Optical Data Storage

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/234963135
Optical Data Storage
Article in Proceedings of SPIE - The International Society for Optical Engineering · July 1991
DOI: 10.1117/12.45900
CITATIONS READS
6 106
5 authors, including:
Mark Wang
KLA Corporation
12 PUBLICATIONS 44 CITATIONS
SEE PROFILE
All content following this page was uploaded by Mark Wang on 19 October 2023.
The user has requested enhancement of the downloaded file.

Optical Data Storage
Optical data storage is found in popular consumer products. Compact Discs (CDs),
Digital Versatile Discs (DVDs), and MiniDiscs (MDs), are all forms of optical data storage.
More advanced forms of optical data storage include high-speed devices and library products.
All optical data storage devices use optical principles to achieve high data density, rugged
packaging, reliable information retrieval, and cost-effective production.
In general, optical data storage relates to placing information on a surface so that, when a
light beam scans the surface, the reflected light can be used to recover the information. There
are many forms of optical storage media and many types of optical systems used to scan data.
This chapter discusses the basic principles of optical data storage, types of commercial optical
media available in 2002, several performance parameters and some interesting prospects for
future systems.
1.0 Inspiration for the Invention
The optical disk was envisioned in 1958 by an eclectic engineer named David Paul
Gregg, who patented the idea in 1961.[1,2,3] At the time, magnetic disks were in their infancy,
and there was no low-cost “videodisk” system that could be marketed as a consumer product. As
told by the inventor, inspiration for the videodisk system came from the picture shown in Fig. 1,
which is a photograph from a trade magazine in the mid 1950’s. The photograph was produced
from an early scanning electron microscope. The narrow lines in the lower-left corner of the
Author: Tom D. Milster (Prepared for Laser Handbook) 1

©2002 Tom D. Milster
picture, which were written by the electron beam, are 0.030 micrometer wide and are separated
by 0.070 micrometer.1 Gregg saw the picture and, while on horseback in Mexico, imagined a
rotating plastic disk with tracks of data marks read by an inexpensive optical system. As
originally envisioned, a master disk is first written with an electron beam. The master disk is
used to produce low-cost replicas, which are then sold to consumers.
In many ways, Gregg was prophetic. The CD, a mature development and improvement
of Gregg’s invention, is an immensely successful consumer product. DVD systems are advanced
versions of CDs. Today’s most advanced systems, which use blue laser diode optical systems,
often utilize electron-beam machines to make master disks. Someday, it may be possible to
record information with the 30 nm linewidth observed in Fig. 1, which would result in a disk
with data density several thousand times the data density found on DVDs.
2.0 Data Marks and Spaces: The Information Carriers
Digital information [xref?] is stored on optical disks in the form of arrangements of data
marks in spiral tracks. Small sections of CD and DVD surfaces are illustrated in Fig. 2, which
also displays representations of laser spots that are focused on the surfaces to write and read data.
CDs typically use 1.6 micrometer track pitch, which is the radial distance between turns of the
spiral tracks. Data marks are nearly one-half as wide as the track pitch. Lengths of data marks
1
For a sense of reference, if we were to magnify a 1.0 micrometer wide line to the length one
yard, edge thickness of a piece of paper under the same magnification would be about the length
of a football field. The lines in Fig. 1 would be about one inch wide. (1 micrometer = 10-6 m =
0.000039 in)
and spacings between marks are determined by the encoding scheme [4,5] used to translate user
data into mark patterns along each track, which is described in more detail in Section 5.8. Width
of the CD laser spot is slightly smaller than the track pitch. DVD media are similar to CD
media, except track pitch is smaller (0.74 micrometers), data marks are shorter and narrower, and
the laser spot diameter s is smaller. Since there are more data marks per unit area on a DVD
compared to a CD, the DVD can hold more data.
3.0 Optical Data Storage Principles
Storage and retrieval of data on optical disks can be described in two simple steps. First,
data marks are recorded on a surface. Data marks can be prerecorded, like on a music CD, or
they can be recorded by users on blank disks, like with CD-recordable products. The second step
is retrieval of information from the disk, where a light beam scans the surface. Modulation in the
reflected light is used to detect the data-mark pattern under the scanning spot.
The process for exposing data marks on a recordable optical disk is shown in Fig. 3,
where an input stream of digital information is converted with an encoder and modulator into a
drive signal for a laser source. The laser source emits an intense light beam that is directed and
focused onto the surface with illumination optics. As the surface moves under the scanning spot,
energy from the intense scan spot is absorbed, and a small, localized region heats up. The
surface, under the influence of heat beyond a critical writing threshold, changes its reflective

properties. Modulation of the intense light beam is synchronous with the drive signal, so a
circular track of data marks is formed as the surface rotates. The scan spot is moved slightly as
the surface rotates to allow another track to be written on new media during the next revolution.
Data marks on prerecorded disks are fabricated by first making a master disk with the
appropriate data-mark pattern. Masters for prerecorded CDs and DVDs are often exposed in a
similar manner to exposing data marks on recordable optical disks, except that the light-sensitive
layer is designed to produce pits in the master that serve as data marks in the replicas.
Inexpensive replicas of the master are made with injection-molding equipment.
Readout of data marks on the disk is illustrated in Fig. 4, where the laser is used at a
constant output power level that does not heat the data surface beyond its thermal writing
threshold. The laser beam is directed through a beam splitter into the illumination optics, where
the beam is focused onto the surface. As the data marks to be read pass under the scan spot, the
reflected light is modulated. Modulated light is collected by illumination optics and directed by
the beam splitter to servo and data optics, which converge the light onto detectors. The detectors
change light modulation into current modulation that is amplified and decoded to produce the
output data stream.
4.0 Commercial media
There are several types of optical disks, which can be differentiated by the type of data
marks on the recording layer. The most popular disks are based on pit-type, magneto-optic,

phase-change and dye-polymer data mark technologies. Several commercial trade names are
associated with the four technologies, as shown in Table I. In this section, basic data-mark
technologies are reviewed, and commercial formats are listed.
Pit-type data mark technology for CD-read-only-memory (ROM) and DVD-ROM
products is based on a very simple scattering phenomenon. Most CDs, like music and data
distribution CDs, are ROM disks. The small data-mark pits are arranged in spiral tracks around
the center of the disk, as shown in Fig. 2. The pit lengths are about one micrometer to three
micrometers long. The widths of the pits along a track are nearly uniform and measure about
one-half to eight-tenths of a micrometer. As the light spot passes over a pit, most of the reflected
light scatters away from the illumination optics. The remaining light collected by the objective
lens is small compared to the amount of light that gets collected when the spot is over a smooth
portion of the track, where the disk surface acts as a mirror to the focused light. The data signal
is derived from the detector that senses the amount of collected light, as shown in Fig. 4.
The amount of light scattered from each pit depends on the depth of the pit and the size of
the laser beam illuminating it. A simplistic explanation is that the portion of the laser spot
reflected from the pit exhibits a phase change due to the additional path that the light must
traverse compared to the portion of the laser spot that is reflected from the surrounding flat area
of the recording surface. The two portions of the spot interfere destructively [xref:?] upon
propagation of the light back to the objective lens if the effective depth of the pit is one-eighth of
the illuminating laser wavelength. In practice, the pit depth profile is designed to not only
provide good data signal modulation, but also good tracking performance, as explained in
Section 5.6, which is optimized at a slightly different pit depth.[6]

In order to increase the amount of light reflected to the detector, the entire recording
surface of the CD, including both pits and areas between pits, is coated with aluminum. The
aluminum is then coated with a lacquer or other protective layer, onto which the label is printed.
The read out optical system focuses light through the clear surface of the disk, as shown Fig. 5,
so the data mark patterns and track are actually located nearer the label than the clear side
through which the user views the rainbow diffraction pattern [xref:] that forms due to the close
radial spacing of the tracks. The plastic CD substrate is 1.2 mm thick, which is designed so that
contamination, like fingerprints or scratches, on the surface of the disk does not adversely affect
disk performance during read out.[7] DVD substrates are only 0.6 mm thick, which implies that
DVDs may be more sensitive to contamination than CDs.
Dye-polymer or dye-monomer technology is used in CD-R products. Dye
polymers/monomers are organic films that are ablated to form pits along tracks.[8, 9, 10] To
form a pit, a high-power focused spot locally heats a micron-sized area. The dye polymer
absorbs a large percentage of the laser energy. Due to the low thermal conductivity of dye
polymers, extremely high temperatures can be reached. Although the exact mechanism of pit
formation in CD-R is not known, a simple explanation is that, in the heated area, the dye material
is vaporized or heated to the point that material flows to form a pit. To read data, a low-power
laser beam scans the track, and the collected light is sensed with a simple detector. The collected
light is modulated by light scattering, which is similar to the phenomenon described above for
pit-type media. Since the recording process for CD-R’s is destructive, the user can only write
data marks once. Data marks cannot be erased and rewritten.

There are three classifications of dyes used to make CD-Rs, which are cyanine,
metallized azo, and phthalocyanine. At a laser wavelength to 0.78 micrometers, where CD
players are designed to operate, there are only slight differences in performance between the
dyes during the writing process. All the dyes absorb laser light and heat the recording surface.
In addition to the dye layer, CD-Rs have a reflective layer, like the CD-ROM. However, the
reflective layer is usually silver or gold instead of aluminum.
Different combinations of dyes and reflective layers influence visual appearance of the
CD when viewed from the clear side of the substrate. Since the visible spectrum [xref:] is
shorter wavelength than the recording laser wavelength for CDs, the dye usually appears with a
characteristic semitransparent color. The reflective layer appears either silver or gold. Table II
lists some combinations of dyes, reflectors, and the resulting synthesized color to the eye. For
example, the green CD-R, the cheapest of combination, uses the cyanine dye.[11] By itself, the
cyanine dye is blue in color, but, together with the gold reflector, the recording surface appears
green. Cyanine’s ability to maintain reflectivity is poor, which gives it an expected lifetime of
only about 10 years. Improved-formula cyanine dyes in combination with silver reflectors (blue
synthesized color) have shown better performance, which is better than twenty-year lifetime after
recording. The gold-colored CD-R uses a phthalocyanine dye and a gold reflector. The dye is
transparent by itself, so the gold color shines through. Modulation in the reflected light caused
by writing on the gold medium is the best of all CD-R media, and lifetime of such CD-R’s is said
to be over 100 years.[12] Blue media are made of azo dyes. Like cyanine, the azo dye is blue,
but azo disks use a silver reflector, which result in a blue synthesized color.

CD-RW products use a different recording layer material than CD-R’s. In addition to
allowing the user to write data marks, data marks can also be erased with multiple cycles before
degradation. CD-RW’s use phase-change technology, which is based on differences of the
crystalline and amorphous states of semi-metal alloys, like AgInSbTe or GeSbTe.[13, 14, 15] To
record or erase phase-change data marks, a high-power focused spot locally melts the medium in
micron-sized regions as the disk spins. The thermal cycle of the local regions determines if the
region will stabilize in a crystalline or amorphous state. By controlling the energy in the focused
spot, the thermal cycle and the state of the material can be controlled. For example, a high
power laser pulse and rapid cooling quenches the material into the amorphous state, as shown in
Fig. 6 when the laser is pulsed to the peak power level. A lower-power laser beam and slow
cooling anneals the material into a crystalline state. Usually, marks are in the amorphous state
and the background is in the crystalline state. CD-RW media is a “write dark” media, which
means that the amorphous state of the data marks does not reflect as much light as the crystalline
background. Some commercial media are “write bright”, in that the recording layer is initially in
the amorphous state, and bright crystalline marks are written on the dark amorphous
background.[16] In Fig. 6(a), a virgin track is exposed to a certain mark pattern. Three clearly
defined, dark amorphous marks are formed along the track, where one each data mark
corresponds to a peak-power laser pulse. The bias power ensures that regions between marks
anneal into the high-reflectivity crystalline state. When new data are written on the track, as
shown in Fig. 6(b), a similar sequence of laser pulses are used, except that the laser pulses
correspond to the new data-mark pattern. Old data marks are overwritten and replaced with
either crystalline or amorphous material of the new pattern. The phase-change process inevitably
involves a mechanical deformation of the material. Therefore, the number of direct overwrite
cycles is limited to several thousand. Like a CD player, the detector in a CD-RW player simply
senses the amount of collected light. The data signal is derived from the detector current.
The readout signal contrast is optimized in a CD-RW player by adding several thin-film
layers around the recording material. The effect of the layers is to produce a thin-film reflective
filter [xref:?]. A typical configuration of the storage layers in CD-RW products is shown in Fig.
7. The laser beam focuses through the clear substrate material and into the thin-film recording
layers. The first layer is a transparent dielectric. The second layer is the recording material, and
the third layer is another layer of transparent dielectric. The fourth layer is a reflector. The
thicknesses of these four layers are designed to tune the reflective properties for maximum signal
contrast, and they are also adjusted to provide adequate absorption so that a reasonable amount
of laser power can be used for writing.[17] A protective coating and a label and ink layer
separate the thin films from the environment.
Magneto-optic (MO) products store information in small magnetic data marks, which are
about the same size as pits on a CD. The recording layer is initially erased so that all magnetic
domains are aligned in one direction perpendicular to the recording surface, as shown in Fig.
8(a). In this configuration, the magnetic domains are extremely stable. A large magnetic field of
several thousand oresteds [xref:?] is required to overcome the magnetic moment of the domains.
The magnetic field required to reorient domains is called the coercivity.[xref:?][18] To record
data marks, a high-power focused spot is used to locally heat the recording surface. Heat reduces
coercivity, so magnetic domains in the region of the focus spot can be reoriented with an external
magnetic field. When the laser beam is switched to low power between data marks, the
recording layer is not heated, and the eternal magnetic field has no effect on domain orientation.

As described in Fig. 3, the laser beam is modulated between high power and low power as the
disk spins in order to write a pattern of data marks along each track. Each mark contains
magnetic domains oriented in the opposite direction compared to the magnetic domains of the
background.[19] The marks have the property that, as a low-power focused light spot passes over
it, polarization [xref: polarization] of the reflected light is rotated, as shown in Fig. 8(b).
Polarization rotation on reflection is due to the polar Kerr effect.[20] When the laser beam
illuminates a data mark with domains oriented away from the laser beam, linear incident
polarization is rotated slightly in the counter-clockwise direction. When the laser beam
illuminates the region between data marks, linear incident polarization is rotated slightly in the
clockwise direction. In order to detect the data signal, a detector is used to sense change in
polarization of the reflected light. For example, an indication that the reflected light is rotated in
the counter-clockwise direction implies that the laser spot illuminates a data mark. In order to
erase data, the external magnetic field in Fig. 8(a) is reversed, and the laser beam heats an entire
section of the track. A major difference between CD and MO products is that the MO marks are
produced in a track with an almost undetectable change in the topology of the track. That is,
there is almost no mechanical deformation of the track as the marks are recorded or erased. This
property enables MO products to exhibit over one million erase cycles with little if any
degradation in performance.[21, 22]
A collection of the available CD and CD-like formats are listed in Table III, along with
their associated data-mark technology.[23] DVDs used for movie or data distribution are pit-
type ROM disks. The tack pitch and pit size is smaller than in CDs, as shown in Fig. 2. DVD
products can also be erasable, and there are a multitude of formats available, as listed in Table

IV. These products use erasable-change technology. DVD-R products, like CD-R, use a write-
once dye polymer recording layer. Unlike CDs, DVDs can use more than one storage layer per
disk. DVDs can be double sided, use two layers on one side, or use two layers on each side.
Adding layers increases the total capacity of the disk.
5.0 Technology
Several important aspects of optical data storage technology are associated with the
optical-mechanical-electrical system that is used to write and read data to the disk. This section
reviews basic concepts necessary to understand how these systems work.
5.1 Data density and spot size
Capacity of an optical disk is determined by its data density, which is the number of bits
of information stored per unit area on the surface, and the recording area. Data density is often
specified in gigabits (109 bits) per square inch of recording surface area (Gb-in-2 ). For example,
a 0.65 gigabyte (GB) CD has a recording area of about 14.5 square inches, so the data density is
(0.65)(8)/14.5 = 0.36Gb-in-2 , where one byte = 8 bits.
A fundamental limitation to the data density is due to the size of the focused laser beam
that illuminates the surface. Figure 7 shows a detailed picture of the laser irradiance approaching
the surface, where irradiance is defined as the laser power per unit area. Ideally, maximum
irradiance is located at the recording material, along with the smallest spot size s. As the
distance increases away from the ideal focus, the spot size increases and the peak irradiance
decreases. A defocus distance ∆z of only 3 micrometers dramatically reduces peak irradiance
and increases spot size. An approximate formula used to estimate the ideal spot size is s = λ/(sin
θ), where θ is the marginal ray angle of the illumination optics, as shown in Fig. 3. Spot size s is
the full width of the irradiance distribution at the 1/e2 (13.5%) irradiance level relative to the
peak. The value of sin θ is often called the numerical aperture or NA of the optical
system.[xref:?] CD systems exhibit λ = 0.78 micrometers and NA = 0.47, which produce a spot
size of 1.7 micrometers. DVD systems exhibit λ = 0.65 micrometers and NA = 0.60, which
produce a spot size of 1.1 micrometers.
5.2 Thermal recording
In order to write data onto the spinning disk, the laser must be pulsed to a high power
level. The time duration of the high-power pulse determines the length of the data mark that is
written onto the surface. Laser writing is possible because the medium is thermally sensitive.
That is, the medium exhibits a thermal threshold.[8] Below the threshold, medium properties do
not change significantly. Above the threshold, a physical change occurs in the medium.
Figure 9 shows lines of constant temperature, which are called isotherms, generated on an
aluminum surface for a 200 nano-second (200x10-9 sec) focused laser pulse. The surface is
moving at 10 meters per second, so the isotherms are spread out along the scan direction. This

profile is representative of those found in DVD optical disks. Notice that the end of the pulse
generates a wider isotherm than at the beginning of the pulse, due to the fact that heat builds up
and spreads out in the direction perpendicular to the scan. This effect is called thermal
blooming, and is a serious problem, especially in mageto-optic systems, if not corrected by
varying the properties of the laser pulse.[24] Fig. 9 indicates that, if the threshold temperature of
the medium is equal to the 200o C isotherm, a data mark of approximately 0.6 micrometers wide
and 2.5 micrometers long will be written at this location on the surface of the disk.
5.3 Frequency response and equalization
The ultimate limit to the size of the data marks on the recording surface is determined by
the frequency response of the optical system. Spatial frequency is 1/T, where T is the period of
the data-mark pattern. As the period decreases, spatial frequency increases. The frequency
response can be understood simply by recognizing the behavior of the reflected light and how the
reflected light is collected by the objective lens. For example, Fig. 10 shows the reflected light
distribution for a periodic pattern of data marks along a track. The reflected light consists of
three cones. The direct reflection is the central cone. The two outer cones are called diffracted
orders. They are very similar to the central cone in appearance, but they are spread apart by
angle ψ. As T decreases, ψ increases, and the diffracted orders spread more widely apart. ψ is
also inversely proportional to the laser wavelength. Shorter-wavelength lasers exhibit smaller ψ.
When the spot scans over data marks, the optical phase of each diffracted order changes,
but the phase of the central cone does not change. The phase difference between the diffracted
orders and the central cone produces a modulation in the overlap area due to interference. That
is, as the spot scans over data marks, the overlap areas get brighter and darker as a function of
the relative position between the spot and each mark. Brightness of the central cone does not
vary. Therefore, the contrast of the signal modulation received at the detectors is determined by
the amount of overlap area. More overlap area produces a higher contrast data signal. As T
decreases, so does the overlap area. At some critical mark period, there is no overlap and,
consequently, no signal modulation at the detector. This critical mark period is called the
resolution limit TR of the optical system. A numerical value for TR is found from
λ s
TR = = ,
2 NA 2
where NA is the numerical aperture of the objective lens and s is the spot size.
Figure 11 shows the modulation transfer function for the optical system,[xref:?] which
plots the contrast of the current signal modulation versus the spatial frequency of the data-mark
pattern. Maximum modulation is observed for long marks. The contrast reduces gradually to the
resolution limit. The maximum mark frequency in practical devices is well above the resolution
limit. For example, the shortest mark period in CDs is about 1.8 micrometers, which is about a
factor of two longer than the resolution limit of 0.85 micrometers. Of course, real data patterns
are more complicated than simple periodic patterns, but each real data pattern can be
decomposed into a collection of weighted periodic patterns. Therefore, the modulation transfer
function is also useful in describing system behavior for real data patterns.
When the real data pattern contains both high-frequency and low-frequency components,
a significant contrast difference exists in the current signal. These differences in contrast make

detecting signal transitions difficult. In order to minimize the contrast difference, electronic
circuits are often employed during signal amplification, as shown in Fig. 4. The electronic
circuits partially equalize the modulation transfer function and provide more reliable signal
decoding.[25] An example of an equalized modulation transfer function is shown in Fig. 11,
where the high-frequency contrast is boosted with respect to the low-frequency contrast.
Unfortunately, physical limitations of electronic circuitry and noise considerations do not permit
ideal equalization, which would exhibit uniform contrast for all spatial frequencies out to the
resolution limit.
5.4 Effects of defocus
An example of the effects of defocus is shown in Fig. 12, where irradiance of a DVD-like
system with NA = 0.6 and λ = 0.65 micrometers is plotted for several values of ∆z. In focus at
∆z = 0, the spot is well confined and only a small fraction of the spot energy is contained in the
diffraction rings surrounding the central lobe. At ∆z = 0.5 micrometers, the peak irradiance has
reduced slightly and a small amount of energy has shifted to the first diffraction ring closest to
the central lobe. At ∆z = 0.5 micrometers, changes in the spot shape will not dramatically affect
device performance. However, as defocus increases beyond ∆z = 0.5 micrometers, peak
irradiance degrades rapidly and a significant amount of energy is spread into the diffraction
rings. An estimate of the allowable depth of focus is ∆z = +/- λn/(4 sin2 θ) = +/- λn/(4NA2 ),
where n is the refractive index of the disk substrate.[xref:?] For example, with n = 1.5, ∆z = +/-

0.67 micrometers. The effect of defocus on an unequalized modulation transfer function is
shown in Fig. 11, where the mid-frequency response of the system suffers severe degradation.
5.5 Servo optics
The size of the focused spot is very small in the direction along the tracks, which allows
many data marks to be written for each revolution of the disk. Since the light spot is also small
in the direction perpendicular to the track, tracks can be spaced closely together. In CD systems,
track pitch, where pitch is defined as the center-to-center track spacing in the radial direction, is
typically 1.6 micrometers. In DVD systems, track pitch is 0.74 micrometers. Data-mark width
is typically less than one half the track pitch in order to reduce the effects of crosstalk from
marks on adjacent tracks. The optical spot must be centered over the marks as the disk spins in
order to obtain maximum signal amplitude at the decoding electronics. A typical requirement is
that the spot must be kept on track center to better than one-tenth the track pitch, or 0.16
micrometers for CD systems and 0.07 microns for DVD systems. In the focus direction, the spot
must be controlled to better than one-tenth the depth of focus, which is about 0.25 micrometers
for CDs and 0.13 micrometers for DVDs.[26, 27]
This demanding control of the spot center and focus position is complicated by the fact
that the optical disk and the electric motor that rotates the disk suffer from loose tolerances that
induce large variations in the track position as the disk spins.[28] For example, thickness of CDs
can vary by more than 50 micrometers. Registration errors during the molding process can offset

the center of the track radii by more than 30 micrometers from the center of the disk rotation.
Wobble of the motor shaft can induce variations in the focus position by several hundred
micrometers.
In commercial optical data storage systems, position control is accomplished with closed-
loop feedback servos. A basic diagram that illustrates the servo technique is shown in Fig. 13.
The difference between the desired spot position and an error signal that is derived from the
actual spot position is amplified with some gain G and used as input to an actuator H.[29] The
actuator is usually a mechanical device, like a voice coil, which moves an optical element that, in
turn, repositions the spot in either the focus direction or across the tracks. The spot position is
instantly determined by the feedback sensor, and the new information is fed back into the control
loop.
Both focus and tracking actuators are usually combined into one mechanical unit that
moves the objective lens. A photograph of an actuator assembly from a commercial CD player
is shown in Fig. 14. The objective lens is mounted in a suspension that has a range of motion of
a few millimeters in both the focus and tracking directions. The flexure of the suspension is very
stiff with respect to motion in any other direction. The fixed part of the suspension also has
permanent magnets mounted on it that are aligned with electric coils on the moving part of the
actuator assembly. As electric current is passed through the coils, the induced magnetic field
presents a force on the permanent magnets and moves the suspension. If the tracking coils are
activated, the suspension moves in the cross-track direction. If the focus coils are activated, the
suspension moves in the focus direction. By moving the suspended objective lens, position of
the focus spot on the disk can be changed. When the actuators are combined with a servo loop,

accurate control of the spot position is possible, even in extreme environments like those found
in portable disk players.
5.6 Feedback sensors
An important part of the servo loop pictured in Fig. 13 is the feedback sensor. In fact, it
is not possible to control the spot position better than the sensor can detect position errors.
Usually, separate tracking and focus sensors are implemented in optical data storage devices.
The error signal generated from the tracking sensor is called the tracking error signal (TES), and
the error signal generated from the focus sensor is called the focus error signal (FES).[7] The
following TES and FES sensors are described:
Push-pull tracking is a method to provide a tracking error signal using a groove
pattern on the disk, where the period of the grooves is equal to the
track pitch.
Three-spot tracking is a method to provide a tracking error signal by detecting the
signals from three spots focused onto the disk, where the central
spot is centered over a track and two neighboring spots are slightly
each side of the central spot in the direction across the track.
Astigmatic focusing is a method to provide a focus error signal by fabricating a small
amount of astigmatism into the servo optics.

Generation of a TES signal using grooves is shown in Fig. 15. In addition to the data
marks, each track contains a land and groove area. Data marks are usually written in the land
areas, which are closer to the objective lens than the grooves. The collection of lands and
grooves forms a diffraction grating in the cross-track direction. Like with single-frequency data-
mark patterns, the land/groove diffraction grating produces separated cones in the reflected light
due to diffraction. When the spot is centered on the groove, the phase of each diffracted order is
equal, so the overlap areas are of equal brightness. When the spot is off center, the phase of the
diffracted orders change, and brightness of the overlap regions become unbalanced. This
brightness asymmetry is detected with a split-cell detector. Current signals A and B are
subtracted to form the TES. When the spot moves in one direction off center, the total power on
detector A becomes brighter than the total power on detector B. The TES signal is positive. As
the spot continues to move in the same direction, the detector signals become more unbalanced,
which creates a more positive TES. If the spot moves in the opposite direction, the detectors
become unbalanced in the opposite sense, which creates a negative TES. Near the center of the
track, the TES is linear and provides a good quality feedback signal that is directly proportional
to the position error. The TES is periodic with a period equal to the track pitch. This type of
TES signal is called push-pull tracking, which is descriptive of the light-pattern behavior on the
detectors as the spot moves off track center.[30] Push-pull tracking is often used in CD-R and
CD-RW players, where some form of tracking reference is necessary before data can be written.
Both regular data-mark patterns and grooves produce diffracted orders in the reflected
light. These orders overlap at the objective lens, as shown in Fig. 16. Diffracted orders from the

data-mark patterns spread in the direction parallel to the scan direction, and the amount of spread
is proportional to the data-mark frequency. Diffracted orders from the grooves spread in the
direction perpendicular to the scan direction, and their separation is constant. Modulation is
observed in the overlap areas between the data-mark orders and the objective lens as the spot
scans along the track. In addition to the modulation due to the data pattern, brightness changes
can be observed in the overlap areas between the groove orders and the objective lens as the spot
moves off track.
A second method used to generate a TES is shown in Fig. 17, which is called three-spot
tracking.[31] In addition to a central laser spot, two additional laser spots are generated by the
illumination optics. These three spots are imaged onto separate detectors A, B and C by the
servo optics. The leading spot is imaged onto detector A, and the trailing spot is imaged onto
detector B. The central spot, which is used to detect the data signal, is imaged onto detector C.
The brightness of the spots at the detectors is determined by the amount of overlap between the
spot and the data marks. When the spot is centered over a data mark, its corresponding light
level at the detectors is reduced the most. When the data spot is centered over a track, the
leading and trailing spots are slightly offset from the center in opposite directions and by an
equal amount. Their brightness at the detectors is equal, and the difference between detector
signals A and B is zero. When the spots are slightly off track, as shown in Fig. 17, the leading
spot is now overlapping less of the data mark and the trailing spot is overlapping more of the
data mark. Therefore, the brightness on detector A increases, and the brightness on detector B
decreases. The difference between detector currents A and B is now positive. When the spot on
the disk shifts in the opposite direction, the difference current becomes negative. The TES is

generated from the difference between the currents from detectors A and B, and provides a good
quality feedback signal that is directly proportional to the position error in the center of the track.
Like with push-pull tracking, the TES for three-spot tracking is periodic with a period equal to
the track pitch. Unlike the push-pull technique, three-spot tracking requires that the spots be re-
imaged onto the detectors. Three-spot tracking is often used in music CD players, where there is
not a land and groove pattern and the TES must be generated from only the data marks.
The popular astigmatic focus method to generate a focus error signal (FES) is shown in
Fig. 18.[32] The reflected light is directed into the servo lens, which affects light on the detectors
in a special way. That is, the light spot on the detector plane changes shape as a function of the
disk defocus. When the disk is too close to the objective lens, the light spot elongates along the
right diagonal on detector quadrants A and C. When the disk is in focus, the light spot is
circular. When the disk is too far from the objective lens, the light spot elongates along the left
diagonal on detector quadrants B and D. The FES signal is created by summing diagonal
quadrants and then subtracting the results. If the disk is too close to the objective lens, the FES
is positive. When the disk is in focus, the FES is zero. If the disk is too far from the objective
lens, the FES is negative. Near the focus condition, the FES is nearly linear and provides a good
quality feedback signal for the servo loop. The elongated spot behavior is due to a small amount
of astigmatism in the fabrication of the servo lens. Astigmatism is a difference in the focusing
power in diagonal directions, and is similar to astigmatism that commonly occurs in the eye.
5.7 Noise and jitter

As shown in Fig. 11, the resolution limit of the optical system is determined by the zero
of the modulation transfer function. In practice, it is not possible to obtain this limit due to noise.
Noise limits the ability of the detection electronics to determine the proper bit pattern from the
detector signal.[xref:] Sources of noise include reflectivity variations across the disk, photon
noise in the laser beam, detector noise and other sources.[33, 34]
A detector signal with noise is displayed in Fig. 19 for a large spot and a small spot. The
same amount of random noise is assumed for both signals. In order to detect the bit pattern, a
threshold level is established based on the signal amplitudes. When the signal level falls below
the threshold, a transition has occurred. The data bit value changes from 1 to 0 at the transition
marked in Fig 19, where there is one data bit for each mark and one data bit for each space
between marks. This minimum mark length is a function of the spot size s, and is generally
found to be 0.6s. Marks and spaces can be longer than the minimum mark length, but they
cannot be shorter.
In very simple recording schemes, the data bit may represent the desired output data
stream. In practice, the transition signals the change of a channel bit value, where more than one
channel bit is present for each data bit.[35, 36] It is the channel bits that determine the output
data stream and the data density. Each channel bit is defined by a channel bit window. Size of
the window is determined by how small the window can be before noise degrades the reliability
of detecting the transition.
A magnified portion of the transition region is shown in Fig. 20. The amplitude noise σN
on the signal creates uncertainty σw in the position of the transition. A detection error occurs if
the noise shifts the transition from the ideal window into a neighboring window. The widow size
is usually specified so that uncertainty σw produces no more than one error per 104 transitions.
Variation of the transition within the timing window is called jitter. As shown in Fig. 20, the
amount of width variation (or jitter) is a function of signal slope. That is, σw = m/σN, where m is
the signal slope.[37] Small spots yield high slope, small σw and short channel bit windows.
5.8 Data coding and formatting
The different ways of organizing ones and zeros on the disk are called formats.[38] There
are several different formats in use today, as illustrated in Tables III and IV, with new ones being
invented all the time. Some are more popular than others; some require special drives to access
them, while others are compatible with each other to some degree. This section describes the
CD-DA (digital audio) format. Details of other formats for CDs are available through the
standards set by the industry in a “rainbow” of reference books.[39] That is, a particular book
“color” corresponds to a particular standard. For example, CD-ROM format follows the
“Yellow Book”, CD-DA follows the “Red Book”, and CD-R and CD-RW follow the “Orange
Book”. These standards are important, because they insure interchangeability between different
players.2
2
Similar standards are available for the DVD family of products in specification books
labeled as “A” (DVD-ROM) through “E” (DVD-RAM). The DVD coding and formatting
philosophy is distinctly different than CDs, in that DVDs are designed from the outset to be more
compatible with computers, and a subset of the Universal Disk Format (UDF) is implemented for
the file system in both writeable and read-only versions.
The set of rules used to convert user data bits into their physical data-mark representation
and back again are called channel codes. The channel code for CD-ROM is called eight-to-
fourteen modulation (EFM).[35, 36] EFM interprets user's data along with error correction data,
address data, synchronization data, and other content into the stream of channel bits recorded in
the data-mark pattern. An example of an EFM sequence is shown in Fig. 21, where there are a
minimum of two zeros following each transition and a maximum of ten zeros following each
transition. The minimum number of zeros is set by the jitter requirement, as explained in Section
5.9, and the maximum length is set by the need to provide a synchronization signal for the
reference clock shown in Fig. 19. The conversion of an eight-bit user byte under these
restrictions leads to a fourteen-bit channel sequence, from which this code scheme derives its
name. A fourteen-bit channel sequence is called a symbol. During readout, the EFM decoder of
the CD-ROM works in the opposite direction, as shown in Fig. 4, converting the current signal
into a binary data stream, which is then cleared of any miscellaneous data by the drive's
electronics.
A problem can exist if two symbols follow each other. If a ‘1’ ending the first symbol is
adjacent to a ‘1’ of the second symbol, the ‘minimum of two zeros’ separation rule is violated.
To solve this problem, three special merge channel bits are placed between the two symbols.
Thus, for each eight-bit user byte, seventeen channel bits are used. The seventeen-bit sequence
is also called a symbol.
A basic unit of information stored on a CD is called an EFM channel frame, which
contains a twenty-seven-bit synchronization pattern and thirty-three seventeen-channel-bit
symbols. The sequence of symbols by themselves is called an EFM frame, which consists of

twenty four user data symbols, a control and display symbol, and eight error correction (ECC)
symbols, as shown in Table V.[5] The EFM channel frame is the smallest recognizable physical
sequence of data marks along a track. In digital audio CDs, sequential music data are scrambled
and cross interleaved throughout multiple frames, in a manner similar to digital audio tape
(DAT) formats.[40] This scrambling and cross interleaving is done to protect against
contamination on the disk or readout errors creating perceptible loss of audio data. During
readout, the decoder shown in Fig. 4 first determines the sequential symbols from the disk, then
decodes and unscrambles them to produce the output data stream.
In the CD-ROM (Mode 1) format, data are logically organized in sectors, which contain
2048 bytes of user data, a twelve-byte synchronization pattern, four bytes of sector identification,
and eight additional bytes that are used for other formats, like CD-I. A representation of a CD-
ROM (Mode 1) sector is shown in Table VI. Since sectors are simply a logical organization of
data, individual bytes in the sector can be distributed across the surface of the CD according to
the cross-interleaving scheme. A CD-ROM (Mode 1) sector spans 98 frames over the surface.
This sector organization is quite different from physical sectors on magnetic disk drives, where a
logical sector corresponds to a continuous length of track.[41] Other CD formats vary in the way
the sector is defined and use of symbols within the sector, but the basic EFM channel frame and
cross-interleaving strategy is common on all formats.
5.9 Configurations for optical media

Optical media can be produced in several different configurations. Figure 22 displays
four configurations that are in commercial use or have been tested in laboratories. The most
common configuration is the single-layer disk, like the compact disc (CD), where data are
recorded in a single storage layer
In order to increase data capacity of the disk, several layers can be used. Each layer is
partially transmitting, which allows a portion of the light to penetrate throughout the thickness of
the layers. The scan spot is adjusted by refocusing the illumination optics so that only one layer
is read out at a time. Some of the DVD formats in Table IV use two layers on one side of the
disk.
Data can also be recorded in volumetric configurations.[42, 43] Like with the multiple-
layer disk, the scan spot can be refocused throughout the volume of material to access
information. Volumetric configurations offer the highest efficiency for data capacity, but they
are not easily paired with simple illumination optics.
The final configuration is to place the information on a flexible surface, like ribbon or
tape.[44, 45] Like magnetic tape, the ribbon is pulled under the scan spot and data are recorded
or retrieved. Flexible media has about the same capacity efficiency as volumetric storage. The
advantage of a flexible medium over a volumetric medium is that no refocusing is necessary.
The disadvantage is that a moderately complicated mechanical system must be used to move the
ribbon.
5.10 Laser sources

The semiconductor laser diode [xref:?] is a key technology element in establishing the
optical data storage industry. Although both the communications industry and the optical data
storage industry use laser diodes, the latter consumes orders of magnitude more diodes that the
former. Each CD player on the market uses an AlGaAs laser diode that operates with a
wavelength around 0.780 micrometers. These small light sources are important, because they
can emit a relatively bright beam, they are reliable and they can be directly modulated with
simple electronics.
The operating power depends on whether data are being written to the disk or data are
being detected during readout. Since optical storage media are thermally sensitive, a relatively
high power is required for writing data. For example, CD-R and CD-RW media typically require
5mW to 10 mW at the disk surface during writing. Since there are losses in the optical system
associated with shaping the laser diode beam and directing the beam with beam splitters, the
typical efficiency of the optical path is around 50%.[46] Therefore, laser diodes that operate
with powers greater than 20 mW are generally required. For systems that write data faster than
standard playing time, more power is required because the disk spins faster. During readout, the
required laser power is greatly diminished. Typical read-only systems require only 0.5 mW.
The mode structure if the laser diode is important for two reasons. First, the source must
exhibit good spatial coherence in the transverse mode structure, because most tracking servo
techniques depend on an interference effect between diffracted orders, as explained in Section
5.6. Single transverse mode behavior is commonly achieved in commercial diodes over a wide
range of operating conditions.[xref:?] The longitudinal mode structure is important also. Laser
diodes for data storage applications typically exhibit more than one longitudinal mode.[xref:?]
This mode behavior is, in part, due to the inexpensive nature of the device. When the optical
system reads data, some amount of light reflected from the disk leaks back to the diode.
Feedback effects from this light returning to the laser can influence laser output and increase
laser noise, which is a source of jitter. Since jitter must be minimized in order to achieve high
density, as explained in Section 5.7, laser diodes are typically modulated at a high frequency
(around several hundred MHz) in order to mix the longitudinal mode structure and prevent mode
hopping noise due to feedback. This high-frequency modulation produces a low temporal
coherence, due to the relative large number of modes observed over the bandwidth of the data
detection electronics.
Since the resolution limit is improved by using short-wavelength lasers, modern DVD
systems use strained multiple quantum well diodes produced in AlGaInP/GaInP by
MOCVD.[47] These diodes typically emit with a wavelength from 0.635 micrometers to 0.680
micrometers. Power levels from the diodes reach 35 mW to 50 mW. Next-generation optical
disks will use InGaN violet laser diodes with operating wavelengths around 405 nm.[48, 49]
6.0 Performance
Three important performance characteristics of optical data storage devices are the
capacity, data rate and access time.

Capacity is the maximum amount of data that can be stored on a single disk.
Capacity is usually specified in terms of giga-bytes (GB), or 109 bytes.
(One byte = eight bits).
Data rate is the number of digital bits per second that are recorded or retrieved from
a device during transfer of a large data block. Data rate is usually specified
in terms of megabits-per-second (Mbps), or 106 bits per second.
Access time is the latency experienced between when a request is made to access data
and when the data starts flowing through the communication channel.
Access time is usually specified in terms of milliseconds, or 10-3 seconds.
Together, data rate and access time determine the throughput of the device. That is, throughput
determines the time required to locate and transmit data to and from the storage device.
The data rate can be different for writing and reading data on a disk. During writing, the
data rate is determined by the highest medium velocity that produces clearly defined marks.
During reading, the data rate is determined by the highest medium velocity that produces
sufficient signal-to-noise ratio. One straightforward way to increase data rate is to use more than
one laser beam at a time. The increase in data rate is nearly proportional to the number of
beams. A consumer CD product based on using multiple beams for readout that are generated
with a diffraction grating has been shown to dramatically improve data rate without large
increases in disk rotation rate.[50]

The access time is determined by the mechanical latency due to the disk rotation. The
highest latency is the time it takes the disk to make one revolution. Reduction of latency requires
spinning the disk faster.
Important considerations for storage are the performance requirements of new and
existing applications. An illustrative example is found in the CD/DVD marketplace. As
introduced in 1991, the CD-ROM exhibited a capacity of 0.64GB and a data rate of 1.2 Mbps.
Although today’s CD-ROM has the same capacity, market forces have driven the data rate to
over 50 Mbps. The increased data rate of CD-ROM drives may have, in part, been responsible
for the delayed market acceptance of DVD-ROM. Introductory DVD products exhibit a data rate
of 10 Mbps. Thus, computer applications eagerly accept higher speed devices.
A serious limitation exists with disk-based optical data storage. As the data rate
increases, the playing time for a fixed capacity decreases. Applications that require long playing
times (and correspondingly high capacities) must use multiple disks. For example, a CD-ROM
drive operating at 50 Mbps takes only 102 seconds to read the entire disk. Correspondingly, a
hypothetical DVD-ROM drive operating at 400 Mbps (a similar speed multiplier compared to
the fast CD drive) takes less than 100 seconds to read a 4.7 GB disk.
A useful figure of merit is the capacity-rate product (CRP), which is the product of the
capacity in GB and the data rate in Mbps. The CRP and other performance characteristics of
disk-based products are given in Table VII. The data-rate speedup factor is shown as “1X” or
“40X”, where 1X refers to the data rate of products first introduced into the marketplace, like the
CD-ROM in 1991. 40X refers to a data rate that is forty times faster that the 1X rate. Also

included in Table VI are preliminary data concerning the digital video recorder (DVR) or Blu-
ray, which is under development.[51]
7.0 Future systems
Future illumination optical systems will use high NA and shorter λ in order to obtain
smaller spot size and higher data density. The effects of using higher NA and shorter λ are
shown in Fig. 23. For example, if the NA of a DVD system is increased to NA = 0.85, the spot
size is reduced by 30%. However, the allowable defocus is reduced by 50%. Alternatively, a
blue laser operating at λ = 0.405 micrometers and NA = 0.60 achieves nearly a 40% reduction in
best-focus spot size at a penalty of reducing the allowable defocus by the same 40% factor. In
general, it is desirable to decrease wavelength rather than increase NA due to the difficulty of
decreased depth of focus.
Parameters of three generations of optical disk products are shown in Table VII. These
systems are evolutionary products. Shortening laser wavelength and increasing NA reduce spot
size and increase capacity. The Blu-ray system, which operates at λ = 0.405 micrometers and
NA = 0.85, provides a capacity of 27 GB per layer. However, the Blu-ray system is near the
limit for conventional optical systems with standard optical materials. For example, increasing
NA beyond 0.85 is possible with a conventional optical system, but the engineering challenges
are substantial. In addition, most plastic substrates exhibit poor transmission below 0.400
micrometers. Even if a new laser diode becomes available with a wavelength shorter than 0.400

micrometers, it is not clear whether substrate, detector and media technologies can support it.
Instead, recent research points to two promising technologies that may provide the fourth
generation optical disk.
The first technology is called near-field optics.[52, 53] Near-field optics use a
transducer, like a small hole in a metal film or a special lens element, to produce a light spot that
is smaller than the ideal spot size given by s = λ/NA. However, the tradeoff for smaller spot size
is that the recording layers now must be in proximity to the transducer. The evanescent energy
in the spot that couples from the transducer to the recording layers falls off exponentially with
distance.
Invented by Prof. Gordon Kino and colleagues at Stanford University, the solid
immersion lens (SIL) is under investigation as a possible candidate near-field transducer.[54]
The basic SIL system is shown in Fig. 24, where the optical system is supplemented with a
hemispherical lens element. When the focused light from the objective lens enters the SIL, the
velocity of the light slows down according to n, the index of refraction of the lens.[xref:]
Marginal ray angle θ is not deviated by the hemisphere as it enters the lens material, so θ′ = θ.
Since the laser frequency does not change, the effective wavelength of the light reduces and the
spot size is now given by s = λ/(n sin θ′) = λ/NAeff, where NAeff is the effective numerical
aperture. In laboratory systems, NAeff approaching 2.0 have been demonstrated.[55] When
coupled with a blue laser diode, the potential of a SIL system is to increase capacity beyond a
factor of four above the Blu-ray system. However, control of the gap that separates the SIL from
the recording layers is a difficult engineering problem, especially if the optical disk is removable
from the optical drive.

A second possible technology for fourth-generation optical disks is called magnetically
amplifying magneto-optical system (MAMMOS), which is similar to MO systems described in
Section 4. MAMMOS technology take advantage of the fact that the primary limitation to
resolution in optical data storage systems is reading data, as explained in Section 5.3. With a
pulsed laser and a modulating external field, magnetic domains can be written in the recording
layer that are much smaller than the resolution limit.[56] Readout of these marks in a
MAMMOS system is illustrated in Fig. 25, where a multiple-layer MO stack is used. Each MO
layer reacts differently to the heat deposited by the laser beam. The bottom layer, which is called
the recording layer, contains the written information in the form of small bits. This layer has a
high coercivity, and it is not easily affected by the relatively low temperature profiles generated
by the readout beam. The top layer is the expansion layer, and it has a low coercivity, among
other special properties. The middle layer is a thin nonmagnetic layer. When the readout beam
heats the expansion layer, magnetic energy from the recording layer couples into the expansion
layer and forms an expanded copy of the recording layer in it. Only a small region of the storage
layer around the center of the laser spot is copied. Expansion of the bit pattern produces a
magnified image in the expansion layer. To the readout optical system, it appears that the light
spot travels over relative large marks, which produce good signal-to-noise ratio. Capacity of
MAMMOS systems have been demonstrated to be about three times greater than Blu-ray
disks.[57] Potential difficult in MAMMOS systems lies in economically producing disks and
player systems.
Other systems are also worthy of mention, because disruptive technologies are always
possible. For example, near-field super-resolution structures (SuperRENS) combine a nonlinear

optical material layer with a conventional phase-change media structure.[58] This system has
the advantage of using near-field optical effects with a conventional readout system. It is not
necessary to maintain a small gap, as required in the SIL systems. Also, volumetric storage
systems show promise. Instead of recording only on one or two layers, volumetric bit-wise
systems store data on several hundred layers through the thickness of the disk.[59] Volumetric
bit-wise system may need to use nonlinear properties of the recording layers in order to record
data marks without interference from other layers. Finally, optical data storage and magnetic
disk storage may converge into hybrid recording, which uses the optical beam only as a heat
source to lower the coercivity for magnetic writing.[60] Hybrid readout may be accomplished
with magnetic sensors, and hybrid recording may use near-field optics.

REFERENCES
[1] D. P. Gregg, “Transparent recording disc,” U.S. Patent 3,430,966, issued March 4, 1969.
[2]2.) M. Bellis, “David Paul Gregg and the Optical Disc,” from
http://inventors.about.com/library/inventors/blopticaldisk.htm, Nov. 7, 2002.
[3] D. R. Cellitti, “World On A Silver Platter: A Brief History of Optical Disc,” published in
Widescreen Review’s Laser Magic 1988 Guide. An updated copy of this article can be found at
http://www.oz.net/~blam1/DiscoVision/WRLM98_main.htm, Nov. 11, 2002.
[4] D. G. Howe, “Data Reliability and Errors,” Ch. 2 in Handbook of Magneto-Optical Data
Recording, T. W. McDaniel and R. H. Victora, eds. Noyes Publications, New Jersey, 1997.
[5] K. A. S. Immink, Codes for Mass Data Storage Systems, Shannon Foundation Publishers, The
Netherlands,, 1999.
[6] M. Kubo, S. Harada, H. Takeshima, T. Kobayashi, T. Ohmori and Y. Kobayashi,
“Prerecorded signal characteristics of high-density optical disks,” Japanese Journal of Applied
Physics 32(Part 1, No. 11B), pp. 5329-5334 (1993).
[7] A. B. Marchant, Optical Recording: A Technical Overview, Addison-Wesley Publishing
Company, Reading, Massachusetts, 1990.
[8] B. Tieke, G. R. Langereis, E. R. Meinders, J. G. F. Kablau, R. Woudenberg and R. A. J. van
Kollenburg, “Thermally balanced writing for high-speed compact disk recordable (CD-R)
Recording, Japanese Journal of Applied Physics 41(Part 1, No. 3B), pp. 1735-1738 (2002).
Author: Tom D. Milster (Prepared for Laser Handbook)

[9] A. N. Burgess, K. E. Evans, M. Mackay and S. J. Abbot, “Comparison of transient thermal
conduction in tellurium and organic dye based digital optical storage media,” Journal of Applied
Physics 61(1), pp. 74-80 (1987).
[10] Y. J. Huh, J. S. Kim, T. Y. Nam and S. C. Kim, “Deformation effects and recording
characteristics of compact-disk recordables,” Japanese Journal of Applied Physics, 36(Part 1, No.
1B), pp. 403-409 (1997).
[11] “CD Dye” a web document published by CD Media World, Nov 11, 2002,
http://www.cdmediaworld.com/hardware/cdrom/cd_dye.shtml
[12] D. Stinson, F. Ameli and N. Zaino, “Lifetime of Kodak Writable CD and Photo CD media,”
a white paper available from Digital and Applied Imaging Division of Eastman Kodak Company,
1669 Lake Avenue, Rochester, N.Y. (1995). see also http://www.cd-
info.com/CDIC/Technology/CD-R/Media/Kodak.html.
[13] T. Handa, J. Tominaga, S. Haratani and S. Takayama, “In-Ag-Te-Sb phase change
recording media at compact disk linear velocity,” Japanese Journal of Applied Physics 32(Part 1,
No. 11B) pp. 5226-5229 (1993).
[14] M. Terada, K. Furuya, T. Okamura, I. Morimoto and M. Nakao, “Optimized Disk Structure
and Ge-Tb-Sb Composition for Overwritable Phase Change Compact Disk,” ,” Japanese Journal
of Applied Physics 32(Part 1, No. 11B) pp. 5219-5222 (1993).
[15] E. Hamada, Y. Takagishi, T. Yoshizawa, T. Fujii, R. Negishi and T. Nakajima, “Ten-Year
Overview and Future Prospects of Write-Once Organic Recordable Media,” Japanese Journal of
Applied Physics 39(Part 1, No. 2B), pp. 785-788 (2000).

[16] Y. S. Tyan, D. R. Preuss, F. Vazan and S. J. Marino, “Laser recording in tellurium suboxide
thin films,” Journal of Applied Physics 59(3), pp. 716-719 (1986).
[17] M. Suzuki, K. Furuya, K. Nishimura, K. Mori and I. Morimoto, “Disk structure for high
performance phase change erasable optical disk,” in Optical Data Storage 1990, Maarten de
Haan and Yoshito Tsunoda, Editors, Proceedings of SPIE Vol. 1316, pp. 374-381 (1990).
[18] see, for example, C. D. Mee and E. D. Daniel, Magnetic Recording Volume I: Technology,
McGraw-Hill, New York, 1987.
[19] J. E. Hurst, Jr., and T. W. McDaniel, “Writing and Erasing in Magneto-Optical Recording,”
chapter 7 in Handbook of Magneto-Optical Data Recording, T. W. McDaniel and R. H. Victora
eds., Noyes Publications, New Jersey, 1997.
[20] C. D. Wright, “The Magneto-Optical Readout Process,” chapter 8 in Handbook of Magneto-
Optical Data Recording, T. W. McDaniel and R. H. Victora eds., Noyes Publications, New
Jersey, 1997.
[21] M. H. Kryder, “Magneto-optic recording technology,” Journal of Applied Physics 57(8) pt.
2B, pp. 3913-3918 (1985).
[22] C. Brucker, “Magneto-optical thin film recording materials in pratice,” ,” chapter 5 in
Handbook of Magneto-Optical Data Recording, T. W. McDaniel and R. H. Victora eds., Noyes
Publications, New Jersey, 1997.
[23] A portion of this information was obtained from web document
http://www.cilect.org/NCDRFF.htm, Nov. 12, 2002.

[24] H. Ide, T. Toda, F. Kirino, T. Maeda, F. Kugiya, S. Mita and K. Shigenatsu, “Precise mark
shape control in mark length recording on magnetooptical disk,” Japanese Journal of Applied
Physics, 32(Part 1, No. 11B), pp. 5342-5348 (1993).
[25] F. Yokogawa, S. Miyanabe, M. Ogasawara, H. Kuribayashi, Y. Tomita and K. Yamamoto,
“Signal Processing for 15/27 GB Read-Only Disk System,” Japanese Journal of Applied Physics
39(Part 1, No. 2B), pp. 819-823 (2000).
[26] G. Brouwhuis and J. J. M. Braat, “Video disk player optics,” Applied Optics, 17(13), pp.
1993-2000 (1978).
[27] B. I. Finkelstein and E. R. Childers, “The effects of focus misregistration on optical disk
performance,” in Optical Data Storage 1991, James J. Burke, Thomas A. Shull and Nobutake,
Editors, Proceedings of SPIE Vol. 1499, pp. 438-449 (1991).
[28] K. W. Getreuer and L. J. Grassens, “Servos and Actuators,” ,” chapter 3 in Handbook of
Magneto-Optical Data Recording, T. W. McDaniel and R. H. Victora eds., Noyes Publications,
New Jersey, 1997.
[29] R. C. Dof, Modern Control Systems, Addison-Wesley, Reading, Massachusetts (1986)
[30] J. J. M. Braat and G. Brouwhuis, “Position sensing in video disk readout,” Applied Optics,
17(13), pp. 2013-2021. (1978).
[31] D. G. Stork, “CD ROM apparatus for improved tracking and signal sensing,” U. S. Patent
5,642,341 issued June 24, 1997.
[32] D. K. Cohen, W. H. Gee, M. Ludeke and J. Lewkowicz, “Automatic focus control: the
astigmatic lens approach,” Applied Optics, 23(4), pp. 565-570 (1984).

[33] B. I. Finkelstein, “Sources of noise in magneto-optical readout,” chapter 9 in Handbook of
Magneto-Optical Data Recording, T. W. McDaniel and R. H. Victora eds., Noyes Publications,
New Jersey, 1997.
[34] C. Peng and M. Mansuripur, “Sources of noise in erasable optical disk data storage,” Applied
Optics 37(5), pp. 921-928 (1998).
[35] K. A. S. Immink, “A survey of codes for optical disk recording,” IEEE Journal on Selected
Areas in Communications, 19(4),pp. 756-764 (2001 ).
[36] K. A. S. Immink, “Run-length limited sequences,” Proceedings of the IEEE , 78(11) pp.
1745 –1759 (1990).
[37] D.G. Howe, “Signal-to-noise ratio (SNR) for reliable data recording,” in Optical Mass Data
Storage, R P Freese, A A Jamberdino, M R de Haan, Editors, Proceedings of SPIE Vol. 695, pp.
255-261 (1986).
[38] Good reference sites to download information about CD and DVD formats are
http://www.disctronics.co.uk/technology/index.htm or
http://www.discusa.com/cdref/cdbooks/books.htm.
[39] These books can be obtained from Philips International B.V., System Standards &
Licensing, Building SFF 8, Glaslaan 2, 5616 LW Eindhoven, or P.O. Box 80002, 5600 JB
Eindhoven, The Netherlands, Fax: +31-40-2732113, URL: www.licensing.philips.com
[40] E. Tan and B. Vermuelen, “Digital audio tape for data storage,” IEEE Spectrum, 26(10 ), pp
34 –38 (1989).
[41] R. Grossblatt, “Floppy-disk data storage,” Computer Digest, 4(12),pp. 91-4, 100 (1987).

[42] J. Ashley, M. P. Bernal, G. W. Burr, H. Coufal, H. Guenther, J. A. Hoffinagle, C. M.
Jefferson, B. Marcus, R. M. Macfarlane, R. M. Shelby, G. T. Sincerbox, “Hologaphic Data
Storage,” IBM Journal of Research and Development 44(3), pp. 341-368 (2000).
[43] S. Hunter, F. Kiamilev, S. Esener, D. A. Parthenopoulos and P. M. Rentzepis, “Potentials of
two-photon based 3-D optical memories for high performance computing,” Applied Optics
29(14), pp. 2058-2066 (1990).
[44] H. Tokumaru, H. Okumura, K. Arai, N. Kawamura and S. I. Yoshimura, “Multi-Beam
Optical System for Optical Tape Recording,” Japanese Journal of Applied Physics 37(Part 1, No.
4B) pp. 2241-2244 (1998).
[45] G. W. R. Leibbrandt, J. A. H. Kahlman, G. E. van Rosmalen and J. J. Vrehen, “Optical tape
system: evaluation of recorder and media,” in Optical Data Storage 1997, Henryk Birecki and
James Z. Kwiecien, Editors, Proceedings of SPIE Vol. 3109, pp. 106-115 (1997).
[46] T. D. Milster, M. K. Benedict and R. P. Stahl, “Laser diode requirements for magneto-
optical storage devices,” in Optical Data Storage 1990, Maarten de Haan and Yoshito Tsunoda,
Editors, Proceedings of SPIE Vol. 1316, pp. 143-149 (1990).
[47] R. Hiroyama, D. Inoue, Y. Nomura, M. Shono and M. Sawada, “High-Power 660-nm-Band
AlGaInP Laser Diodes with a Small Aspect Ratio for Beam Divergence,” Japanese Journal of
Applied Physics, 41(Part 1, No. 2B), pp. 1154-1157 (2002).
[48] J. Piprek and S. Nakamura, “Pohysics of high-power InGaN/GaN lasers,” IEE Proc.-
Optoelectron. 149(4), pp. 145-151 (2002).

[49] G. Hatakoshi, M. Onomura, S. Saito, K. Sasanumaand and K. Itaya, “Analysis of Device
Characteristics for InGaN Semiconductor Lasers,” Japanese Journal of Applied Physics, 38(Part
1, No. 3B), pp. 1780-1785 (1999).
[50] see, for example, http://www.kenwoodtech.com/white_paper6.html, Nov. 11, 2002.
[51] S. Furumiya, J. I. Minamino, H. Miyashita, A. Nakamura, M. Shouji, T. Ishida, H. Ishibashi,
“Optical disk recording system of 25GB capacity,” in Optical Data Storage 2001, Terril Hurst,
Seiji Kobayashi, Editors, Proceedings of SPIE Vol. 4342, pp. 186-193 (2002).
[52] T. D. Milster, “Near-field optics: A new tool for data storage,” Proceedings of the IEEE,
88(9), pp. 1480-1490 (2000).
[53] T. D. Milster, “Near-field optics: Avenues for increased performance,” Optical Engineering
40(10), pp. 2255-2260 (2001).
[54] S. M. Mansfield, W. R. Studenmund, G. S. Kino and K. Osato, ‘‘High numerical-aperture
lens system for optical data storage,’’ Optics Letters, 18(4), pp. 305–307 (1993).
[55] M. Shinoda, K. Saito, T. Kondo, T. Ishimoto and A. Nakaoki, “High density near field
readout over 50GB capacity using a solid immersion lens with high refractive index,” Paper
WC.2, , Technical Digest of the Joint International Symposium on Optical Memory and Optical
Data Storage, Waikoloa, Hawaii, 7-11 July, 2002, IEEE catalog number 02EX552, pp. 284-286.
[56] H. Awano, S. Ohnuki, H. Shirai, N. Ohta, A. Yamaguchi, S. Sumi, and K. Torazawa,
“Magnetic domain expansion readout for amplification of an ultra high density magneto-optical
recording signal,” Applied Physics Letters, 69(27), pp. 4257-4259 (1996).

[57] H. Awan0, M. Sekine, M. Tani, N. Kasajima, N. Ohta, K. Mitani, N Takagi and S. Sumi,
“0.04 µm Domain Expansion Readout for the Magnetic Amplifying Magneto Optical System,”
Japanese Journal of Applied Physics, 39(Part 1, No. 2B), pp. 725-728 (2000).
[58] J. Tominaga, F. Hiroshi, A. Sato, T. Nakano, T. Fukaya, and N. Atoda, ‘‘The near-field
superresolution properties of an antimony thin film,’’ Japanese Journal of Applied Physics,
37(Part 2, No. 11A), pp. L1323–L1325 (1998)
[59] Y. Kawata, M. Nakano, and S. C. Lee, “Three-dimensional optical data storage using three-
dimensional optics, “ Optical Engineering 40(10) 2247-2254 (2001).
[60] N. Ota, M. Sekine, H. Awano, S. Imai, J. Hohlfeld and T. Rasing “Hybrid recording
capability toward tera bit/in 2 and 100 Gbps storage device,” paper TuC.1, Technical Digest of the
Joint International Symposium on Optical Memory and Optical Data Storage, Waikoloa, Hawaii,
7-11 July, 2002, IEEE catalog number 02EX552, pp. 156-158.

Table I. Four technologies of commercial optical disks.
Disk Technology Description

CD-ROM or DVD- Compact-disc (CD) and digital versatile disc (DVD) products use
ROM pit-type technology. CD and DVD products are read-only
memories (ROMs), that is, they are used for software or
entertainment distribution and cannot be used for recording
information.
CD-R Compact-disk-recordable (CD-R) products use dye-polymer
technology. CD-R products can be used for recording information,
but, once the information is recorded, it cannot be erased and
reused.
CD-RW Compact-disk-rewriteable (CD-RW) products use phase-change
technology. Data can be erased and the disk reused.
MO Erasable disks using magneto-optic (MO) technology are popular
for work-station environments. Data can be erased and the disk
reused.

Table II. CD-R materials, reflective layer, and the resulting color as observed through the clear
side of the substrate.
Disk Color
Dye Material Gold Reflector Silver Reflector
Cyanine Green Green/Blue
PhthaloCyanine Gold -
Metallized Azo - (Dark) Blue
Advanced PhthaloCyanine Gold -
Formazan (hybrid
Cyanine/PhthaloCyanine
combination) Green/Gold -

Table III. CD formats
Format Data-Mark Technology Characteristics

pit Computer storage medium with the capacity of up to 700
CD-ROM
MB
CD-R dye polymer Write-once computer and audio storage > 700 MB
CD-RW erasable phase change Erasble > 700 MB storage with > 1000 erase cycles
pit Audio CD plus additional graphics and/or text
CD+G (CD+Graphics
information recorded in R-W subcodes
pit Audio CD plus MIDI music information in the R-W
CD+Midi subcodes to enable sounds from CD to be re-voiced or
remixed through the MIDI compatible device.
pit Audio CD with the smaller diameter - (3") with playing
CD-3 (3"CD)
time reduced to 20 minutes. Also called CD-single
pit Original, audio compact disc, containing up to 74
CD-A (CD Audio) minutes (suggested by "Red Book") of stereo digital
audio along with 8 subcode tracks labeled P-W
CD-E (Erasable) erasable phase-change Audio or data CD which can be recorded and erased many times
CD-EG (CD Extended pit Enhancement to the CD+G format adding 256 colors and
Graphics) instant mix of two pictures
pit Extension of the CD-ROM format aimed specifically to
CD-I the consumer market. System offers high resolution
graphics, still and (recently) moving pictures and
(CD-Interactive)
stereo sound. The CD-I player also plays back CD-A,
CD+G and Photo CD.
pit A CD-A that contains additional data "Hidden" in a
CD-I Ready space before the first track. Loaded into a CD-I player
the disc offers many features of full CD-I.
pit Development of the CD-ROM which has been designed to
CD-ROM XA
meet the multimedia and interactive needs. Audio,
(Extended
graphics and (some) video information have been added
Architecture)
to original CD-ROM format.
pit Also known as the LaserDisc containing up to 2 hours of
CD-V (CD Video) analogue video and digital audio information. There are
three formats 12", 8" and 5".
pit A version of the CD-V containing up to 5 minutes of
CD-V Single
video plus 20 minutes of CD audio only.
dye polymer Audio CD allowing direct recording (only once) of
CD-WORM (Write
musical information. Used also for storing of computer
Once, Read Many)
data or copying of CD-ROMs.
pit A version of the CD-ROM discs written in the AMIGA
CDTV language with the interactive and graphics
possibilities.
magneto-optical A SONY 2" rewritable magnetooptical disc which is using
MINI DISC data reduction system to record up to 74 minutes of the
audio information (possible use for computer data
storage)
dye-polymer A KODAK development which is using a CD-WORM system to
record up to 100 still pictures. The player is
Photo CD
compatible with the CD-I system). To be able to read
these disks, CD-Player must be "multi session"
pit CD using MPEG-1 encoding process to compress video
(including feature films) on CD. The picture quality is
Video CD
higher than in standard CD-ROMs, but additional
hardware (MPEG encoder) is necessary.

Table IV . DVD formats
Data-Mark
Format Characteristics
Technology
DVD-5 pit Single layer 4.7 GB read-only DVD.
pit Dual layer 7.95 GB read-only DVD. Both layers are read from
DVD-9
one side of the disk.
pit Double sided 8.7 GB DVD. Must turn disk over to read
DVD-10
second side.
pit Dual layer, double sided 17.1 GB read-only DVD. Can read
DVD-18
two layers from each side.
pit One of DVD-5 through DVD-18 with MPEG-1 or MPEG-2
DVD Video
video files, audio, subpictures and navigation data.
DVD Audio pit DVD-5 with high-quality audio files.
pit One of DVD-5 through DVD-18 with computer-friendly file
DVD-ROM
formats.
DVD-R dye-polymer Write-once 4.7 GB/side
dye-polymer Similar to DVD-R, except designed to be compatible with
DVD+R
DVD+RW
erasable phase change Erasable computer-friendly random access with 4.7 GB/side.
DVD-RAM Number of erase cycles > 100,000. Not compatible with all
players.
erasable phase-change Erasable with better compatibility than DVD-RAM and >
DVD-RW
1000 erase cycles.
erasable phase change Similar to DVD-RW, except designed to be compatible with
DVD+RW
DVD-ROM and DVD Video players
pit CD-V authored on a DVDR/W. Audio has to be resampled to
DVD-VCD
48 khz.
pit SVCD authored on a DVDR/W. Higher quality video than
DVD-SVCD DVD-VCD. Audio has to be resampled to 48 khz like the
DVD-VCD.
DVD-MP3 dye polymer MP3s burned on a DVDR/W.
pit DVD format on a CD-R(W) instead of a DVD disc. miniDVD
is also sometimes called cDVD. A miniDVD only fits about
miniDVD
15 minutes video on a 650 MB CD-R(W) . This fis not a
supported format.

Table V. EFM Channel Frame of a Compact Disk (contiguous in along track)
Control and
Synch User Data ECC
display
27 bits 17 bits (1 symbol) 408 bits (24 symbols) 136 bits (8 symbols)
ß------------------------------------------ 561 bits (33 symbols) = 1 EFM frame----------------------------------------à
ß--------------------------------------------------- 588 bits = 1 EFM channel frame ---------------------------------------------------à

Table VI. CD-ROM Logical Sector After Decoding
Synch Address Mode User Data ECC Reserved

12 bytes 3 bytes 1 byte 2048 bytes 280 bytes 8 bytes
ß----------------------------------------------------------- 2352 bytes ---------------------------------------------------------à
ß-------------------------------------------------------- 98 EFM frames --------------------------------------------------------à

Table VII. CRP
Parameter CD-1X CD-40X DVD-1X DVD-40X Blu-ray 1X
capacity (GB) 0.64 0.64 4.7 4.7 20
data rate (Mbps) 1.2 48 10 400 25
CRP 0.77 30.7 47 4700 500
retrieval time (min) 70 1.7 62.7 1.6 106.7

Table VIII. Parameters of optical disk products
Parameter units CD DVD Blu-ray

Wavelength micrometer 0.78 0.65 0.405
NA 0.45-0.5 0.6-0.65 0.85
Track Pitch micrometer 1.6 0.740 0.320
Shortest Pit micrometer 0.831 0.399 0.138
Density Gb/in2 0.4 2.8 15.9
Capacity GB 0.65 4.7 27

FIGURE CAPTIONS
1.) This picture of lines written by an electron beam was the inspiration for the invention of
the videodisk in the late 1950’s. Lines in the lower left-hand corner of the picture are 0.030
micrometers wide and spaced by 0.070 micrometers. If it were possible to reliably record data at
this density, a 130 mm diameter optical disk would have several thousand times the capacity of a
digital versatile disc (DVD).
2.) Small sections of a compact disc (CD) and a digital versatile disc (DVD) are displayed
relative to the laser spots that are focused onto them during recording and readout.
3.) The process of recording data onto an optical disk starts with the user input data stream
converted to a current drive signal for the laser diode. Intense pulses from the laser cause
physical changes in the surface of the recording medium as the disk spins, which result in spiral
tracks of data marks.
4.) A constant, low power laser beam scans a data track to readout data from the disk.
Reflected light, which is modulated by the data-mark pattern, causes modulation in the reflected
light. The reflected light is directed to servo and data detectors with a beam splitter, which
convert the light modulation in a current signal that is then decoded.
5.) An optical disk is used so that the focusing laser light is passed through the clear side of
the substrate and illuminates the data-mark pattern on the other side. The thickness of the cover
layer is designed to reduce effects of contamination on the surface of the disk, like dust or
fingerprints.
6.) The process of recording data marks on CD-RW media. In (A), a laser power modulation
with three peaks creates a pattern of three amorphous marks in a bright crystalline background.
Overwriting of new data are shown in (B), where a different laser pulse pattern is used. The bias

power is used to produce the annealing of the medium into the crystalline state, while the peak
laser pulses are use to quench the medium into the amorphous state.
7.) Detail of the laser spot irradiance distribution near the focus of the data marks. Energy
falls off rapidly with defocus ∆z, and the spot has a finite width s. Multiple thin-film layers are
used to enhance recording and readout characteristics of CD-RW media.
8.) A) Magneto-optic recording involves heating the recording layer with the focused laser
spot in order to reduce the layer coercivity. Domains in this small region can then be flipped with
an externally applied magnetic field in order to form the data-mark pattern. Once the bits cool,
they are frozen in place until heated again. B) During readout, a low-power focused spot
illuminated the data marks. Linearly polarized light from the laser is rotated either clockwise or
counterclockwise, depending on domain orientation. The data detector senses the polarization
change and converts this information in to a current signal.
9.) Since optical disk media are sensitive to a thermal threshold, simulated thermal contours
of a laser spot scanning the recording surface show isotherms that predict the mark size and
shape.
10.) The data-mark pattern reflects light in a diffraction pattern consisting of three primary
cones, which are the zero and +/- 1st diffraction orders. As the data pattern moves under the laser
spot, relative phase of the +/- 1st diffracted orders changes. In the overlap area between orders,
the phase difference produces an interference effect that modulates the irradiance level. This
light modulation is then converted into a current signal by the detectors. The amount of overlap
area, and hence the amplitude of the data signal, depends on the spatial frequency 1/T of the data-
mark pattern. Higher-frequency patterns produce smaller overlap area.
11.) The modulation transfer function is a plot of signal contrast versus spatial frequency of
the data-mark pattern. Defocus blurs the laser spot and reduces contrast, especially in the mid-

frequency range. Equalization circuits can be used to boost the transfer function after the light
modulation has been converted into a current signal.
12.) The relative irradiance distribution of a focused DVD-like laser spot is shown with
different amounts of defocus. More than 0.5 micrometers of defocus significantly degrades the
peak irradiance and spot quality.
13.) A basic diagram of the servo loop used in optical storage devices shows the gain (G) of
the drive electronics producing current for the actuator (H), which positions the lens over the data
track. A feedback sensor provides an error signal for robust control of spot position.
14.) A photo of a commercial CD actuator assembly that illustrates the objective lens,
suspension flexure, and the coils used to move the suspension.
15.) The push-pull tracking error signal (TES) is generated by using a slit-cell servo detector
and sensing the difference in light level between the cells. Since grooves of the disk diffract light
like a grating, diffracted orders overlap, as describes with Fig. 10. (In this case, spatial frequency
of the grating is fixed, and the diffraction occurs in an orthogonal direction compared to
diffraction from data marks.) As the laser spot moves off track, the relative phase change in the
diffracted orders produces bright and dark patterns on the detector. The TES difference signal
indicates the relative off0track location of the laser spot.
16.) Diffracted from the disk contains orders from the data-mark pattern and the grooves.
17.) Three-spot tracking uses two auxiliary laser spots that ride edges of the track as the disk
spins. The spots are reimaged onto separate detector elements. If the data track is not centered,
the amount of light reflected from each auxiliary spot changes. One detector spot becomes
brighter, and the other dims. A difference signal produces a reliable TES.
18.) The astigmatic focusing technique uses a special lens in the servo optics before a
quadrant-cell detector. The lens introduces a small amount of astigmatism into the beam along a
diagonal direction on the detector. As the spinning disk goes into and out of focus, the

astigmatism forces the light spot to change shape. A difference signal from the detector
quadrants produces an FES that indicates the amount of defocus.
19.) Detector signals with noise are shown for scanning data marks with two spot sizes. The
width of the channel-bit window is determined by the reliability of detecting a transition across a
threshold level. Transitions determine the potions of channel-bit 1’s in the data pattern, so data
marks and spaces represent more than one channel bit.
20.) Small spots produce high slopes in the transition region, and the signal is less affected by
noise as compared to the system with the larger spot. Therefore, laser systems the generate
smaller spots can pack more channel bits into the data mark sequence.
21.) An eight-to-fourteen (EFM) modulation code produces a laser pulse signal and data-mark
pattern that exhibits a specific number of zeros between each transition.
22.) Four possible configurations of optical storage media include a single -layer substrate, like
a CD, a multiple-layer substrate, like a DVD, volumetric configurations, like holographic and bit-
wise storage, and ribbon (tape) media.
23.) Future optical data storage systems will exhibit both higher NA and shorter wavelength
lasers. Decreasing wavelength rather than increasing NA has the advantage of a larger depth of
focus.
24.) A solid immersion lens (SIL) system uses a image-centric hemisphere in near contact to
the recording layers. The SIL increases the effective numerical aperture of the system by a factor
of the lens refractive index, and thus decrease the focused spot size by the same amount.
25.) The magnetically amplifying magneto-topical system (MAMMOS) uses a special
configuration of magnetic layers to produce good signal readout from a very small data mark
pattern.

Milster: Optical Data Storage Figure 1
CD DVD
1.6 µm track pitch 0.74 µm track pitch
1.6:m s scan
direction

input current
data drive
stream signal
110010010111010101010
Encoder/ laser
modulator source
illumination
newly optics
written
storage data intense light
medium beam
θ (half angle = θ)
scan
spot

output
data laser source
stream (low and constant
110010010111010101010 power level)
servo/data
optics
detectors
amplifier/ beam
decoder splitter
current
signal reflected
light
illumination
optics
data to
storage be read
medium
low-power
light beam
scan
spot

contamination
focused laser beam
top of disk (clear)
1.2 mm for CD
0.6 mm for DVD pits
aluminum
protective lacquer
label

peak power
Laser Power (for A)
bias power
read power
time
A.) Before overwrite
B.) After overwrite
crystalline amorphous
(bright) (dark)

90% 100%
relative
irradiance
50%
of the laser 50%
beam
13.5%
10%
3 µm
disk structure*
2 µm above the first
dielectric is the clear
substrate material
1 µm (refractive index = n )
first dielectric
∆z recording material
second dielectric
0 µm reflector
protective coating
s label and ink
∆z = λn/(4sin2θ) *layer thicknesses are exaggerated

s = λ/(n sin θ)

focus beam
(pulsed for
A) MO recording each mark
reversal) spot irradiance at the
external magnetic field
(uniform over recoding area) recording layer
domains initialized in
one direction
disk motion
data marks of varying length
B) MO readout
linearly
polarized cw ccw
illumination rotation rotation

2
Distance from center of pulse

(micrometers) 1
0 300o C
200o C
-1 100o C
-2
-1 0 1 2 3
Distance from start of laser pulse
(micrometers)

overlap area
pupil
-1 order 0 order +1 order
objective lens
(in collection)
reflected light 0
-1 +1
cones due to
diffraction ψ scan
T direction
data marks along track are
like a diffraction grating

Modulation Transfer Function 1
no with equalization
0.5 mi
na
l
with defocus
0
0 NA 2NA
λ λ
Spatial frequency of the data-mark pattern (1/T)

in focus 90% 0.5 µm defocus
50%
13.5%
1.0 µm defocus 2.0 µm defocus

+ Gain Actuator
Σ (G) (H)
Desired Spot
Spot - Position
Position
Feedback Sensor
Error
Signal

tracking and focus
suspension
actuator coils
flexure
objective lens

detectors
amplifier
beam
A splitter
+
TES
-
B
overlap
reflected light areas
TES = A - B
objective diffraction
lens orders from
position tracks
error
scan data
direction marks
track pitch track position land

pitch error
groove

Diffracted orders due to
data marks change
separation as a Diffracted orders due
function of mark to tracks have
frequency and constant separation,
modulate brightness in but change
the overlap areas as brightness in the
the spot scans along overlap areas as the
the data marks. spot moves off track.
objective lens
scan direction
modulation due to
the data pattern
brightness variation due to

spot position error across
the track
servo optics
detectors
amplifier
A beam
+ splitter
TES C
- B
reflected light
TES = A - B
objective
lens
position data marks
error (pits)
leading spot
scan
data spot
direction
trailing spot
track pitch position

error

detectors servo optics
amplifiers
A
beam
+ D
FES + splitter
+
- B
+
+ C
reflected light
FES = (A + C) – (B + D)
D A objective
lens data marks
C B
too close in focus too far

FES scan spot
direction
position
error

marks
high slope
(small spot)
detector
signal
threshold low slope

level (large spot)
perfect transition
(no noise)
channel bit
window
reference
clock
data bits 1 0 1 0
channel bits 0 1 0 0 1 0 0 1 0

low slope high slope
(large spot) (small spot)
amplitude
variation σN
threshold due to noise
level
width variation σW (small spot)

width variation σW (large spot)
adjacent bit ideal channel adjacent bit
window bit window window

001000100100100000000001000000100001001000000100010000 (2,10) EFM coded data
laser pulse signal
data mark pattern

disk - disk -
single layer multiple layer volumetric ribbon
data
layer(s)
scan
spot
substrate
substrate
dust
focused
beam
single flat multiple 3D volume flexible
surface planes information surface
(side view) (side view)

in focus 0.5 µm defocus
λ = 650 nm NA = 0.85 λ = 650 nm NA = 0.85
90%
50%
13.5%
in focus 0.5 µm defocus

λ = 405 nm NA = 0.60 λ = 405 nm NA = 0.60

marginal light from
ray laser
objective lens
θ SIL in near contact
with the recording
layers
θ′
n
data
v

Laser Beam Detail of
Recording Layers
PC substrate
SiN 60 nm
Objective GdFeCo 20 nm Expansion

Lens SiN 20 nm
Expansion
Expanded Layer
Domain
TbFeCo 200 nm Recording
Recording
Layer
Recorded SiN 60 nm
Domain
External Field

View publication stats

Optical Data Storage

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Optical Data Storage

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Optical Data Storage

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Optical Data Storage

The user has requested enhancement of the downloaded file.

packaging, reliable information retrieval, and cost-effective production.

1.0 Inspiration for the Invention

Author: Tom D. Milster (Prepared for Laser Handbook) 1

used to produce low-cost replicas, which are then sold to consumers.

2.0 Data Marks and Spaces: The Information Carriers

compared to a CD, the DVD can hold more data.

3.0 Optical Data Storage Principles

Author: Tom D. Milster (Prepared for Laser Handbook) 3

Inexpensive replicas of the master are made with injection-molding equipment.

output data stream.

4.0 Commercial media

Author: Tom D. Milster (Prepared for Laser Handbook) 4

technologies are reviewed, and commercial formats are listed.

Pit-type data mark technology for CD-read-only-memory (ROM) and DVD-ROM

Section 5.6, which is optimized at a slightly different pit depth.[6]

Author: Tom D. Milster (Prepared for Laser Handbook) 5

DVDs may be more sensitive to contamination than CDs.

Dye-polymer or dye-monomer technology is used in CD-R products. Dye

data marks once. Data marks cannot be erased and rewritten.

Author: Tom D. Milster (Prepared for Laser Handbook) 6

metallized azo, and phthalocyanine. At a laser wavelength to 0.78 micrometers, where CD

reflective layer is usually silver or gold instead of aluminum.

Author: Tom D. Milster (Prepared for Laser Handbook) 7

degradation. CD-RW’s use phase-change technology, which is based on differences of the

separate the thin films from the environment.

Author: Tom D. Milster (Prepared for Laser Handbook) 9

degradation in performance.[21, 22]

Author: Tom D. Milster (Prepared for Laser Handbook) 10

Adding layers increases the total capacity of the disk.

reviews basic concepts necessary to understand how these systems work.

5.1 Data density and spot size

(0.65)(8)/14.5 = 0.36Gb-in-2 , where one byte = 8 bits.

decreases. A defocus distance ∆z of only 3 micrometers dramatically reduces peak irradiance

produce a spot size of 1.1 micrometers.

5.2 Thermal recording

Author: Tom D. Milster (Prepared for Laser Handbook) 12

blooming, and is a serious problem, especially in mageto-optic systems, if not corrected by

5.3 Frequency response and equalization

Author: Tom D. Milster (Prepared for Laser Handbook) 14

decoding.[25] An example of an equalized modulation transfer function is shown in Fig. 11,

5.4 Effects of defocus

device performance. However, as defocus increases beyond ∆z = 0.5 micrometers, peak

Author: Tom D. Milster (Prepared for Laser Handbook) 15

5.5 Servo optics

for CDs and 0.13 micrometers for DVDs.[26, 27]

Author: Tom D. Milster (Prepared for Laser Handbook) 16

Author: Tom D. Milster (Prepared for Laser Handbook) 17

in portable disk players.

5.6 Feedback sensors

following TES and FES sensors are described:

Push-pull tracking is a method to provide a tracking error signal using a groove

Three-spot tracking is a method to provide a tracking error signal by detecting the

Astigmatic focusing is a method to provide a focus error signal by fabricating a small

amount of astigmatism into the servo optics.