zlosynth instruments Kaseta User manual

MANUAL
Kaseta is a multi-purpose module inspired by reel-to-reel
tape machines. It simulates magnetic hysteresis to provide
warm saturation, its four independent delay lines can be
used to sculpt rhythms or feedback loops, and it offers
wow and flutter control. The module also goes beyond
typical tape machine features, with free-moving delays,
trigger sequencing, and an internal oscillator.
Width 20 HP
Depth 28 mm
Power +12 V (117 mA), −12 V (8 mA)
Input impedance 100 kΩ
CV inputs −5 to +5 V, 16-bit, 1 kHz
Trigger output 0 to +5 V, 10 ms
Audio 24-bit, 48 kHz
- Delay with 4 reading heads
- Up to 5 minutes of audio recording
- Tape saturation simulation
- Wow and flutter effects
- Tone control
- Voltage-controlled oscillator
- Trigger sequencer
- Stereo output

1 Installation
Kaseta is 20 HP wide. It is powered by a +12V/−12V 2×5 connector. The
red stripe (−12V) has to be connected on the side of the board marked with
the white line. The module must be mounted in a eurorack case.
2 Controls, inputs and outputs
There is one AC coupled audio input IN, and two stereo outputs LEFT and
RIGHT. They all operate in the range from −5 to +5 V.
There is a total of 23 knobs. With the four identical rows (H) controlling
four independent delay reading heads.
The button (B) serves for tempo tap-in and access to secondary attributes
of some knobs: Position of a filter set through TONE, placement of wow
and flutter effects set through WOW/FLUT, unlimited hysteresis enabled
through DRIVE, and internal oscillator enabled through PRE-AMP.
The four CONTROL inputs accept voltage from −5 to +5 V and can be
mapped to any of the knobs. Values set by the knob and the control input
are summed together.
For most attributes, summing the minimal value of the knob with +5 V
control input would produce the maximum value of the attribute. 0 V on the
control input would not affect the attribute, and −5 V with maximum value
on the knob would lead to the minimum value of the attribute.
For TONE and WOW/FLUT, the maximum control input would only offset
the value to its middle point.
For the internal oscillator, the control input follows the 1V/oct standard,
with the knob adding an offset of −2 to +1 octaves.
The display (X) visualizes dialed-in attributes, warnings and configuration.
The LED (Y) and IMPULSE output are triggered at intervals controlled by
the delay heads.
3 Mapping
Each of the four multi-purpose control inputs can be mapped to any of the
knobs:
1. Plug a cable into one of the control inputs.
2. Display will signalize mapping.
3. Turn the desired target knob.
The mapping is persisted between restarts. Disconnect a cable to unmap it.
4 Calibration
Some of the attributes follow the 1V/oct standard. Calibrate each of the
control inputs to increase the accuracy.
1. While holding the button, connect a jack to an input.
2. The first and second LED should light up.
3. Play note C on the CV source and press the button.
4. Now, the third and fourth LED should light up.
5. Play C one octave higher and press the button again.
The module then enters mapping mode of the given control input.
The calibration is persisted between restarts and disconnects.
5 Reset
Calibration, mapping and all secondary attributes are persisted between
restarts of the module. To reset their values, hold the button pressed while
powering on the module.

6 Signal flow
Pre-amp Drive+Bias Dry/Wet Tone Wow/Flutter
Input Saturation Filter Time effect
Write
Filter Tone Time effect Wow/Flutter
Read N Position N + Speed
Other heads Feedback N
Volume N + Pan N
Other heads
Outputs
Figure 1: The default signal flow within the module. Only a single reading head is shown for clarity.

7 Pre-amp
The PRE-AMP attribute attenuates or amplifies the signal received via INPUT
between silence and +28 dB. If the input signal is boosted too hard and
starts clipping, the display will start blinking.
8 Internal oscillator
The INPUT signal can be replaced with an internal oscillator. This oscillator
consists of two sine waves, one of which is a slightly detuned sub-octave.
To enable this feature, turn the PRE-AMP knob to the max while holding
the button. The PRE-AMP knob controls the pitch. If a control input is
mapped to the pitch, it will follow the 1V/oct standard.
9 Hysteresis
To replicate the warm tape saturation, the modules leverages Jiles-Atherton
magnetization model
1
. DRIVE and BIAS are affecting the character of the
saturation. DRY/WET blends between clear and saturated signals.
9.1 Limitting
The range of drive and bias is limitted to keep the simulation stable. This
limitation can be disabled to reach far harsher distortions, clicks and pops.
Disable it by turning the DRIVE knob to the max while holding the button.
1
Implementation of this algorithm is based on Jatin Chowdhury’s paper https://
ccrma.stanford.edu/~jatin/420/tape/TapeModel_DAFx.pdf
10 Tone
TONE applies a filter on the saturated signal. When the pot is at its 12
hours, the filter is disabled. Turning it to the left or right enables a low-pass
or high-pass filter, respectively.
10.1 Placement
There are three placement options for the filter.
On the input – the filter gets applied on the signal after it leaves the
saturation stage. This is useful to shave off some unwanted frequencies and
can act as a VCF when mapped to a CV input.
On feedback – the filter gets applied on the signal every time it passes
through the feedback. This is useful to keep the loop from getting out of
hand and also to introduce dampening.
Both simultaneously – both of the placements are active. Default.
Hold the button while turning the TONE knob to switch between the input,
feedback, and both placements.

11 Wow and flutter
The module simulates two phenomena known from the physical medium:
Wow, a slow fluctuation of the playback speed, causing the pitch to move
up or down. And flutter, abrupt changes of the playback speed, resulting in
faster momentary increases of pitch.
WOW/FLUT is controlling these time effects. When the pot is at its 12
hours, the effect is disabled. Turning it to the left or right enables wow or
flutter, respectively.
11.1 Placement
There are three placement options for this effect, each with its distinct
behavior.
On the input – the effect will be applied to the signal before it enters the
delay. This is useful when using multiple heads placed in different positions
since the modulation at the time of their reading will vary.
On reading from tape – the effect will be applied when reading the signal
back from the tape. This is useful when using feedback, since each time the
signal passes through the tape it will get further modulated.
Both simultaneously – both of the placements are active. Default.
Hold the button while turning the WOW/FLUT knob to switch between the
input, read, and both placements.
12 Delay
The input signal gets recorded on an imaginary tape by a writing head, to be
then, after a set interval, played back by a reading head. See the figure 1.
Hold the button for 5 seconds to clear the entire tape.
12.1 Heads
There are four such reading heads, each controlled by four knobs (H). POSI-
TION sets the relative position in the delay. VOL controls how much of the
signal will be sent into the output, with PAN controlling the balance between
LEFT and RIGHT. FEEDBACK controls how much of the signal is fed back
to the beginning of the delay.
When the combined strength of all feedback gets too strong, the module
may enter into a feedback loop. To get out of this loop, reduce the feedback,
or use the tone filter.
When the signal written to the tape is too loud, it will not allow any feedback
to fit in. To allow stronger feedback, reduce the pre-amp or drive.
12.2 Speed
SPEED controls the velocity in which the imaginary tape loops around, or
in other words, the length of the delay. By default, this length is between
5 minutes and 10 ms. Turning this knob to the middle while holding the
button switches to a shorter range between 8 seconds and 10 ms, turning it
to the maximum switches to audio range between 14 Hz and 1.8 kHz.
Alternatively, tap the button four times to set the desired tempo. Similarly,
if a clock signal is detected in control input mapped to SPEED, it would set
the tempo, with the SPEED knob acting as a multiplier.
12.3 Impulse
The IMPULSE trigger output will fire in the pattern set through head positions.
It is synchronized with the tap-in or clock-in.

13 Examples
Some basic combinations to get you started.
13.1 Clean slate
Pass through clean unaffected signal.
13.2 Saturation
Saturated signal without any delay or effects. Play with the PRE-AMP,
DRIVE, BIAS and DRY/WET controls to achieve the desired sound. Try
different input sources.
Stronger PRE-AMP usually produces more pronounced saturation, but be
careful not to let it clip.

13.3 Ping pong delay
The two top-most heads play the incoming signal with a delay set via
POSITION. They are placed left and right using PAN. Note that VOL of
both of these heads is reduced to avoid compression or clipping.
The last head is enabled with FEEDBACK, feeding the signal back to the
beginning to sustain the echo.
13.4 Haas effect and reverb
The first two heads, playing left and right with a short delay in between,
model a Haas effect. This effect makes the mono input sound wide in the
stereo output.
The two other heads are enabled with FEEDBACK. The feedback signal is
filtered with a low-pass filter configured using TONE. This produces a very
simple reverb. If the output sound grows into a loud feedback loop, set
TONE or FEEDBACK a little lower.

14 Acknowledgments
Kudos to all the eurorack, DSP, and embedded
programming communities online. Here are some
of the valuable resources that helped shape this
module:
Jatin Chowdhury’s white papers Real-time physical
modelling for analog tape machines
1
and Complex
nonlinearities for audio signal processing
2
, and his
open-source plugin ChowTape
3
. These materials
served as the base for Kaseta’s hysteresis model.
Nigel Redmon’s blog EarLevel Engineering
4
and
specifically his series about oversampling5.
mhampton’s implementation
6
of the Ornstein-
Uhlenbeck algorithm, which was used as part of
the wow effect.
And last but not least, Hainbach
7
and his con-
taigious enthusiasm about tape machines. Which
sparked my initial ideas to create this module.
1https://ccrma.stanford.edu/~jatin/420/tape/
TapeModel_DAFx.pdf
2https://ccrma.stanford.edu/~jatin/papers/
Complex_NLs.pdf
3https://github.com/jatinchowdhury18/
AnalogTapeModel/
4https://www.earlevel.com/
5https://www.earlevel.com/main/category/
digital-audio/sample-rate-conversion/
6https://github.com/mhampton/
ZetaCarinaeModules
7https://www.youtube.com/@Hainbach
15 Changelog
v1.0 The status LED blinks once
v1.2 The status LED blinks twice
Holding button clears the tape.
New wow and flutter placements.
New filter placements.
16 Questions?
You can find more information about the module
on https://zlosynth.com/kaseta.
p[email protected]