Beads

Texture synthesizer
About Beads
Once upon a time there was Clouds. Then came the day to
clean up the mess.
Beads is a granular audio processor. It creates textures and
soundscapes by playing back layered, delayed, transposed
and enveloped fragments of sound (“grains”) taken continu-
ously from the incoming audio signal.

Installation
Beads requires a -12V/+12V power supply (2x5 pin connec-
tor). The red stripe of the ribbon cable (-12V side) must be
oriented on the same side as the “Red stripe” marking on
the module and on your power distribution board.
The module draws 100mA from the +12V rail, and 10mA
from the -12V rail.

Online manual and help

The full manual can be found online at
mutable-instruments.net/modules/beads/manual
For help and discussions, head to
mutable-instruments.net/forum

Please refer to the online manual for detailed infor-

mation regarding compliance with EMC directives
Beads in a nutshell

One way of picturing how Beads operates is to imagine a

tape loop, on which incoming audio is continuously recorded.
Every time you request a grain to be played (in reaction to
a trigger, a button press, periodically, or randomly), a new
replay head positions itself along the tape.
If this replay head doesn’t move, the audio will be played
back at the original pitch and speed, but if it moves closer
to, or further away from the record head, the signal will be
replayed at a different speed and pitch. This replay head has
its own amplitude envelope, and it will leave the tape once
the envelope has reached a null amplitude.
Now imagine up to 30 replay heads flying along the tape.
Imagine you can stop the incoming audio from being record-
ed on the tape so that all these little replay heads can freely
move along and gather sounds. And there’s a reverb...

Beads does not use tape, but RAM. In this manual we use
computer-science terminology and refer to this virtual
piece of tape as a recording buffer.
Recording quality and audio input

Recording quality is chosen with the selector button [A].

Buffer length
LED Quality Rate Resolution Mono Stereo
Bright digital 48 kHz 16-bit 8s 4s
Cold digital 32 kHz 12-bit 16s 8s
Sunny tape 24 kHz 12-bit 20s 10s
Scorched 24 kHz 8-bit 32s 16s
cassette

• The Cold digital setting most accurately reproduces the

sonic character of the late Mutable Instruments Clouds.
• The Sunny tape setting runs the dry audio signal at a
bright and clean 48kHz.
• The Scorched cassette setting emulates wow and
flutter.
 2
3

Beads operates in mono or stereo depending on whether

one, or both, of the audio inputs (1) are patched.
When patch cables are inserted or removed, Beads monitors
for five seconds the level of the incoming signal and adjusts
the input gain accordingly, from +0dB to +32dB. The input
level LED (2) blinks during this adjustment process. The
input gain is chosen to leave some headroom, but in case of
big level changes, a limiter kicks in.
One can manually restart the gain adjustment process by
pressing and holding the audio quality selector button [A]
for one second. Holding this button [A] while turning the
feedback knob allows manual gain adjustments.
The manually-set gain is memorized and applied until a
long press on [A] re-enables automatic gain control.

The FREEZE latching button [B] and the corresponding gate

input (3) disable the recording of the incoming audio signal
in the buffer. Otherwise, Beads records continuously!
If FREEZE remains engaged for more than 10 seconds, the
content of the buffer is backed up, and will be restored the
next time the module is powered on.
Beads will not switch between stereo and mono operation,
or change the recording quality, while FREEZE is engaged.
Grain generation
Latched
Latched grain generation is enabled by holding the SEED
button [C] for four seconds, or by pressing the FREEZE but-
ton [B] while the SEED button [C] is being held. This is also
the default setting when the module is powered on.
The SEED button remains illuminated, and its brightness is
slowly modulated to indicate that latching is enabled.

B C
D

4
In this mode, the grains are generated continuously, at a
rate set by the DENSITY knob [D] and modulated by the
DENSITY CV input (5).
At 12 o’clock, no grains are generated. Turn DENSITY CW
and grains will be generated at a randomly modulated rate,
or CCW for a constant generation rate. The further you turn,
the shorter the interval between grains, reaching at the
extreme the period of a C3 note.

Clocked
When latched grain generation is enabled, and when a
signal, such as a clock or sequence, is patched into the
SEED input (4), the DENSITY knob [D] is repurposed as a
divider or probability control. At 12 o’clock, no grains are
generated. Turn CW to increase the probability (from 0% to
100%) that a grain is triggered by the external signal. Turn
CCW to increase the division ratio, from 1/16 to 1.

Gated and triggered

Disable latched grain generation with a short press on the
SEED button [C].
Grains will then be generated only when the SEED button is
held, or when a gate signal patched in the SEED input (4) is
high. The DENSITY knob [D] controls the repetition rate of
grains. When DENSITY is at 12 o’clock, only a single grain
will be played at each press of the SEED button, or at each
trigger sent in the SEED input (4).

When grain density reaches audio rates, the DENSITY CV

input (5) applies exponential FM on this rate, with a scale
of 1V/octave.
Grain playback control
Four parameters control at which buffer position, pitch, and
with which duration and envelope the grains are replayed.
More precisely, these parameters and their respective
modulations are read once, whenever a grain starts, and
remain unchanged throughout the duration of the grain. If a
parameter then changes, it will only impact the next grain.
For example, turning the PITCH knob will create a trail of
grains with different pitches, rather than change, in lock-
step, the pitch of all the grains that are currently playing.

E F

G H

6
E. TIME controls if the grain replays the most recent (fully
CCW) or oldest (fully CW) audio material from the recording
buffer – shifting the replay heads further apart from the
record head.

Beads does not make use of any time-travel

technology: if you request a grain to be played at double
speed, one second away from the start of the buffer, the
grain will fade out and stop after 0.5s of playback, once
the replay head bumps into the record head. (Suggested
reading: “Light cones in tape recorder cosmology”).

F. PITCH controls the transposition, from -24 to +24 semi-

tones, with virtual notches at selected intervals.
G. SIZE controls the duration and playback direction of the
grain. At the 11 o’clock position, a very short (30ms) grain
is played. Turn CW to increase grain duration up to 4s. Turn
CCW to play a reversed grain, lasting up to 4s.
Turning SIZE fully clockwise (∞) generates never-ending
grains acting as delay taps. Please refer to the “Beads as a
delay” section.
H. SHAPE adjusts the amplitude envelope of the grain.
Fully CCW creates clicky, rectangular envelopes, while fully
CW provides envelopes with slow attacks reminiscent of
reversed grains (Please note, however, that the envelope
shape is independent of the playback direction).
I. Attenurandomizers for the TIME, SIZE, SHAPE and PITCH
parameters. They control the amount of external CV modu-
lation on the corresponding parameters, or repurpose the
CV input (6) as a randomization or “spread” control.
Attenurandomizers
When a cable is patched into the corresponding CV input
(6), turning the attenurandomizer [I] CW from 12 o’clock
increases the amount of external CV modulation. Turning it
CCW increases the amount of CV-controlled randomization.

CV

With no CV patched into an input, the attenurandomizer con-

trols the amount of randomization from an independent in-
ternal random source with a peaky (12 o’clock to fully CCW)
or uniform (12 o’clock to fully CW) distribution. The random
values from the peaky distribution are clustered towards the
middle, with extreme values being infrequently generated.
Patch ideas
• Patch a ramp-down LFO, or a decaying linear envelope
into the TIME CV input for “scrubbing” the buffer, or a
segment of it, at whichever speed the LFO rate or enve-
lope time is set to. Timestretching time!
• The PITCH CV input tracks V/O when the attenurandom-
izer is turned fully CW: one can sequence a melody of
grains or even play them from a keyboard.
• Patch a fast arpeggiated sequence into the PITCH CV
input to create chords: each grain will be played at a
randomly chosen note of the arpeggio.
• Sequence slices of sound (or phonemes from a record-
ing of speech) by patching the CV output of a sequencer
into TIME, and its gate output into SEED.
Mixing and audio output
Beads’ signal flow is as follows:

grains

IN reverb OUT
feedback
dry/wet
crossfade

J K

M L

8
J. Feedback, that is to say the amount of output signal
mixed with the input signal and fed back into the processing
chain. Each quality setting employs a different feedback am-
plitude limiting scheme typical of the medium it emulates –
from clean brickwall-limiting to grungy tape saturation.
K. Dry/wet balance.
L. Amount of reverb. Modeled on the acoustics of
Thoreau’s cabin, or of a strip-mall spa.
The LED under each of these knobs indicates the amount of
modulation they receive from the assignable CV input (7).
Press the button [M] to select to which of these 3 destina-
tions the CV input (7) is assigned. Or hold this button and
turn the knobs [J], [K] and [L] to individually adjust the
amount of CV modulation.
8. Audio output. While the recording buffer can be mono
or stereo, Beads’ signal processing chain is always stereo.
If the R output is left unpatched, both L and R signals are
summed together and sent to the L output.

If one of the grains’ parameters is randomized, or if

the grains are generated at a random rate, their pan
position will also be randomized.

Hold the button [M] and press the SEED button [C] to enable
(or disable) the generation of a grain trigger signal on the R
output. A patch cable will have to be inserted in the R output
for this to work without affecting the L output!
Beads as a delay
Setting the grain SIZE [G] knob fully clockwise (∞) turns
Beads into a delay or beat slicer. Effectively, only one grain
remains active, forever, continuously reading from the tape.
The base delay time (and slice duration) can be manually
controlled, tapped, or set by an external clock.

Manual control
If the SEED input (4) is left unpatched, and if the SEED but-
ton [C] is latched (slowly fading in and out), the delay time is
freely controlled by the DENSITY knob [D] and CV input (5).
At 12 o’clock, the base delay time corresponds to the full
buffer duration. Turn the knob further away to shorten the
delay time up to audio rates, for flanger or comb-filtering
effects. From 12 o’clock to fully CW, the delay will have an
additional, unevenly spaced, tap.

full buffer full buffer

DENSITY

Clocked or tap-tempo control

If an external clock is patched into the SEED input (4), or if
you rhythmically tap the SEED button, the base delay time
will be set as the interval between the taps or clock ticks.
The DENSITY knob [D] selects a subdivision of this duration.
Turn the knob further away from 12 o’clock to use shorter
subdivisions. From 12 o’clock to fully CCW, only binary sub-
divisions will be used. From 12 o’clock to fully CW, a wider
variety of ratios are available.

1/1 1/2
1/2 1/3
1/4
1/4 1/6
1/8
1/8 1/10
1/12
1/16 1/16
DENSITY

Delaying or slicing
When FREEZE [B] is not engaged, Beads operates as a
delay. The TIME knob [E] selects the actual delay time, as
a multiple of the base delay time set by DENSITY and/or by
the external clock or taps.
When FREEZE [B] is engaged, a slice from the recording
buffer is continuously looped. The duration of a slice is equal
to the base delay time. The TIME knob [E] selects which
slice is played.
The SHAPE knob [H] applies a tempo-synchronized enve-
lope on the repeats. For normal operation, turn it fully CCW.
PITCH [F] applies a classic rotary-head pitch-shifting effect
on the delayed signal. At 12 o’clock, the pitch-shifter is
bypassed.
Slow random LFOs are internally routed to the attenuran-
domizers [I].
Beads as a granular wavetable synth
When both audio inputs (1) are left unpatched, and at the
end of a period of ten seconds, Beads loses patience and
granularizes a collection of internally stored buffers of raw
waveforms from Mutable Instruments Plaits’ wavetable
model.

The feedback control [J] selects which one of these 8 banks

of waveforms is played.
The dry/wet control [K] adjusts the balance between the
continuous oscillator signal, and the granularized signal.
The FREEZE button [B] halts the envelope of the grains, and
stops the generation of new grains.
The audio quality selector [A] selects the output resolution.
LED Resolution LED Resolution
16-bit 16-bit dry, 4-bit wet
7-bit 4-bit

Finally, the PITCH CV input always acts as a 1 V/octave CV

input affecting the root note of the grains, irrespectively of
the position of the PITCH attenurandomizer.
The PITCH attenurandomizer always controls the amount of
pitch randomization of the grains.