timbre
change. This also amounts to dynamic expansion, for the level of the fun-
damentals is proportionally too small as signal levels are reduced. There is
good reason to question the 44.1-kilohertz sampling rate used in the CD sys-
tem, which was chosen for easy compatibility with already-existing video
tape recorders. A higher sampling rate, such as the 48 kHz chosen for pro-
fessional digital tape recorders, would allow less steep anti-alias and anti-
image lters, with consequently reduced phase distortion. The problem is
that it would be very difcult to change now, for compatibility with discs
and players in present use would be destroyed. This is the universal prob-
lem with early standardization.
There is no doubt that digital audio systems sound different from the best
analog systems, and it is easy to convince most people that they sound bet-
ter. However, quite a few listening tests where digital and high-quality ana-
log master tape recordings of the same performance are played side by side
to a panel of listeners have been conducted, and in many of these the analog
recordings are judged as being more musical and pleasant, even though
they are admittedly a little more noisy. Extended critical listening of CDs
reveals that there are certain problems in some areas, such as tonal balance
and a sense of ambience or spaciousness.

481
Appendix 8
Musical Scales and the Tuning
of Musical Instruments

The tonal space in which music exists is dened by the subjective pitches of
the notes of musical scales, which are almost exclusively determined by the
objective frequencies of these pitches. If one examines the musical scales that
are used in various cultures around the world, a large variety is found, con-
sisting of quite distinct pitch intervals. However, there is one universal inter-
val that is always recognized as a fundamental basis for musical expression,
and that interval is the octave. It seems that a part of being human is the
recognition of the octave not as an interval like any other but as an over-
riding sense of pitch or tonality. Octave transpositions are frequently unno-
ticed by musicians, as if all the notes spaced apart by octaves have the same
pitch.
The familiar do re mi fa sol la ti do notes of the diatonic scale arise nat-
urally from the fact that the overtones of musical instruments consist of fre-
quencies that are integral multiples of their fundamental frequency. Let us
investigate this situation and attempt to understand how this musical scale
is logically built.
The octave, the most musically consonant interval, is simply a frequency
ratio of 2 to 1. The second harmonic of a musical sound has a frequency and
a pitch one octave above that of the fundamental. Two simultaneously sound-
ing notes one octave apart blend together extremely well, for the higher one
has the same frequency as the second harmonic of the lower one. Each har-
monic of the lower one will have a harmonic of the higher one at the same
frequency.
The next most consonant interval is the perfect fth, which has a frequency
ratio of 1.5 to 1, or 3 to 2. The second harmonic of the higher tone will have
the same frequency as the third harmonic of the lower tone, and the two tones
blend together extremely well. The musical interval between the second and
third harmonics of a musical sound is a fth.
The musical interval between the third and fourth harmonics is a perfect
fourth, with ratio of 3 to 4, and the interval between the fourth and fth har-
monics is a major third, with a ratio of 5 to 4. The next two harmonics, num-
bers ve and six, span a musical interval of a minor third.
All these harmonic frequencies can be transposed downward in octaves until
they are in the octave just above the fundamental, or tonic. If this is done,
and if we consider the tonic to be C, the third harmonic becomes G, called the
dominant. The transposed fourth harmonic is F, the subdominant, and
the fth harmonic becomes E, or the mediant. The C-E-G together form the

482
 Appendix 8

major triad, a very important chord in music. If we start from G and form
another major triad above it, the notes are G-B-D, giving us the leading tone
B and the second D. In like manner, we can start from the subdominant F
and build a major triad on it. This produces the submediant A and brings
us back to C. All eight notes of the diatonic scale in the key of C are now present,
and they are tuned to the perfect intervals known as just intonation. These
intervals produce chords that have no beats and sound perfectly smooth.
This construction of a just scale by taking various harmonic frequencies
of the tonic C is illustrated in musical terminology in gure 1.

Figure 1

It is tempting to think that the just system of tuning the scale would be an
ideal basis for music and that if a musician had a free choice of frequency for
each note, she would always play in just intonation. Surprising as it may seem,
however, this is impossible, for the diatonic scale and these perfect intervals
are not compatible. To illustrate this, consider the sequence of chords shown
in gure 2, which contains ascending intervals of a sixth and a fourth, and
then descending intervals of two fths. Refer to the black notes in gure 2.
Let f be the frequency of the tonic C. The rst sixth produces A, with a fre-
quency of (53)f. In going to D, an interval of a fourth is required, and this is a
frequency ratio of 43, so its frequency will be 43 of A, which is 53 of f, or (43)(53)f,
which is (209)f. The descending fth gives G, at a frequency of (23)(209)f = (4027)f.
The last fth brings us back to the tonic C, with a frequency of (23)(4027)f =
(8081) f. But this is not the same frequency f we started with. This error inter-
val, with a frequency ratio of 8081, is called the syntonic comma, and amounts
to about one fourth of a semitone. This means that after this simple ve-chord
progression, our tonic is no longer at the same frequency at which it started.
Many players, especially string and brass players, think that when play-
ing together without keyboard instruments, they play all perfect beatless inter-
vals in just intonation. As we have just seen, this is impossible without allowing
the pitch of the tonic to wander aboutand this is never allowed in music.
Another way to look at this situation is to consider twelve ascending per-
fect fths (the circle of fths), followed by seven descending octaves. The
pitch discrepancy between the starting note and the ending note is called the

483
Appendix 8

Figure 2

diatonic comma or the comma of Pythagoras, after its most famous investi-
gator, and amounts to a little over 1%, or about one sixth of a semitone.
It is surprising that this state of affairs should exist, for it seems that since
our basis for musical harmony is the naturally occurring harmonic series of
overtones consisting of simple frequency ratios of small integers, they should
all be commensurate with each other, and there shouldnt be such glaring
errors as the syntonic and diatonic commas. It seems that the octave, which
is really a single pitch percept, is incompatible with the simple frequency ratios
of the intervals of the diatonic scale. Another way of saying this is that the
following integer equation must be false for all integers:
n

YX  f = 2mf
where X, Y, n, and m are integers, and f is the frequency of the tonic. The left-
hand side of the equation represents successive steps of musical intervals, and
the right-hand side represents octave transpositions.
It can be shown with elementary number theory that this equation can-
not in fact ever be satised, except for the trivial case where the left-hand side
also represents an octave, and therefore, it is true that the octave percept is
not a subset of the harmonic, or intervalic, percept. The two percepts are inde-
pendent of one another, like the X and Y coordinates of the Cartesian plane.
Moreover, it can also be easily shown that intervals of major thirds are not
commensurate with perfect fths, the difference being the syntonic comma.
Any system of tuning a musical scale, especially in the case of a keyboard
instrument or fretted instrument, must contend with this basic difference
between the concept of the octave and the noncommensurability of musical
intervals with respect to the octave. We must introduce some sort of com-
promise in tuning, and this compromise is called a temperament, of which
there are theoretically an innite number.
One of the earliest known schemes used for tuning keyboard instruments
was the so-called Pythagorean intonation. Strictly speaking, this is not a tem-
perament, for it is based on pure perfect fths. It consists of eleven perfect
fths starting on E at and ending on G sharp. The fth between these last
two notes is one comma too small, and the interval is so discordant as to be
unusable. This poorly tuned fth is called the wolf, presumably because it
howls so.

484
 Appendix 8

Pythagorean tuning also has very poor thirds, most of which are too wide.
It sounds better when used melodically rather than harmonically, and those
keys must be avoided where the wolf would occur. Pythagorean tuning was
probably used for keyboard instruments until near the end of the fteenth
century.
An early attempt to improve Pythagorean tuning was the so-called mean-
tone temperament, in which the thirds are tuned pure, and all the fths are 14
comma too narrow. The circle of fths cannot be closed, and the wolf, between
E at and G sharp, amounts to 134 comma and is even worse than the wolf in
Pythagorean tuning. Meantone temperament produces very rich and satisfy-
ing harmony in keys near C major because of the pure thirds. All the whole
tones are equal and are precisely half a major third, hence the name meantone.
Much effort was expended in trying to modify meantone tuning to reduce
the size of the wolf in order to make more keys available and still keep most
of the beautiful pure thirds. This led to hybrid Pythagorean-meantone, or
baroque, temperaments. One such system tunes four of the fths 14 comma
narrow and all the others pure, eliminating the wolf by spreading the comma
over four fths. The thirds range from pure to Pythagorean. Andreas Wer-
ckmeister designed several such tunings around 1690, his number III being
the best known. In this variation, the keys with the fewest accidentals have
the best thirds and sound the purest, whereas keys with the most accidentals
sound the wildest or most brutal. Incidentally, Werckmeister said this
shouldnt be a problem because the ordinary organist cant play in those keys
anyway(!). Nevertheless, it was theoretically possible to play in all keys, and
this led to these temperaments being called wohl temperirt, or well tem-
pered, and it is probable that J. S. Bach wrote his famous Well-Tempered Clavier
with one of these tunings in mind, although there are those who believe that
he favored equal temperament.
From about 1800, the most used type of tuning has been equal tempera-
ment, in which the comma of Pythagoras is evenly divided among the twelve
half-steps of the octave. Although the idea of equal temperament predates
Bach, it was not generally used in the baroque era because it was thought to
be boring. All the half-steps are of equal size and are exactly one twelfth of
an octave. They span a frequency ratio of the twelfth root of 2, which is about
6 percent. In equal temperament, all the intervals are the same regardless of
the key in which one is playing, and none except the octave is perfectly tuned.
This makes it very easy to modulate from one key to another, although the
keys lose their individuality because they all have equal intervals.
Many theorists complain about the fact that the intervals in equal tem-
perament are all impure, but in fact the fourths and fths are within 0.001%
of just intervalsso close that most musicians have a hard time telling them
apart. The thirds, however, are not very satisfying, being about 0.01% away
from pure thirds, and they have annoying audible beats. And of course all
the thirds in all keys are equally bad.
With the advent and almost universal use of equal temperament, music
today lacks the variety and richness that it had when many different systems

485
Appendix 8

of tuning were in common use. Of course, we must live within the capabili-
ties of the musical instruments we have, such as the piano with only twelve
keys per octave and its equal temperament. But electronic music has no such
tuning limitations. A composer could dene any number of intervals within
an octave, even one thousand or more, and this is within the range of even
relatively simple synthesizers. One can only hope that the vast variety and
richness of intonation that is readily available to us will be recognized more
fully and exploited by our musicians. They have only to look into the past
for countless inspirational examples.

486
Appendix 9
Some Notes on the History
of High Fidelity

The roots of high-quality sound reproduction lie not in the early days of radio
broadcasting and receiving but in the sound motion picture industry, which
has its beginning about 1930. A great many of the principles that govern music
recording and reproduction today were discovered and developed in the thir-
ties, with World War II putting an effective damper on further exploitation
during the decade of the forties. Most of the breakthroughs that occurred in
this fertile decade were the direct result of research at the Bell Telephone Lab-
oratories under Harvey Fletcher and at the RCA Research Laboratories under
Harry Olson. It was Fletcher and Munson who discovered and quantized the
very nonlinear behavior of the human hearing mechanism about 1933 (the
famous Fletcher-Munson effect), and this represents the beginning of psy-
choacoustic research, at least in this country.
The Bell Labs had an interest in this type of research and development
partly because they were interested in the perfection of the telephone and
partly because the manufacturing arm of the Bell company (Western Electric)
was interested in building sound systems for movies and also for public
address use. The equipment built during this time by Western Electric in-
cluded microphones, ampliers, preamps, phonograph reproducers, radio
broadcasting equipment, and loudspeakers, and was of excellent quality. In
fact much of this venerable gear is in great demand today and is eagerly sought
after by many acionados. The feedback amplier, the moving-coil phono
cartridge, the exponential horn loudspeaker, the compression driver for high-
frequency loudspeakers, the crossover network, the compander, SVA sound-
on-lm recording, stereophonic recording and reproduction, and many more
techniques and devices were introduced by Western Electric and the Bell Tele-
phone Laboratories people in the thirties.
Alan Blumlein, chief engineer at EMI in England, carried out much re-
search in audio and also developed a two-channel stereo system for motion
picture sound. It seems that he and Bell Labs were unaware of each others
work.
Harry Olson, the director of the RCA acoustics research laboratory for many
years, was responsible for over one hundred patents relating to audio, most
notably pertaining to directional microphones and many improvements to
loudspeaker designs, including acoustic suspension.
The consumer began to get interested in high delity after World War II,
but had very little to choose from because the major radio manufacturers paid
no attention to what they thought was far too small a market. An interesting

487
Appendix 9

article in Fortune magazine of October 1946 stated that according to research

by CBS, the public did not prefer wide-range sound reproduction, and musi-
cians especially advocated the deliberate limitation of high-frequency response
to below 5 kHz, even when they could have had response to 10 kHz at the
time. Hi- enthusiasts had to buy equipment from commercial sources, such
as theater equipment dealers and broadcast equipment suppliers, and had to
build their own systems.
But it was not long before small companies began to spring up to service
the growing hi- market, and the 1950s saw the birth of the consumer audio
business. One of the early entries into consumer audio was the Altec Lans-
ing company, which had originally been formed as the All Technical Service
Corporation and which had acquired all the Western Electric patents and
designs for theater sound systems. Western Electric had left the theater sound
business just before the start of World War II. The Lansing Manufacturing com-
pany, which had been making commercial loudspeakers, merged with All
Technical, and the name was shortened to Altec Lansing company. James B.
Lansing went on to form JBL about 1946 and was considered by many to be
the preeminent consumer loudspeaker maker of the time. Lansing himself
became despondent over business indebtedness and tragically committed sui-
cide in 1949.
Another pioneer in consumer audio was Avery Fisher, after whom Avery
Fisher Hall in New York is named. He founded the Fisher Radio company
and became one of the rst manufacturers to produce a complete high-delity
system, including even the cabinet. Fisher also built components, and his
receivers were especially well respected.
The modern microgroove LP record was introduced by Columbia in 1948,
and it was an immediate success. Of course, Edison had a long-playing record
as early as 1929, so Columbias innovation was not exactly new. It did estab-
lish 3313 rpm as a standard. (This was also borrowed from Western Electric,
which rst used that speed for the Vitaphone motion picture sound system.)
Before the Columbia engineers announced the LP publicly, they approached
RCA, hoping to get an agreement from them to join them in making LPs, but
RCA did not do so. RCA then introduced the 7-inch microgroove 45-rpm disc
and player in 1949, thus initiating the famous battle of the speeds. The ration-
ale was that the most popular musical selections were no more than 5 min-
utes long, and the 7-inch disc was easier to handle than a 12-inch disc! The
loser in the battle of the speeds was of course the consumer, for he now had
to buy another turntable, or at least a two-speed model.
The introduction of the quiet vinyl microgroove records made existing
78-rpm record players obsolete, and the reduced distortion and noise levels
attainable placed much higher demands on existing ampliers and preamps,
initiating a new wave of improved components. Around this time, stereo-
phonic sound was becoming known, and some two-channel tape recorders
were becoming available, along with prerecorded stereo tapes. Emory Cook,
one of the pioneers in LP technology, introduced a stereo LP record that had
two separate grooves and was played by two cartridges in a double-headed

488
 Appendix 9

tonearm made by Livingston. It was not commercially viable, for it was impos-
sible to preserve anything like correct phase relationships between the two
channels.
The stereophonic LP record was introduced in 1958, and it was again an
outgrowth of stereophonic experiments conducted in the 1930s by Western
Electric. About this same time, FM stereo broadcasting began. These devel-
opments of course made existing sound systems obsolete, for now two pre-
amps, ampliers, and loudspeakers were needed; so another revolution in
components was born.
In the eld of loudspeakers, one notable development was the introduc-
tion of the Bass Reex principle by the Jensen company in the 1930s. This was
the rst of the vented loudspeaker systems and resulted in extending the
low-frequency limit about one-half octave below what was common for that
day. Another signicant loudspeaker development was the Klipschorn,
patented by Paul W. Klipsch in 1941. This system had a folded low frequency
horn designed to t into the corner of a room so that the walls and oor of
the room effectively extended the horn. It was characterized by prodigious
low-frequency output for its day and also had very high efciency. Its high-
frequency section was a straight horn coupled to a high-efciency compression
driver. Even though the Klipschorn was good at radiating low frequencies,
it also had a series of resonances in the mid-bass range, causing it to sound
hollow and cavernous, especially on voices. A true revolution in loudspeaker
design occurred in 1954 when Edgar Villchur introduced the acoustic sus-
pension principle in the form of the AR-1. This bookshelf-sized system was
capable of smooth, highly damped, and extended low-frequency response,
and it outperformed almost all other speakers of the day. Villchur had tried
to interest several speaker makers in his idea, but none would accept it, so he
and three other people started the Acoustic Research Corp. It is interesting
that in the late 1950s, the courts ruled that Villchurs patent on acoustic sus-
pension was invalid because of a disclosure by Harry Olson much earlier. Had
RCA elected to commercialize the idea in the thirties, the course of future loud-
speaker development would have been far different! Today, a great many loud-
speaker systems use the acoustic suspension principle or variations of it to
attain good low-frequency performance from relatively small boxes. In the
1970s, Thiele and Small undertook a theoretical analysis of the vented loud-
speaker box, greatly adding to the understanding of the principles involved.
This work led to the ability to optimize the performance of systems of dif-
ferent sizes and further led to the introduction of many small-sized systems
with excellent low-frequency performance by many different manufacturers.
While the principle of magnetic recording was described as early as 1888
by the American Oberlin Smith, the professional tape recorder as we know
it today was developed by the Magnetophon company in Germany just before
World War II. John Mullin brought back two of these machines after the war,
and he, along with promised nancing by Bing Crosby, convinced Alexan-
der M. Poniatoff of the Ampex Electric company to build the rst commer-
cial American machine. Col. Richard Ranger built a copy of the Magnetophon

489
Appendix 9

in 1947, which he then used for motion picture sound recording. He had devel-
oped a synchronizing technique, called Rangertone, which allowed the tape
to run in synch with the movie camera.
The rst commercial Ampex machine was the model 200, and it outper-
formed the Magnetophons. It gained instant acceptance as a recorder for delay-
ing radio broadcasts. The Brush Development company had in 1946 made a
home-type tape recorder called the Soundmirror, which used DC bias. Tape
for the machine was being made by the Minnesota Mining and Manufactur-
ing company (3M). The German machines used tape made by the Bayrische
Analin und Soda Fabrik (BASF), some of which was brought back by John
Mullin and used by Ampex in the development of their machine. From this
point, the development of the reel-to-reel tape recorder progressed rapidly.
But the audio compact cassette is by far the most important tape medium for
the consumer. It is the noise reduction systems that have made the cassette
format acceptable for recording music. When originally conceived by Philips
in the 1950s, the audio cassette was never meant to be anything more than a
format suitable for voice dictation, and the strict standards under which Philips
licensed manufacturers prevented true high-delity response. This was espe-
cially true of stereo, where the two tracks are side by side, very narrow, and
very close together.
Although the compander principle of noise reduction had been used much
earlier in motion picture sound, it was Ray Dolby who introduced the Dolby
A noise reduction system for master tapes in the 1960s and who deserves credit
for its modern exploitation. Henry Kloss, one of Edgar Villchurs original part-
ners, who started KLH Corporation and Advent Corporation, convinced Ray
Dolby to let him market a simplied Dolby compander for use with audio
cassette machines. Dolby has been modifying and improving companders for
many years, and his company produces Dolby A and SR for professional use
and Dolby B, C, and S for home use. Other manufacturers of compander sys-
tems have come forward, but the Dolby systems are dominant.
The compact disc system, jointly developed by Philips and Sony, is unques-
tionably the highest quality sound reproduction system for the consumer for
a long time, but improvements have become recently available. The Super
Audio Compact Disc (SACD), introduced by Sony and Philips, is noticeably
better in reproducing stereo music and DVD Audio, derived from the video
DVD system can present music in a variety of formats, from two or three-
channel stereo to 5.2 channel with surround sound.

490
Appendix 10
Some Notes on Decimal
and Binary Arithmetic
as Used in Computer Terminology

Computers and data processing machines use binary numbers in their cal-
culations rather than decimal numbers because it greatly simplies the han-
dling of data. The binary number system uses only 0s and 1s to represent
numbers of any size. The 0s and 1s are the bits (short for binary digits)
of information. When talking about computer data quantities, memory sizes,
and storage medium capacities, many thousands, millions, or trillions of bits
are involved.
The International System of measurements (SI) long ago standardized arith-
metical prexes to indicate multiples of numerical quantities. For instance, it
is well known that kilo means a multiple of 1000 and that mega is one million
times, etc. The SI prexes most commonly used in computerese are these
two plus giga, standing for one billion, and sometimes tera, for one trillion.
So, according to the SI, one kilobit is 1000 bits, and a megabit is 1,000,000 bits.
So far, so good.
Binary arithmetic, which is used in computers almost exclusively, is based
on powers of 2, contrasted with the decimal system, which is based on pow-
ers of 10.
Computer professionals realize that 210 is equal to 1024, which is only about
2% greater than 1000, and since this is quite a small error, they started calling
210 bits (1024 bits) a kilobit and calling 220 bits (1,048,576 bits) a megabit.
This was acceptable practice before the advent of data storage for gigabytes,
and even terabytes, but the storage devices were not constructed on binary
arithmetic, which meant that, for many practical purposes, binary arithmetic
was less convenient than decimal arithmetic. When discussing computer mem-
ory, most manufacturers use megabyte to mean 220 = 1,048,576 bytes, but the
manufacturers of computer storage devices usually use the term to mean
1,000,000 bytes. Some designers of local area networks have used megabit per
second to mean 1,048,576 bit/s, but all telecommunications engineers use it
to mean 1,000,000 (10 6) bit/s. The result is that today everybody does not
know what a megabyte is.
Floppy disk manufacturers are even more confusing. The prex M means
(1000 1000) in SI, and (1024 1024) in standard computing. However, the
standard 1.44 MB oppy holds (1.44 1000 1024) bytes.
In 1999, the International Electrotechnical Commission (IEC) published
Amendment 2 to IEC 60027-2: Letter symbols to be used in electrical

491
Appendix 10

technologyPart 2: Telecommunications and electronics. This standard,

which had been approved in 1998, introduced the prexes kibi, mebi, gibi, tebi,
pebi, exbi, to be used in specifying binary multiples of a quantity. The names
come from shortened versions of the original SI prexes and bi which is short
for binary. It also claries that, from the point of view of the IEC, the SI prexes
only have their base-10 meaning and never have a base-2 meaning.

New IEC Standard Prexes

Name Abbr Factor

kibi Ki 210 = 1024
mebi Mi 220 = 1 048 576
gibi Gi 230 = 1 073 741 824
tebi Ti 240 = 1 099 511 627 776
pebi Pi 250 = 1 125 899 906 842 624
exbi Ei 260 = 1 152 921 504 606 846 976

As of 2004 this naming convention has not yet gained widespread use. The
IEC did not give names for the prexes beyond exa-, but if they had given
them names, they would probably be zebi and yobi.

492
Appendix 11
Some Notes on Impedance
and Frequency Response

The denition of impedance (pronounced im-peed-ents), according to the

International Standards Organization (ISO), is the ratio of the harmonic exci-
tation of a system to its response (in consistent units), both of which are com-
plex quantities, and both of whose arguments increase linearly with time at
the same rate. The term applies to linear systems only and the terms and
denitions relating to impedance apply to sinusoidal conditions only.
Let us take a closer look at this denition and attempt to clarify it. First, it
is very important to note that impedance is a characteristic of a device or sys-
tem and is not dependent in any way on the input signal to the system. We
are concerned here with electrical impedance, although a similar treatment
can be given to mechanical systems, and mechanical impedance is a perfectly
legitimate and useful concept.
The term harmonic excitation means a sinusoidal voltage, or a sine wave,
is applied to the device whose impedance we wish to investigate. The
response is the electric current in the device that is a result of the excitation
voltage. That this excitation voltage and the response current are complex
means that we are to be concerned with their phase as well as their magni-
tude. In other words, the impedance will let us know if the device will apply
a time delay to the excitation signal, and if so, how long a delay. The state-
ment about both arguments increasing linearly with time means that the ratio
of excitation to response is to be measured as a function of frequency. This
implies that the impedance will be different at different frequencies, and this
will be true for nearly all cases.
The fact that the concept of impedance applies to linear systems only sim-
ply means that a nonlinear system or device cannot be said to have an imped-
ance. A nonlinear system behaves differently at different excitation levels, and
its response is not generally predictable from its excitation. Ideal audio sys-
tems and components would be truly linear, but we know that real devices
suffer from certain nonlinearities, and these nonlinearities cause most of the
various distortions that we struggle to reduce. Because audio systems are at
least conceptually linear, and also because they approach true linearity, we
are justied in dening their impedances, even though the denition of imped-
ance would strictly speaking preclude it. In some casesfor instance, when
a loudspeaker is overdriven into nonlinearitywe could say that its imped-
ance varies with drive level.
The denition says that impedance is only meaningful with sinusoidal exci-
tation and response conditions, but this does not mean that we cannot apply

493
Appendix 11

the concept of impedance to systems that are excited by complex waveforms

such as music. The denition simply means that to measure the impedance,
we must consider the ratio of excitation to response at each frequency inde-
pendently, and this implies sinusoidal voltage and current signals. Once the
impedance is known as a function of frequency, the behavior of the system
with complex waveform excitation will be predictable, because the system is
essentially linear.
Because impedance has three dimensionsmagnitude, phase, and fre-
quencyit is difcult to graph on two-dimensional paper. One solution is to
graph the magnitude of impedance versus frequency on one plot, and to graph
the phase versus frequency on another plot.
The three-dimensional impedance curve can be thought of as the locus of
a vector that extends from the frequency axis a distance proportional to the
magnitude of the impedance. The angle the vector makes with a xed coor-
dinate system is the phase angle of the impedance. This curve will in general
be some sort of spiral extending along the frequency axis. The shape of this
curve can be resolved by projecting it onto two planes that intersect along the
frequency axis. One of these planes extends in the zero phase angle direction
from the frequency axis, and the other one extends in the 90-degree phase
angle direction. See Heyser Spiral for graphic details. Mathematically, if a
vector is multiplied by  -1 (abbreviated i by mathematicians and j by
engineers), it is rotated 90 degrees with respect to its coordinate system or in
other words is simply given a 90-degree phase shift. These two projections
of the curve are called, respectively, the real and the imaginary parts of
the impedance. The real part is that part of the impedance that is in phase
with the excitation, and the imaginary part is that part that is 90 degrees out
of phase with the excitation. The actual phase at any frequency can be deter-
mined mathematically by taking the arctangent of the real part divided by
the imaginary part. The term imaginary is unfortunate, for it is a perfectly
valid concept. A more descriptive term for it is the quadrature part, and this
term is preferred. The term imaginary comes from the early mathemati-
cians, who could not quite believe that the square root of negative 1 could
exist.
Another scheme is the so-called Nyquist plot, in which a rectangular coor-
dinate system is used to represent magnitude of impedance as the distance
from the origin and positive phase as a rotation to the right from vertical,
with negative phase as a rotation to the left from vertical. The Nyquist plot
is useful for identifying resonances, because the phase of the impedance will
smoothly vary from zero degrees below the resonance frequency to 90
degrees at the resonance frequency to 180 degrees above the resonance fre-
quency. The Nyquist graph of the impedance of such a resonance will be a
circle, and a system with several resonances will have a circle for each one
in the Nyquist plot. The diameter of the circle is an indication of the magni-
tude of the impedance at the resonant frequency and is thus a function of
the damping of the resonance. The Nyquist plot is often used to represent
the impedance of loudspeakers, where resonances can readily be identied.

494
 Appendix 11

The Nyquist plot does not identify the frequency of resonances; in a sense
it is like looking at the three-dimensional impedance curve by sighting down
the frequency axis.
The phase of the current will be shifted either positively or negatively
depending on whether the impedance is capacitive or inductive. If the phase
is not shifted, the impedance is said to be resistive. The phase of the imped-
ance of a real circuit or device will generally vary from positive to negative,
and it is called inductive if the current lags the voltage and capacitive if the
current leads the voltage.
As mentioned above, mechanical impedance can be treated in a manner
similar to electrical impedance, and is often used in loudspeaker analysis. The
excitation can be a sinusoidal force, and the response can be a resulting motion.
In a dynamic loudspeaker, inductive motional impedance means the cone
motion lags behind excitation and acts like a mass (or is mass-controlled),
whereas capacitive motional impedance means the cone motion leads the exci-
tation and acts like a spring (or is stiffness-controlled). If the impedance
looks resistive (zero phase angle), it means the system is controlled by fric-
tion. At resonance, the mass and stiffness effectively cancel each other and
the phase angle is 90 degrees and the impedance magnitude is controlled by
the damping (friction) of the system.
Another type of mechanical impedance is the so-called acoustic imped-
ance, which is dened as the complex ratio of the sound pressure at a point
in a medium to the complex acoustic volume velocity at the same point.
Acoustic impedance is thus a complex quantity, meaning it can be resistive

495
Appendix 11

and reactive, having phase angles between pressure and velocity that are zero,
leading, or lagging. Acoustic impedance considerations are important in loud-
speaker design, for the impedance is what provides the loading for the speaker.
For instance, the horn in horn-type loudspeakers acts as an impedance trans-
former to better match the driver impedance to the impedance of the air in
the room.
Impedance is expressed as the symbol Z and has units of ohms.
The frequency response function is closely related to the impedance func-
tion; in fact they are reciprocals, for frequency response is dened as the com-
plex ratio of the output signal of a device or system to the input signal. It is
also a three-dimensional quantity, having magnitude, phase, and frequency,
and it is often displayed in the same ways as impedance, such as in Nyquist
plots or real and quadrature parts versus frequency.
In audio devices, such as ampliers, sometimes it is useful to dene two
types of impedanceinput impedance and source impedance. If measured
at the input terminals of the device, it is dened as input impedance, and if
measured at the output terminals, it is source impedance, or sometimes out-
put impedance. However, output impedance usually means the impedance
the output of the device is designed to be connected to. An example is the
output of an audio power amplier that is labeled 8 ohms or 16 ohms.
This means the amplier will work best whan connected to speakers with
input impedances of 8 or 16 ohms, respectively. The source impedance of the
amplier, however, will be very much lower than the speaker impedance, usu-
ally in the neighborhood of less than a tenth of an ohm. This is to increase the
efciency of the power transfer to the speaker.

496
Bibliography

Some of the books listed here are out of print and may be difcult to obtain,
even in public libraries, but we have included them because of their impor-
tance at their time of publication. Many of them offer valuable practical infor-
mation today, and all are of historical interest.

Books and Articles

Ando, Yoichi. Concert Hall Acoustics. New York: Springer Verlag, 1985.
A very detailed mathematical study of concert-hall acoustics, this book
represents a distillation or summary of Andos considerable research into
the subject. In many ways, it represents the state of the art, but it is not easy
reading. The introduction by Manfred Schroeder is well written and enter-
taining. Andos discussion of the interaural cross-correlation function and
its importance is rst-rate.

Ballou, Glen, ed. Handbook for Sound Engineers. Boston: Focal Press, 2002.
This very large (1,552 pages) book is the replacement for the similar-sized
Audio Cyclopedia of Howard Tremaine, which has been out of print for many
years. The third edition drops the Audio Cyclopedia part of the title. It cov-
ers many subjects, such as acoustical design, transducers, audio electronic
circuits, sound system design, electronic components, and audio measure-
ments. It is primarily for the audio professional but contains plenty of infor-
mation of interest to the hobbyist.

Bksy, Georg Von. Sensory Inhibition. Princeton: Princeton University Press,

1967.
This important compact book by the Nobel-Prize-winner Von Bksy
describes a series of experiments and draws conclusions from them that
apply to many aspects of our hearing. The author is extremely ingenious
in his experimental technique, and his ndings help to explain such phe-
nomena as masking and precedence effect. He also points out many basic
similarities between hearing and our other senses.

Benade, Arthur H. Fundamentals of Musical Acoustics. New York: Oxford Uni-

versity Press, 1976.
This introductory text uses a minimum of mathematics and describes
many acoustic phenomena in somewhat unconventional ways. It covers a
lot of ground but is a little jumbled in its organization, and some explana-
tions are difcult to understand.

497
Bibliography

Blesser, Barry A. Digitization of Audio: A Comprehensive Examination of

Theory, Implementation, and Current Practices. Journal of the Audio Engi-
neering Society, October 1978.
An excellent overview of the pros and cons of PCM, delta modulation,
etc. Blesser has written many good articles in the popular press on digital
audio.

Blesser, Barry; Bart Locanthi; and Thomas Stockham. Digital Audio. Collected
papers from the premier conference of the Audio Engineering Society, June
1982.
A very advanced volume of 270 pages dealing with basics, converters,
measurements, rate conversion, recording formats, error correction, manu-
facturing, and applications.

Bohn, Dennis. Rane Pro Audio Reference. Mukilteo, Wash.: Rane Corp., 2002.
A 320-page large-format paperback with a large glossary and several good
technical tutorials; with CD-ROM.

Burroughs, Lou. Microphones: Design and Application. New York: Sagamore,

1974.
This well-written and interesting book contains a great deal of practical
and useful information about microphones and their use in recording and
broadcasting. The emphasis is on professional-quality dynamic micro-
phones, although other types are also discussed.

Chamberlin, Hal. Musical Applications of Microprocessors. 2d ed. Indianapolis:

Howard Sams & Company, 1985.
This is an expanded and revised version of Chamberlins classic origi-
nal. It is more up to date in its coverage of new technology and describes
commercial music synthesis products in some detail. It also contains pro-
gram listings for doing computerized spectrum analysis.

Colloms, Martin. High Performance Loudspeakers. London: Pentech Press,

1978.
This is a very detailed book on the design and construction of loud-
speakers. It is intended for the professional and knowledgeable engineer
rather than for the beginner, but the casual reader will nd much of inter-
est nevertheless. It is well illustrated and well written, and it covers the sub-
ject with unusual thoroughness.

Eargle, John. Sound Recording. New York: Van Nostrand Reinhold Co., 1976.
A good book covering most aspects of sound recording from a somewhat
theoretical viewpoint, although it does not rely heavily on mathematics.
There is a good chapter on psychoacoustics. Digital recording and signal
processing are not covered.

498
 Bibliography

Eargle, John, ed. Stereophonic Techniques. New York: Audio Engineering Soci-
ety, 1986.
This anthology of reprints of articles on stereophonic techniques is
divided into ve parts: historical papers, analysis and experimentation, stu-
dio technology, broadcasting, and the consumer interface. The historical sec-
tion begins in 1881, which actually seems to be the rst time a stereophonic
sound was transmitted. The writings of many famous people are to be found
here, and it is enlightening to read about key audio developments in the
words of the developers themselves.

Fletcher, Harvey. Speech and Hearing. New York: Van Nostrand, 1929.
This is a classic work on speech and hearing by one of the most impor-
tant investigators in psychoacoustics.

Hall, Donald E. Musical Acoustics, An Introduction. Belmont, Calif.: Wads-

worth, 1980.
A good introductory textbook on the subject that uses mathematical for-
mulas sparingly. There are chapters on physical acoustics, how musical
instruments work, hearing, and sound reproduction.

Helmholtz, Hermann von. On the Sensations of Tone. New York: Dover, 1974.
This is a reprint of an English translation of the monumental German
publication of 1863. Helmholtz was an accomplished scientist and also a
physician, and he performed many truly ingenious experiments in the inves-
tigation of the human hearing mechanism. He was one of the rst to show
that complex sounds are made up of component parts, or harmonics, and
he devised a method for hearing these partials via special resonators of his
design. He also made a mechanical model of the human speech mechanism,
and it was able to intone vowel sounds. Today, the book is of historical inter-
est and is a good source for anyone interested in the scientic method and
its application to basic research. His sections on music and harmony are
very well done.

Huber, David M. and Robert E. Runstein. Modern Recording Techniques. 5th

ed. Boston: Focal Press, 2001.
This is an enlarged and more up-to-date version of Robert E. Runsteins
original Modern Recording Techniques (1974). It covers digital audio tech-
nology quite well.

Isom, Warren Rex, ed. The Phonograph and Sound Recording After 100 Years.
New York: Journal of the Audio Engineering Society, Centennial Issue, Octo-
ber 1977.
This greatly expanded issue of the AES journal commemorates the 100th
anniversary of the invention of the phonograph and is full of historical
information on all aspects of phonograph record production and record-

499
Bibliography

ing. The section on the manufacturing of records is particularly complete.

There is an interesting article on digital audio by Tom Stockham.

Langford-Smith, F. Radiotron Designers Handbook. Harrison, N.J.: Radio Cor-

poration of America, 1953.
This is a comprehensive handbook (over 1,400 pages), primarily on the
applications of vacuum tubes to audio. Although it may seem outdated,
there is a great deal of useful information here. Such subjects as equaliz-
ers, transformers, harmonic analysis, tuned circuits, distortion, loud-
speakers, oscillators, etc., are treated in detail. Lately, there has been an
increased interest in the use of tubes for audio, and this work is the most
complete we know of on this subject.

Mathews, Max V. The Technology of Computer Music. Cambridge, Mass.: MIT

Press, 1969.
A good overview of the generation of computer-realized and computer-
composed music.

Miyaoka, Senri. Digital Audio Is Compact and Rugged. IEEE Spectrum,

March 1984.
A short but precise description of the optical aspects of the compact disc
concept, by the leader of the development staff at Sony.

Olson, Harry F. Modern Sound Reproduction. 1972. Reprint, New York: Robert
E. Krieger, 1978.
An interesting book by one of the giants of electroacoustics, who headed
the RCA Research Laboratories for many years. The book is outdated now,
but contains much historical information.

Philips Technical Review. Vol. 40, No. 6, 1982, published by Philips in the
Netherlands.
A very good and extensive (180 pages) general description of the tech-
nical aspects of the compact disc format.

Pierce, John R. Almost All about Waves. Cambridge, Mass.: MIT Press, 1974.
While not specically directed toward audio, this well-written little
book describes waves in a good deal more detail than almost any other. It
is not for the beginner, requiring as it does an understanding of mathematics
through calculus, but it offers the careful reader a most clear and penetrating
insight into wave motion, vector and complex notation, polarization, dif-
fraction, and radiation, etc.

Pierce, John R. The Science of Musical Sound. New York: Scientic American
Library, 1983.
This is a fascinating book from John Pierce, who consistently addresses
technical subjects with fresh insight and penetrating powers of observa-

500
 Bibliography

tion. The book comes with two small phonograph records that illustrate
various topics that are discussed.
His chapter on sound reproduction has some interesting history, but it
is not very complete. The sections on the ear and hearing are detailed and
rst-rate, and are written in a way that makes this difcult subject easy to
understand.

Pohlmann, Ken C. Principles of Digital Audio. 4th ed. New York: McGraw-
Hill Professional, 2000
This is one of the rst popular books on digital audio by a well-known
columnist on the subject. Pohlmann writes in an amusing and easy-to-read
style, and the book is well illustrated. The rst chapter, on the basics of sound
and acoustics, has some inaccuracies, but the rest of the book is very inform-
ative and quite complete.

Read, Oliver. The Recording and Reproduction of Sound. 2d ed. Indianapolis:

Howard W. Sams, 1952.
This classic volume, even though not up to date, contains a wealth of infor-
mation on all aspects of audio, and it makes fascinating reading today. It
is well illustrated and covers the entire subject with ne detail.

Read, Oliver, and Walter L. Welch. From Tin Foil to Stereo. 2d ed. Indianapo-
lis: Howard W. Sams, 1976.
This fact-lled book is the most complete history of the acoustic phono-
graph from its invention by Edison up through the development of stereo-
phonic recording. It is biased heavily toward acoustic recording, and
perhaps reveres Edison unduly, but it is a fascinating anecdotal history of
the many events, inventions, patent ghts, etc., that constituted the devel-
opment of audio recording and reproduction.

Rossing, Thomas D. The Science of Sound. Reading, Mass.: Addison-Wesley,

1982.
This is a ne if conventional text on musical acoustics, covering percep-
tion and measurement of sound, how musical instruments work, etc. Its
best features are the chapters on electrical production of sound, which cover
electronic circuitry, and a fair amount of detail on audio devices. Electronic
music and digital sound are also covered.

Schubert, Earl D., ed. Psychological Acoustics. Stroudsburg, Penn.: Dowden,

Hutchinson, and Ross, 1979.
A collection of previously published papers, providing a fascinating
chronology of major discoveries from 1876 to 1970.

Sinclair, Ian R., ed. Audio Electronics Reference Book. Boston: BSP Professional
Books, 1989.
This interesting technical book from Britain includes 20 chapters written

501
Bibliography

by various experts in their respective elds and contains a wealth of infor-

mation on most aspects of audio electronics. Emphasis is given to British
audio products and developments. It also covers elementary acoustics and
psychoacoustics as related to sound reproduction. The compact disc sys-
tem and tape recording are well represented, as are loudspeakers, loud-
speaker enclosures, and automobile audio.

Woram, John M. The Recording Studio Handbook. Commack, N.Y.: Elar Pub-
lishing Co., 1983.
An interesting, quite well-illustrated book covering most aspects of
recording studio operations. It is detailed and accurate in its descriptions
of audio equipment, although it does not cover digital equipment.

Periodicals

Audio Media. IMAS Publishing Group, 5827 Columbia Pike, Third Floor, Falls
Church, VA 22041 (monthly).
A magazine with several world region editions, it aims for sound pro-
fessionals in various elds, especially recording studios. Many reviews of
equipment, facility showcases, and general articles of audio production top-
ics. It is free to qualied individuals.

EQ. United Entertainment Media, P.O. Box 0532, Baldwin, NY 11510-0532

(monthly).
A recording magazine suited for the project studio owner and record-
ing musician, it features articles on recording and studio techniques,
equipment reviews and several well-known columnists. It is free to qual-
ied individuals.

Journal of the Audio Engineering Society (JAES). 80 East 42nd Street, New York,
NY 10165 (monthly).
A technical journal, containing detailed articles on all aspects of audio
engineering. Many of the articles are manuscripts of papers presented at
the semi-annual Audio Engineering Society conventions and represent the
most up-to-date information available. The journal is intended for the pro-
fessional in audio, and the articles are on an advanced technical level, with
a good deal of mathematics included. The AES also publishes occasional
anthologies or collections of papers on various subjects such as loudspeaker
design and digital audio. The journal is included with paid membership
in the AES.

Mix. Mix Publications, Inc., 6400 Hollis Street, Suite 12, Emeryville, CA 94608
(monthly).
This relatively large magazine is somewhat similar to RE/P and db, but
reects a more popular approach. It includes semi-technical articles on most

502
 Bibliography

aspects of audio and studio techniques, some articles on video, and reviews
of studio equipment. Like S&VC, it is heavily supported by advertising. It
is free to qualied individuals.

Sound & Video Contractor (S&VC). Primedia, Inc., 6400 Hollis Street, Suite 12,
Emeryville, CA 94608 (monthly).
Distributed free to qualied professionals in sound reinforcement and
closed-circuit video system contracting. There are articles on theater sound,
signal processing equipment and techniques, digital equipment for sound
reinforcement and semi-technical articles on the operation of audio devices
such as loudspeakers. The magazine is heavily supported by advertisers.

Periodicals No Longer Published

The following periodicals all ceased publication since the last edition. Many
issues contained articles of merit and still make good reading when taken in
historical context

Audio. 1515 Broadway, New York, NY (monthly).

This was one of the oldest magazines devoted to audio, and was aimed
at the knowledgeable consumer. It contained semi-technical articles on a
wide range of subjects of interest to the stereo enthusiast, and also contained
reviews of consumer audio equipment as well as reviews of records, tapes,
and compact discs. One of its most endearing features was the monthly arti-
cle by the late Edward Tatnall Canby, which ran continuously since the early
1950s.

db, The Sound Engineering Magazine. Sagamore Publishing Co., Plainview, L.I.,
NY (bi-monthly).
An interesting magazine for the recording engineer and studio operator,
including reviews of professional and semi-professional audio products.
The continuing column on digital audio by Barry Blesser was very well writ-
ten and informative.

Recording Engineer/Producer (RE/P). P.O. Box 2449, Hollywood, CA (monthly).

This semi-technical magazine was aimed at the recording studio opera-
tor and engineer, and also had articles of interest to the recording musi-
cian. Articles covered a wide range of subjects, from historical pieces to
digital recording and motion picture sound. The emphasis was always on
practicality, and much useful information is included. A typical issue ran
to over 200 pages.

Studio Sound and Broadcast Engineering. Link House Publications, Croydon,

England (monthly).
This publication was available free to audio professionals, and was con-

503
Bibliography

sidered very important. Although published in England, it was still rele-

vant to American readers interested in all aspects of audio, with emphasis
on the recording studio and its operation. One of its most valuable features
was the equipment review section, where professional audio devices were
subjected to extensive testing and reporting. They included high-level arti-
cles on new and little-known techniques, and the letters to the editor make
interesting reading. It seemed the British have raised the writing of such
letters to an art form, and it was well represented here.

The Internet
One should approach information on the internet with caution, as much advice
is ill-informed. One should seriously evaluate the authority of the source of
information. It is, however, a practical form of give and take communication.

Usenet Newsgroup Rec.Audio.Pro, also known as RAP

An unmoderated newsgroup where participants post questions and
replies, and chat. Read the FAQ (Frequently Asked Questions) which gen-
erally answers many beginners questions about audio production rea-
sonably well. The noise or useless chatter can be daunting, but many
front-line audio workers post questions and answers. Very few well-known
authorities make their presence, if any, known.

Live Audio Board (LAB) at <www.prosoundweb.com/forums> or <www

.live-audio.com>
Afavorite of many in the sound reinforcement eld. The prosoundweb.com
site has other forums for recording and stage lighting. There is advertising
to support the site, but no fees for users.

504