Monsoon

Texture synthesizer
 Monsoon by jakplugg 28
Table of Contents Microcell / Storm / nanoCell
Typhoon by jakplugg
29
29

About Monsoon 6 Parasite 30

Why Monsoon over Clouds? 7
Power requirements 7 Overview 31
Switching modes with Parasite 32
Granular synthesis 8 Important warnings 32
How Monsoon does granular synthesis 8
Notable changes 33
Front panel 10 Granular processor 33
Controls 10 Looping delay and pitch-shifter 33
Inputs and outputs 14 Looping delay 34
Calibration procedure 35
Advanced topics 16
Miscellaneous 35
Audio quality 16
Saving and loading buffers 17 Oliverb 36
Calibration procedure 18 Controls 36
Updating or replacing firmware 19
Resonestor 40
The infamous alternate modes 20 Controls 41
Pitch-shifter/time-stretcher 21 Inputs and outputs 43
Spectral madness 22
Looping delay 24 Kammerl Beat Repeat 46
Tips and patch ideas 25 Overview 47
Switching modes with KBR 48
Understanding the Clouds derivatives 26 Important warnings 48
Clouds by Mutable Instruments 26 Reverb change 49
Supercell by grayscale 27
μBurst (microBurst) by jakplugg 28 Kammerl beat repeat mode 50

2 3
 Controls 50
Inputs and outputs 54

Spectral clouds mode 56

Controls 56
Inputs and outputs 58

Appendix 60
Credits, licensing, and honorable mentions 60
Contact 61
Disclaimer 61
Selling or purchasing this guide 61

This page is unintentionally left blank.

4 5
 About Monsoon Why Monsoon over Clouds?
Monsoon’s primary difference lies in Clouds’ four blend
Monsoon is a granular audio processor based on the open-
mode parameters: Wet/dry mix, stereo panning, feedback
source hardware and firmware for Clouds, a discontinued
amount, and reverb amount. While Clouds provides a single
Eurorack module by Mutable Instruments. It creates
knob and a blend-mode-switch button, Monsoon gives each
textures and soundscapes by combining multiple
blend parameter its own knob and its own CV modulation
overlapping, delayed, transposed, and enveloped segments
input.
of sound, taken from an audio recording buffer.
The four main control parameters, POSITION, DENSITY,
Clouds differs from other granular Eurorack modules in that
SIZE, and TEXTURE, are changed from knobs to sliders. The
it focuses on the real-time processing of audio sources from
slider knob contains a red/green LED which displays the
your modular system itself, rather than the playback of pre-
corresponding CV modulation value.
recorded samples from a storage device. It rewards the
freezing, fragmentation and dissolution of the unexpected Additionally, Monsoon cuts the size of the module from 18HP
instant, rather than the careful planning of what might very down to 12HP.
well not come. You can read more in the “Understanding the available
You can find more information on Clouds at Clouds derivatives” section.

mutable-instruments.net/modules/clouds
Power requirements
Monsoon requires a standard Eurorack power supply which
can supply both +12V and -12V (2x5 pin connector). The red
stripe on 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 approximately 120mA from the +12V rail,
and 10mA from the -12V rail.
Like Clouds, Monsoon does not require +5V power.

6 7
 Granular synthesis recorded audio data, the module synthesizes a sonic texture
by playing back short, overlapping segments of audio
Granular synthesis is a sound synthesis method based on (grains) extracted from it.
the same principle as sampling. However, the samples are Monsoon allows you to control the buffer location from
instead split into small pieces of around 1 to 50 which grains are taken (POSITION), the length of the grains
milliseconds. These pieces are called grains. Multiple grains (SIZE), the speed/pitch of the grains when they are re-played
may be layered on top of each other, and may play at (TUNE), the amount of overlap between grains (DENSITY),
different speeds, phases, volume, and frequency, among whether the distribution of grains in time is constant or
other parameters. random, and the envelope curve which is applied to the
At low playback speeds, the result is a kind of soundscape, grains, giving the impression of a rough or smooth TEXTURE.
often described as a cloud. At high speeds, the result is In addition, to create textures with a blurry feel, a diffuser (a
heard as a note or notes of a novel timbre. By varying the network of all-pass filters, like a reverb without a tail) can
waveform, envelope, duration, spatial position, and density be applied.
of the grains, many different sounds can be produced. The module plays grains continuously, at a rate determined
Both have been used for musical purposes: as sound by the DENSITY and SIZE settings. A TRIGGER input is also
effects, raw material for further processing by other present, to explicitly instruct the module to start the
synthesis or digital signal processing effects, or as complete playback of a new grain. The maximum number of
musical works in their own right. Conventional effects that concurrent grains is quite large - between 40 and 60. This
can be achieved include amplitude modulation and time specificity brings Monsoon closer to the roots of granular
stretching. More experimentally, stereo or multichannel synthesis, and allows the synthesis of varied textures even
scattering, random reordering, disintegration and morphing from basic waveforms - there are indeed many more
are possible. dimensions to granular synthesis than keeping a playback
pointer moving through an SD card sample!
How Monsoon does granular synthesis It is possible, at any time, to FREEZE the audio buffer from
Monsoon continuously records the incoming audio signals which the grains are taken. In this case, the incoming audio
into a short amount of sample memory. While recording is no longer recorded. Somehow, Monsoon is the exact
time varies from 1 to 8 seconds by reducing the audio opposite of a sampler; by default, the modular always
quality, you ought to feel very guilty every time you think of samples the audio it receives, except when it is in the frozen
this as tape - think of it as a space; a room. Using this state.

8 9
 Front panel C
A B
Controls DISP

A. DISPLAY button. A single press selects the display mode

of the four LED bar [C] at the top of the module. By default,
it will show the input signal [6] level. The alternate mode
displays colors which represent the current value of the four D
corresponding blend parameter knobs [J-M]. Holding the
button for one second will allow selection of the current TUNE
audio quality level (see the Audio quality section).
B. LOAD/SAVE button. Enables saving and loading. See the E
Saving and loading buffers section. Some panels may label
this as WRITE.
INPUT LEVEL
C. Indicator LEDs. In their default mode, they act as an input POSITION DENSITY SIZE TEXTURE

VU-meter. When FREEZE [N] is active, they monitor the

output level. Soft-clipping occurs when the last LED is on.
When in their secondary mode, each LED displays a color
which represents the current value of the corresponding
WET/DRY STEREO FEEDBACK REVERB
blend parameter knob [J-M]. They can also indicate the
audio quality setting (red). Lastly, when loading or saving
buffers, the LEDs are used to display the memory bank
being selected.
D. TUNE knob. Controls the playback frequency of the audio
buffer. At 12 o’clock, the buffer is played at its original V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE

frequency. Some panels may label this as PITCH.

E. INPUT LEVEL knob. Controls the input audio [6] gain
L IN R FREEZE L OUT R
level, from -18dB to +6dB. Some panels may label this as
IN GAIN.

10 11
 F. Grain POSITION slider. Selects the part of the audio
buffer from which audio grains are selected. Move the slider
upwards to travel back in time.
DISP
G. Grain DENSITY slider. At center, no grains are generated. F G H I
Move it upwards and grains will be sown randomly,
downwards and they will be played at a constant rate. The
further from center, the higher the grain overlap.
H. Grain SIZE slider. The center value has no particular
significance. Moving the slider upwards selects bigger TUNE
grains, and downwards, smaller.
I. Grain TEXTURE slider. Morphs through various shapes of
grain envelopes: square (boxcar), triangle, and then Hann
window. At a point shortly above the center, a diffuser is
INPUT LEVEL
activated, which smears transients.
POSITION DENSITY SIZE TEXTURE
J. WET/DRY knob. Fully counter-clockwise passes only the
J K L M
input signal to the output, fully clockwise passes only the
wet signal.
K. STEREO knob. Controls the amount of random panning
WET/DRY STEREO FEEDBACK REVERB
applied to the grains. Fully counter-clockwise performs no
stereo panning. Some panels may label this as PAN.
L. FEEDBACK knob. Controls the amount of the output
signal that gets fed back into the input audio buffer. Has no
effect when FREEZE [N] is enabled.
V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE
M. REVERB knob. Controls the amount of reverb applied to
the input.
N. FREEZE button. This latching button stops the recording L IN R FREEZE L OUT R
of incoming audio. Grain sampling is then performed on the
N
last few seconds of audio kept in the module’s memory.

12 13
 Inputs and outputs
1. Blend control CV inputs. Each of the four inputs
corresponds to the blend control parameter knob [J-M] DISP

above it.
2. V/OCT CV input. Grain transposition (TUNE knob [D]) CV
input, with 1V per octave response.
3. TRIGGER input. Generates a single grain. By moving the
grain DENSITY slider [G] to center, and sending a trigger to
TUNE
this input, Monsoon can be controlled like a micro-sample
player.
4. Slider parameter CV inputs. Each of these CV inputs
correspond to a slider parameter [F-I]. The slider knob
LEDs will turn red or green based on the CV sent to these INPUT LEVEL
inputs. POSITION DENSITY SIZE TEXTURE

5. FREEZE gate input. When the gate signal sent to this

input is high, FREEZE [N] is enabled.
6. Stereo audio inputs. The right channel is normalled to the
left channel, so patching only the left channel input will also WET/DRY STEREO FEEDBACK REVERB
send the signal to the right channel. This aids in processing
mono signals, and allows generating true stereo audio from 1
a mono source.
7. Stereo audio outputs. The right channel is normalled to 2 3
4
the left channel, so patching only the left channel input will
V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE
cause the module’s stereo output to be mixed down to a
5
mono output. 6 7
L IN R FREEZE L OUT R

14 15
 Advanced topics Saving and loading buffers
Up to 4 frozen audio buffers can be saved and reloaded. The
Audio quality audio data, along with the current audio quality setting, are
Monsoon can only keep so much recorded audio in its buffer, saved to one of the module’s internal memory banks.
so four audio quality modes are provided, which vary in their To save the current buffer to memory:
sample rate, resolution, number of channels. As a result,
1. Hold the LOAD/SAVE button for one second.
the memory buffer is able to hold a different number of
seconds for each mode. 2. Press the DISPLAY button repeatedly to select one of the
four memory slots, indicated by a blinking red LED.
Note that Monsoon’s 8-bit resolution modes use a lovely
flavor of 8-bit called μ-law companding. It sounds like a 3. Press the LOAD/SAVE button to confirm.
cassette - less hiss, more distortion.
To load a buffer from permanent memory:
Mode Rate Resolution Channels Buffer length 1. Press the LOAD/SAVE button.
32kHz 16-bit stereo 1 second 2. Press the DISPLAY button repeatedly to select one of the
32kHz 16-bit mono 2 seconds four memory slots, indicated by a blinking green LED.
16kHz 8-bit μ-law stereo 4 seconds 3. Press the LOAD/SAVE button to confirm.
16kHz 8-bit μ-law mono 8 seconds If you press the LOAD/SAVE button by mistake, do not press
any button for a few seconds, and the module will reset.

16 17
 Calibration procedure Updating or replacing firmware
The module, depending on who manufactured it, is likely to The process of updating your module’s firmware, or
have been factory-calibrated using precision voltage replacing it entirely with an available alternate firmware, is
sources. Follow this procedure only if you want to the same as a genuine Clouds module. To do so:
compensate for inaccuracies in your CV sources, or if your 1. Power off and unplug all cables from the module.
module has lost its calibration settings following a fault or
2. Connect the output of your computer’s audio interface
the installation of alternative firmware.
into the left channel audio input [6].
1. Disconnect all CV inputs.
3. Power on the module while holding FREEZE [N]. The
2. Connect the note/pitch CV output of a well-calibrated button should blink.
keyboard interface or MIDI-CV converter to the V/OCT input.
4. Make sure that your computer will make no additional
3. Press the LOAD/SAVE button, and while holding it down, sound during the procedure, and that your speakers are not
press the DISPLAY button. The first two LEDs will blink connected to your audio interface - the noises emitted
orange. during the procedure are aggressive and can harm your
4. Play a C2 note, or send a 1V voltage from your CV source. hearing.
5. Press the LOAD/SAVE button. The four LEDs will blink 5. Play the firmware audio file on the computer, into the
orange. module. While the module receives data, the LEDs will show
6. Play a C4 note, or send a 3V voltage from your CV source. the signal level. Signal reception is optimal when two or
three LEDs are lit. You can use the INPUT LEVEL [E] knob to
7. Press the LOAD/SAVE button.
control the signal level. When the end of the audio file is
reached, the module automatically restarts - if not, start
again.
6. In case the signal level is inadequate, the LEDs will blink
in red. Adjust the INPUT LEVEL [E], and press the
FREEZE [N] button to retry. If this does not help, retry the
procedure from another computer/audio interface, and
make sure that there are no extra pieces of equipment
(equalizers, FX processors) in the signal chain.

18 19
 The infamous alternate modes Pitch-shifter/time-stretcher
This mode is quite similar to the granular mode, except that
In its tumultuous teenage years, Clouds tried to be
it uses two overlapping grains, synchronized with the most
everything, including a delay/pitch-shifter, a spectral
salient period of the sound. The grains are carefully spliced
processor, a projectionist, and a cab driver in Rouen.
so that they mesh well with each other (“deglitching”).
This experimental code is still available in the module, by
A. Modulating POSITION when recording is frozen will scrub
pressing the DISPLAY button for 5 seconds until one of the
through the audio buffer.
LEDs is orange, and then repeatedly pressing the button to
select one of the four functions. B. The DENSITY slider creates a granular diffusion effect
based on all-pass filters.
LEDs Mode name C. The SIZE slider controls the size of the overlapping
Granular processor (normal operation) windows used for pitch-shifting and time-stretching - from
Pitch-shifter / time-stretcher an extremely grainy “drilling” sound, to smooth bits of loops.
Looping delay D. The TEXTURE slider acts as a low-pass/high-pass filter.
Spectral madness 5. Sending a trigger to the TRIGGER input creates a clock-
synchronized loop (when FREEZE is enabled) or stuttering
What follows is a description of the three (2 through 4) effect - equivalent to applying a tempo-synchronized
hidden alternate modes. These features are highly decaying envelope on the POSITION parameter.
experimental, and the module controls might be reassigned
to functions very different from the label on the front panel.
It is not recommended to use these alternate modes if you A B C D
plan to let Tony Rolando come play with your system.
Each mode description will include a diagram which
includes only those inputs and outputs which are different 1
from the normal granular processor mode.
TRIGGER

POSITION DENSITY SIZE TEXTURE

20 21
 Spectral madness that determines how frequencies are mapped between the
analysis and synthesis buffers. It’s like a single-slider GRM
In this mode, the incoming signal is converted into “frames”
Warp. Over the course of the slider, it’ll do spectral shifting,
of spectral data, that are stored, transformed, and re-
but also spectral reversal.
synthesized as a time-domain signal.
D. The TEXTURE slider does two things. Below center, it
A. The POSITION slider selects the buffer into which the
increasingly quantizes the amplitudes of the spectral
audio is poured (when FREEZE is not active), or from which
components, like a very low-bitrate audio file. Above center,
buffer the audio is synthesized (when FREEZE is active).
it increasingly weakens the strongest partials and amplifies
Depending on the quality settings, there are between 2 and 7
the weakest ones. This has the effect of making the
buffers laid out over the course of the POSITION slider. What
spectrum more noise-like.
the module does is crossfade between a “wavetable” of fast
E. The TUNE knob controls the transposition (pitch-shifting).
For example, set POSITION to its minimum value. Press FREEZE. 1. Sending a trigger to the TRIGGER input creates various
You get a first texture. Set POSITION to its maximum value. frequency-domain glitches typical of corrupted (encoded)
UNFREEZE. Wait for something else to happen in the incoming audio files. It works as a kind of build-up/feedback effect -
audio. Press FREEZE again. By moving POSITION, you the shorter the pulse, the smaller the effect. With a
interpolate between the two textures which had been captured at continuous gate, it’ll really start to rot...
the press of FREEZE.

Fourier transform (FFT) slices.

E A B C D
B. The DENSITY slider determines how results from the
analyzer are passed to the re-synthesizer. Below center,
there’s an increasing probability that a given FFT bin won’t
get updated, causing a kind of partial freeze. Above center,
adjacent analysis frames are increasingly merged together TUNE

(like a low-pass filter in the amplitude for each frequency

1
bin). At extreme settings, random phase modulation is
applied to smooth things, giving you different flavors of TRIGGER
POSITION DENSITY SIZE TEXTURE
spectral muddling/reverb.
C. The SIZE slider changes the coefficients of a polynomial

22 23
 Looping delay Tips and patch ideas
This mode continuously plays back audio from the buffer
1. If you need a rapidly changing noise source to randomize
without any kind of granularization.
grain position or pitch, try using one of the audio outputs
A. The POSITION slider controls the delay time. Modulating through a multiple.
POSITION will create effects similar to vinyl scratching or
2. Very dense clouds sound best when at least one
manual manipulation of tape. When FREEZE is activated, it
parameter receives random modulations. Otherwise, the
controls the loop start position.
identical echoes created by the repeating grains will sound
B. The DENSITY slider has no effect unless FREEZE is like a feedback comb filter.
activated, where it creates a granular diffusion effect, based
3. Raw material like sawtooth or sine waves sound very
on all-pass filters.
good, especially with heavy random modulation. Experiment
C. The SIZE slider controls the size of the overlapping with capturing many short waveforms and using them as a
windows used for pitch-shifting - slider at its minimum for a kind of wavetable.
smooth result that might smear transients, at its maximum
4. Contact microphones, touch controllers or random
for a grainy, almost ring-modulated sound. When FREEZE is
sources are Monsoon’s best CV friends!
activated, it controls the loop duration.
1. Sending a trigger to the TRIGGER input creates a clock-
synchronized stuttering loop. Otherwise, the period of the
trigger pulses sets the delay time.

A B C

1
TRIGGER

POSITION DENSITY SIZE TEXTURE

24 25
 Understanding the Clouds derivatives Supercell by grayscale
Supercell doubles down on a highly-accessible and feature-
Since Clouds’ unexpected discontinuation in 2017, a number
rich 34HP design. The massive list of new features includes:
of implementations of Clouds’ open-source and open-
hardware platform have come to market. • Potentiometers, CV inputs, and attenuverters for all four
blend parameters
Over time, the number of implementations has grown to the
point where it can be difficult to keep track of the differences • Mute switches, stereo VCAs, and separate VU meters
between them. This section aims to give a short, general for both audio input and output
description of each one. • Stereo output gain control, allowing for +6 dB additional
Each hardware variation also supports a different set of gain to compensate for low output level patches
firmware distributions. The specifics of each of these • Dedicated switches and LEDs for switching audio quality
firmware versions will be discussed in a later section. and loading/saving buffer presets
Additionally, because the naming of the front panel controls • Extensive modulation options for all parameters,
is not standardized, you might also find that different including direct patching of external CV, distribution of
manufacturers have different names for some controls. one CV to all parameters via the AUX input, or an
internal smooth random CV that is normalled to any
Clouds by Mutable Instruments unpatched inputs when AUX is also unpatched
Released in 2015, Clouds has become a Eurorack mainstay. • Separate exponential V/OCT input and linear PITCH
The module was discontinued in 2017. While still an input
excellent module, a significant pain point of the module is • A manual TRIGGER switch
the fact that all four “blend parameters” are controlled with
• A more powerful microcontroller which allows loading
a single knob, switched between with a mode button. This
of larger, more comprehensive firmware variants
also means that there is only a single CV input for
modulating these parameters. If space is no issue, Supercell is the module for you. These
additional features, plus the aforementioned microcontroller
Supported firmwares: Stock, Parasite, and Kammerl Beat
change, mean that Supercell uses its own firmware variant.
Repeat
Supported firmwares: Supercell Core (stock),
Size: 18HP
SuperParasites (combined Parasite and Beat Repeat)
Size: 34HP

26 27
 μBurst (microBurst) by jakplugg Microcell / Storm / nanoCell
μBurst is a miniature version of Clouds. It adds no All technically the same module, Microcell is a minified
functionality, but it manages to decrease the size of the version of Supercell, coming in at 14HP. Microcell retains all
module from 18HP to 8HP. functionality of the Supercell except for the CV input
The smaller layout has the tendency to feel cramped due to attenuverters. Microcell uses the same firmware as
the densely-packed mini potentiometers. Nonetheless, it is Supercell.
a very capable module and recommendable to anyone for The issue with the name stems from the fact that Microcell
whom space is at a premium. is a trademarked name. Depending on where you buy the
Supported firmwares: Stock, Parasite, and Kammerl Beat module, it might be named Storm (After Later Audio) or
Repeat nanoCell (Momo Modular).

Size: 8HP Supported firmwares: Supercell Core (stock),

SuperParasites (combined Parasite and Beat Repeat)
Monsoon by jakplugg Size: 14HP
The result of taking their μBurst design and adding some
welcome features, jakplugg designed Monsoon - the Typhoon by jakplugg
primary subject of this guide. Taking another step in the μBurst and Monsoon lineage,
With the size increased to 12HP (still 6HP smaller than jakplugg designed Typhoon similarly to Microcell - a
Clouds), Monsoon features individual pots and CV inputs for minified Supercell. Typhoon, like Microcell, omits the input
all four blend parameters. It also opts for sliders with CV CV attenuverters found on Supercell. Following the choices
indicator LEDs instead of potentiometers for the four made on Monsoon, Typhoon also includes the friendly four-
primary module parameters. Due to these changes, slider layout. Typhoon uses the same firmware as Supercell.
Monsoon requires slightly modified firmware. Supported firmwares: Supercell Core (stock),
Supported firmwares: Monsoon Stock, Monsoon Parasite, SuperParasites (combined Parasite and Beat Repeat)
Monsoon Beat Repeat Size: 16HP
Size: 12HP

28 29
 Overview
The Parasite firmware is an alternate firmware for Clouds.
This guide will be focusing on Monsoon, but the principles
are functionally the same.
NOTE: You must re-calibrate the module’s tuning after you
install Parasite. See the Calibration procedure section of the
Parasite firmware section.

Parasite Parasite adds two new operation modes:

• Oliverb, a full-featured and CV-controllable modeless
reverb with some twists.
Firmware
version 2.0.1 for Monsoon • Resonestor, a polyphonic resonator processor with
built-in support for Karplus-Strong synthesis. Plucked
or fed with your favorite sound, it spits out strange
guitar sounds, lush chords, evolving drones, and more.
In addition to these new modes, Parasite:
• Adds more envelopes for your grains, accessible
through the TEXTURE slider.
• Increases the available range of grain sizes by allowing
for much smaller grains.
• Adds asymmetrical grain envelopes (saws, ramps).
• Enhances the built-in “experimental” pitch-shifter/
time-stretcher and looping delay modes with a number
of new features.

30 31
 Switching modes with Parasite Notable changes
To access the mode selector, you now have two options:
1. Press the DISPLAY button for 2.5 seconds until some
Granular processor
number of the LEDs become orange. Repeatedly press the • Long pressing the FREEZE button will toggle reversed
same button to step through the list of available modes. playback. When this setting is enabled, the FREEZE
button will flash, and each grain will play in reverse.
2. Press both the DISPLAY and the LOAD/SAVE buttons at
the same time. Press the two buttons to move forward or • The TEXTURE slider now as asymmetric shapes of grain
backwards through the list of available modes. envelopes. It morphs between square, ramp up, ramp
down, triangle, and triangle with diffuser. In its lowest
LEDs Mode name position, the square shape has particularly sharp edges,
Granular processor (normal operation) and may click. This is desired (more glitch!)
Pitch-shifter / time-stretcher • The range of the SIZE slider has been adjusted to
Looping delay produce much smaller grains. At its lowest position,
grains are barely audible spikes; at its highest, the grain
Spectral madness
size is the same as the stock firmware.
Oliverb
• The response curve of the DENSITY slider has been
Resonestor altered to enable easier access to slowly sown grains.
Important warnings • To cope with the additional computation requirements,
grain polyphony has been reduced slightly. This should
• Modules purchased on or after July 2015 must use
not be noticeable.
firmware version 1.3 or newer.
• Some cases of bricking have been reported with Looping delay and pitch-shifter
previous firmware versions, very roughly 1% of the time.
• When a clock is fed into the TRIGGER input, POSITION
Their causes are unknown, but all were fixed with an
becomes a clock divider/multiplier for the pre-delay.
FTDI or JTAG adapter and a computer.
With the slider at center, the pre-delay takes the value
• Genuine Mutable Instruments Clouds modules with of the clock length. Above center, the clock is divided.
serial numbers ending in a number less than 020 may Below center, the clock is multiplied. The multipliers for
have a faulty bootloader that could brick the module this slider are borrowed from the Make Noise
during an update. Contact Mutable Instruments.

32 33
 Echophon: 1⁄16, 3⁄32, 1⁄8, 3⁄16, 1⁄4, 3⁄8, 1⁄2, 3⁄4, 1⁄1, 3⁄2, 2⁄1, 3⁄1, 4⁄1, 6⁄1, Calibration procedure
8⁄1, and 12⁄1. Note that the clock division is limited to the
Calibration has been moved such that it is only possible to
largest division not exceeding the maximum delay;
access it on module startup. To calibrate, hold the
beyond this point, the POSITION slider has no effect.
LOAD/SAVE button while turning the module on. Then,
Also note that clock synchronization will be more
follow the Calibration procedure instructions in the
accurate with the SIZE slider at its lowest position.
Advanced topics section.
• In the mono quality modes (2 and 4), the STEREO knob
crossfades between the left and right inputs. If nothing Miscellaneous
patched into the left channel, this will have no effect.
• The length of the long-press for accessing the mode
• In the stereo quality modes (1 and 3), the STEREO knob switcher has been changed from 5 seconds to 2.5
gradually swaps both output channels. Fully clockwise, seconds.
it creates ping-pong delay effects; each time the sound
• The response of the FEEDBACK knob has been changed
is fed back, the two channels are reversed.
to be square-root-ish.

Looping delay • The constant low-frequency tone buildup has been

avoided at high FEEDBACK levels.
• Long-pressing the FREEZE button will toggle reversed
loop playback. • The TUNE knob has a clearer plateau at ±12 semitones,
so that octaves are easier to find on the knob.
• The TUNE knob bypasses the pitch shifter completely
when at 12 o’clock. This drastically enhances the delay • The WET/DRY response curve has been set to equal
quality when no pitch-shifting is done, avoiding the power crossfading (no volume drop at 50% anymore).
flanging effect when the delay was used with feedback. • The post-processing reverb has been placed after the
• When FREEZE is active and delay time is synchronized WET/DRY attenuation, so that it affects the dry signal
to an external clock, SIZE controls the repeat time (so that you can use the module as a simple reverb, plus
multiplier. its main function).

• The POSITION slider response has been tweaked to • The reverb is no longer increased when FREEZE is
allow easier dialing of short delays. Slider and CV enabled and FEEDBACK is up.
changes are much faster due to less slew.

34 35
 Oliverb
Oliverb is a full-featured, creative reverb mode based on the
DISP
reverb contained within the stock firmware, and inspired in
design by the MakeNoise Erbe-Verb module. It is largely
A B C D
modeless, probably unrealistic, and goes far beyond your
traditional reverb.

Controls
TUNE
A. The POSITION slider controls the pre-delay, ranging from
zero to about ½ of a second. When a clock is sent to the
TRIGGER input, POSITION becomes a clock divider/
multiplier for the pre-delay. Above center, the clock is
divided; below, it is multiplied. The multipliers for this slider INPUT LEVEL
are borrowed from the Make Noise Echophon: 1⁄16, 3⁄32, 1⁄8, 3⁄16, POSITION DENSITY SIZE TEXTURE

1⁄4, 3⁄8, 1⁄2, 3⁄4, 1⁄1, 3⁄2, 2⁄1, 3⁄1, 4⁄1, 6⁄1, 8⁄1, and 12⁄1. Note that the clock
division is limited to the largest division not exceeding the
maximum delay; beyond this point, the POSITION slider will
have no effect.
WET/DRY STEREO FEEDBACK REVERB
B. The DENSITY slider controls the decay length of the
reverb tail. Near the slider’s top, the signal is amplified and
the reverb enters self-oscillation.
C. The SIZE slider controls the reverb size - the size of the
emulated room.
V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE
D. The TEXTURE slider controls the dampening of the
reverb. From the bottom-most position to center, a low-pass
filter is applied, simulating the absorption of the room. From L IN R FREEZE L OUT R
center to the top-most position, a high-pass filter is applied
for unnatural, crystalline effects.

36 37
 E. The TUNE knob controls the pitch shift amount. Every
time the sound is fed back into the reverb, it can be pitch
shifted. The knob controls the amount of shift with a ±1
DISP
octave range. At 12 o’clock, no pitch shifting is applied. Fully
clockwise, you get the classic shimmer effect. Lots of
oddities can be found in between. To hear the effect of the
pitch shifter, some sound has to be fed back into the reverb E
by increasing the decay amount. Note that SIZE [C] has an
impact on how well the sound is pitch-shifted; the larger the
TUNE
room size, the more accurate the pitch shift.
F. The FREEZE button sets the reverb to (near) infinite
decay, and mutes the input. This works best with no pitch
shifting and a large size.
G. The WET/DRY knob continues to be a dry/wet mixer. INPUT LEVEL
POSITION DENSITY SIZE DAMPENING
H. The STEREO knob controls the diffusion amount - how
much the sound is “smoothed” by the diffusers each time it G H J I
goes through the reverb loop. Fully clockwise you get more
dense, continuous sound, fully counter-clockwise you get a
clearly repeated sound, like in a multi-tap delay. WET/DRY STEREO FEEDBACK REVERB

I. The REVERB knob controls the modulation amount

applied by 9 smooth random LFOs within the reverb. Small
modulations result in subtle chorus and ghost tones, and
large modulations result in random pitch shifts.
J. The FEEDBACK knob controls the modulation speed of V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE

the aforementioned random LFOs. It ranges from about

1⁄100Hz to about 100Hz. It has no effect if the modulation
amount is zero. L IN R FREEZE L OUT R

38 39
 Resonestor Controls
A. The POSITION slider controls the timbre and duration of
Resonestor is a dual-voice, four-part resonator (or comb
the noise burst. Below center will be longer and more
filter) effect with built-in capability for polyphonic Karplus-
dampened; above, it will be shorter and pitched higher. At
Strong plucked string synthesis, and more.
both ends the burst will be inaudible - convenient “mutes”.
To start the discovery process in this mode, it is
B. The DENSITY slider controls the decay length of the
recommended you set:
current voice. Above the ¾ mark, decay is infinite and the
• SIZE [C] and DENSITY [B] to their lowest positions sustained sound can be used as a traditional oscillator.
• POSITION [A], TEXTURE [D], TUNE [E], and INPUT C. The SIZE slider selects the chord for the current voice.
LEVEL to their center positions Borrowed from Mutable Instruments Braids, the chord
• STEREO [G], FEEDBACK [H], and REVERB [I] to fully morphs between unison, fat, superfat, fat power, fat octave,
counter-clockwise octaves, power, major, major 7th, minor 7th, minor, sus2,
• WET/DRY [F] to fully clockwise sus4, minor 9th, major 9th, minor 11th, major 11th.

Each audio input [3] feeds one of the two voices/resonators. D. The TEXTURE slider controls the dampening of the
The pitch of the resonator can be modulated, so the feedback loop of the resonator. At center, no filtering is
resonator can be “played” like an oscillator. Each voice applied. Below is a low-pass filter, above is a band-pass
consists of four “parts” (resonators), each of which can be filter with increasing resonance.
set to different pitches to form chords. E. The TUNE knob controls the resonator pitch of the
The controls facilitate sending short noise bursts to “excite” current voice. At 12 o’clock, the pitch is A3 (220Hz).
a resonator, creating the sound of plucked string
instruments like guitars.
A B C D
Because there are multiple voices, when triggered, the
E
module switches the active voice; therefore, you can get
duophony. At any time, the controls affect only the
parameters of the resonators of the active voice; the other
voices keep the last parameters they were on.
TUNE

POSITION DENSITY SIZE TEXTURE

40 41
 F. The WET/DRY knob controls distortion. It randomly Inputs and outputs
distorts the timbre of the voices by modulating each comb
1. The TRIGGER CV input switches the current voice (if
filter with low-pass-filtered noise. Fully counter-clockwise,
FREEZE [J] is not active) and sends a short burst of noise
the modulation amount is maximum but the noise is filtered
into its resonator.
out entirely, so there is no effect. Fully clockwise, the noise
is unfiltered, but the modulation amount is zero, so there is 2. FREEZE gate input. When the input gate signal is high,
no effect. In between, you get interesting effects, from FREEZE [J] is enabled. It will be disabled when the gate
subtle swaying to harsh distortion. signal is low.

G. The STEREO knob controls the stereo spread. Each voice 3. Stereo audio inputs. Rather than left and right channels,
and resonator is assigned to the left or right output [4]. Fully think of the two input channels as two distinct inputs, one
counter-clockwise, each voice goes to a different output. At for each voice. The inputs are still normalled together.
12 o’clock, both voices are equally mixed to each channel. 4. Stereo audio outputs. Rather than left and right
Fully clockwise, parts of both voices are distributed on both channels, think of the two output channels as two distinct
channels for a wide stereo effect. outputs, one for each voice. The inputs are still normalled
H. The FEEDBACK knob controls the harmonics. It together. The normalled output is both voices, mixed.
simulates striking the harmonics on a string. Fully counter-
F G H I
clockwise, it has no effect. Fully clockwise, the second
harmonic will ring; at 12 o’clock, the third, at 10, the fourth,
etc.
WET/DRY STEREO FEEDBACK REVERB
I. The REVERB knob controls the scatter. Random delay is
added to the input or burst sounds before they hit the
resonator of the current voice. Used for Karplus-Strong
synthesis with a chord, this will give the impression that
1
strings are struck sloppily. The delay times are randomized
at each voice switch. V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE

J. The FREEZE button triggers a switch to the next voice, 2

and prevents further voice switching from the TRIGGER CV
3 4
input [1]. L IN R FREEZE L OUT R
J

42 43
 Oliverb cheat sheet Resonestor cheat sheet

DISP DISP

REFLECT PITCH TUNE

INPUT LEVEL INPUT LEVEL

PRE DECAY SIZE DAMPENING TIMBRE & DECAY CHORD DAMPENING
DELAY LENGTH DURATION LENGTH

WET/DRY DIFFUSION MOD SPD MOD AMT DISTORTION STEREO HARMONICS SCATTER

V/OCT PRE-DELAY DECAY SIZE DAMPENING V/OCT TRIGGER TIM & DUR DECAY CHORD DAMPENING

L IN R INFINITE DECAY L OUT R R1 IN R2 FREEZE VOICE R1 OUT R2

44 45
 Overview
The Kammerl Beat Repeat firmware (KBR) is an alternate
firmware for Clouds. This guide will be focusing on

Kammerl Beat Monsoon, but the principles are functionally the same with
other Clouds derivatives.
NOTE: You must re-calibrate the module’s tuning after you

Repeat install KBR. See the Calibration procedure section.

The primary function of KBR is to add the Kammerl beat
repeat mode. This mode analyzes the incoming clock signal
and creates real-time audio input slices. It manages
Firmware multiple slices in real-time, which can be individually
version 0.42 for Monsoon
selected. Each slice can be played back with different loop,
pitch and distortion settings.
KBR also offers an enhanced version of the factory Spectral
madness mode, called Spectral clouds, which creates
“cloud-like frequency spectra” by utilizing a high-resolution
multi-band filter with randomly modulated frequency bands.

46 47
 Switching modes with KBR Reverb change
To access the mode selector, Press the DISPLAY button for 5 The post-processing reverb has been placed after the WET/
seconds until some number of the LEDs become orange. DRY mix, so that it affects the dry signal (so that you can use
Repeatedly press the same button to step through the list of the module as a simple reverb, plus its main function). This
available modes. also applies to the “warm distortion” setting in the Spectral
clouds mode (see later section).
LEDs Mode name
Granular processor (normal operation)
Pitch-shifter / time-stretcher
Looping delay
Spectral clouds
Kammerl beat repeat

Important warnings
• It is always possible to damage your module when
flashing alternate firmware. Please understand the
risks of doing so, before flashing.
• Genuine Mutable Instruments Clouds modules with
serial numbers ending in a number less than 020 may
have a faulty bootloader that could brick the module
during an update. Contact Mutable Instruments.

48 49
 Kammerl beat repeat mode E. The TUNE knob defines playback speed. Pitch is not
time-corrected. This allows for interesting groove effects,
Note: This mode requires a clock signal plugged into the since audio is delayed towards the end of the slice but reset
TRIGGER [3] input. Without a clock signal, it holds the most at the beginning of the slice. Fully counter-clockwise, the
recent audio slice (or silence, if no clock signal has been knob defines a zero pitch, and at fully clockwise, the original
previously provided). playback speed.

Controls
A. The POSITION slider controls the start position of the
loop, relative to the length of the slice. It is quantized to one
of the following values, from bottom to top: Unquantized, 1⁄64,
1⁄32, 1⁄16, 1⁄8, 1⁄4, 1⁄3, 1⁄2, 1⁄1.
DISP
B. The SIZE slider controls the length of the loop, relative to
the length of the slice, as well as the loop mode (regular or A C B D
alternating). It is quantized as follows:
• Regular from the bottom to center: Unquantized, 1⁄64, E
1⁄32, 1⁄16, 1⁄8, 1⁄4, 1⁄3, 1⁄2.
• Alternating from the center to the top: 1⁄2, 1⁄3, 1⁄4, 1⁄8, 1⁄16, TUNE
1⁄32, 1⁄64, unquantized.
C. The DENSITY slider controls the loop length modulation.
This enables a decreasing loop length towards the end of
the slice, creating a ping-pong effect.
INPUT LEVEL
D. The TEXTURE slider controls the slice step pattern - the
POSITION DENSITY SIZE TEXTURE
order in which slices are selected during playback. From
bottom to top, the possible patterns are: Disabled (only CV
selects slices), repeat current slice, skip every 2nd slice, skip
two slices, skip four slices, skip five slices, skip six slices,
randomly select slices.

50 51
 F. The FREEZE button enables slice processing/beat-
repeating. If not enabled, slice processing is randomly
enabled based on the slice probability blend control [G].
DISP
G. The WET/DRY knob controls slice probability - the
probability of disabling bypass and processing an incoming
slice. Note that activating the FREEZE [F] button overwrites
this setting.
H. The STEREO knob selects a clock divisor. Possible
values, in clockwise order, are: 1⁄1, 1⁄2, 1⁄4, and 1⁄8. Slice TUNE
lengths are changed accordingly.
I. The FEEDBACK knob selects one of four pitch
modulation modes, in clockwise order:
• Fixed pitch - no modulation. INPUT LEVEL
• Fixed pitch - reversed playback. POSITION DENSITY SIZE TEXTURE

• Linearly decreasing pitch, starting from the original G H I J

pitch, moving to the target pitch [E].
• Linearly increasing pitch, starting from the target
pitch [E], moving to the original pitch. WET/DRY STEREO FEEDBACK REVERB

• Simulated vinyl scratching - sinusoidal pitch

modulation. The TUNE knob [E] controls the intensity.
J. The REVERB knob controls the feedback amount.

V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE

L IN R FREEZE L OUT R
F

52 53
 Inputs and outputs
1. The TRIGGER input is the main slicing clock signal. A
KBR cheat sheet
clock in this input is required. Without it, the module holds
DISP
the most recent slice, which is silence when no clock signal
has been previously provided.
2. The TEXTURE CV input selects one of the eight most
recently recorded slices. 0V corresponds to the most recent
slice (real-time, no delay). Note that slices are continuously
created from audio input, independent from FREEZE [F] or
PLAYBACK SPEED
the slice probability mode [G]. This acts as an offset to the
slice selected with the TEXTURE slider [D].

INPUT LEVEL
START LENGTH LENGTH SLICE
MOD PATTERN

WET/DRY STEREO FEEDBACK REVERB

PROB CLK DIV PITCH MOD FEEDBACK

1 2

V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE

PLBK SPD TRIGGER START LEN. MOD LENGTH PATTERN

L IN R FREEZE L OUT R

L IN R BEAT REPEAT L OUT R

54 55
 Spectral clouds mode
This mode, a revised version of the stock firmware’s
DISP
Spectral madness mode, is all about creating “cloud-like
frequency spectra”. It utilizes a high-resolution, multi-band
A C B D
filter with randomly modulated frequency bands to do so.
The logarithmic divisions of the frequency spectrum result
in a continuously morphing, but musical sounding filter.

TUNE
Controls
A. The POSITION slider controls frequency band probability.
The defines the probability of a frequency band to become
enabled. At bottom, all frequency bands are disabled.
B. The SIZE slider controls the frequency band division. This INPUT LEVEL
POSITION DENSITY SIZE TEXTURE
is the number of filter bands and their corresponding
frequency width. At bottom, the spectrum is split into four E F
filter bands, and the top, into 128 bands. All frequency band
divisions are applied in a logarithmic scale, in order to
sound musical. WET/DRY STEREO FEEDBACK REVERB
C. The DENSITY slider controls filter smoothing. This is the
smoothing intensity on the frequency band division, as well
as the filter band attenuation changes during TRIGGER [1]
inputs. At bottom, filter changes are applied immediately. At 1
the top, the current filter configuration is held.
V/OCT TRIGGER POSITION DENSITY SIZE TEXTURE
D. The TEXTURE slider controls filter texture. This defines
the degree of phase randomization in the frequency domain.
This mostly affects waveforms which contain transients. L IN R FREEZE L OUT R

E. The STEREO knob controls the probability of random

filter modulation - essentially the chance of simulated

56 57
 TRIGGER [1] input events. Fully counter-clockwise, the
simulated events are disabled, and random filter re- Spectral clouds cheat sheet
configurations only happen when the signal to TRIGGER [1]
is high. DISP

F. The FEEDBACK knob controls the amount of warm

distortion. This effect is applied post dry/wet mixing.

Inputs and outputs

1. The TRIGGER input randomizes the set of active
frequency bands, and controls the degree to which they are TUNE

randomly attenuated. Note that these TRIGGER input events

can also be simulated with the second blend mode knob [E].

INPUT LEVEL
BAND FILTER BAND FILTER
PROB SMOOTHING DIVISION TEXTURE

WET/DRY RANDOM DISTORT REVERB

V/OCT TRIGGER BAND PRB SMOOTH BAND DIV TEXTURE

L IN R FREEZE L OUT R

58 59
 Appendix • Mutable Instruments Clouds’ on-line user manual and
the in-box printed quick-start manual
Credits, licensing, and honorable mentions • After Later Audio’s article titled “Understanding Clouds
A. Mutable Instruments Clouds’ source code is written by and Its Derivatives”
Émilie Gillet, and licensed under the MIT license. • The Clouds Parasite firmware guide page by Matthias
B. Mutable Instruments Clouds’ hardware is designed by Puech, the firmware’s author
Émilie Gillet, and licensed under the Creative Commons • The Kammerl Beat Repeat firmware guide page by
Attribution-ShareAlike 3.0 (CC-BY-SA-3.0) license. Julius Kammerl, the firmware’s author
C. Monsoon’s hardware is designed by jakplugg
(github.com/jakplugg) as a derivative work of Mutable Contact
Instruments Clouds, licensed under CC-BY-SA-3.0. This guide was compiled by Taylor Thurlow. You can find me
D. The Parasite firmware for Mutable Instruments Clouds online at the following locations:
was created by Matthias Puech (mqtthiqs.github.io). The • Personal website: thurlow.io
firmware source code (github.com/mqtthiqs/parasites) is • GitHub: github.com/taylorthurlow
licensed under the MIT license. More information about the
• Email: [email protected]
firmware can be found at mqtthiqs.github.io/parasites.
E. The Kammerl Beat Repeat firmware for Mutable Disclaimer
Instruments Clouds was created by Julius Kammerl
This guide and its contents contain references to Mutable
(kammerl.de). The firmware source code
Instruments. Mutable Instruments is a registered
(github.com/jkammerl/eurorack) is licensed under the MIT
trademark. Use of the name and the names of its products
license. More information about the firmware can be found
within this document does not imply any affiliation with, or
at kammerl.de/audio/clouds.
endorsement by, Mutable Instruments.
F. The Parasite and Kammerl Beat Repeat firmwares for
Monsoon were adapted by Tom Burns (burns.ca). The Selling or purchasing this guide
firmware’s source code (github.com/boourns/eurorack) is
This guide is not currently for sale. If you paid for this guide,
licensed under the MIT license.
or any portion of it, in any format (digital or physical), you
G. Much of the content within this guide is copied or adapted have been scammed.
directly from guides and resources available online:

60 61
 Please refer to the Mutable Instruments Clouds
online information page for detailed information
regarding compliance with EMC directives.

Guide version 1.2.1, June 2020