RECODER was a limited Black Friday /
End of Year 2025 release.
All units that were ordered during
the pre-order time window
have been manufactured
and are now shipping!
A small number of B-stock or
Surplus units may become available.
Leave your email below and we’ll
inform you about availability on
surplus RECODER units on
a first-come first-serve basis…
The RECODER was instantiated
as a finite operational event during
the Q4 2025 Black Friday window.
The fabrication cycle has concluded.
All verified units have entered
the distribution network and
are currently en route to
their assigned human operators.
A small number of statistically
anomalous B-stock variants
may still exist within the system.
Establish a telemetry link below
to receive notification should any
additional units be detected, recovered,
or spontaneously instantiate.
The idea for this project came
out of an unusual creative challenge:
how can we build an effects pedal
where no two units will ever
sound the same?
A continuously morphing effects
engine that reshapes itself around
each player’s musical palette –
producing radically different results
based on the sounds you feed into it,
rather than imposing a fixed,
pre-designed algorithm.
A deceptively simple combination
of two ideas – Recording & Vocoding –
resulting in a fully interactive, input-dependent effects pedal with
a fluid sonic identity.
A Spectral Morphing Workstation
that lets you record any musical phrase
or audio signal into its internal memory, and turn it into an entirely new,
original sound-processing engine.
When you record a phrase, RECODER
does not play it back like a looper.
Instead – it performs a detailed spectral
analysis of the captured signal –
extracting its sonic DNA with extraordinary
precision, and transforms it
into a dynamic audio processing
template which is applied directly
onto your live playing in real-time.
The result is a continuous conversation
between two signals – where every
new note you play pushes the character
of the interaction, and pulls the effect
in new and unexpected directions.
You can think of the RECODER as:
a signal cross-pollinator,
a dynamic filter,
a moving audio prism,
a ghostly vocoder,
a formant resonator,
a granular echo creator, or even
a sound-imprint machine…
…but the truth is that even we don’t know how your RECODER is going to sound, because this pedal is only complete when you fill it up with your own musical ideas…
The foundational architecture of this project emerged from a highly anomalous creative constraint: the engineering of an acoustic transformation node wherein no two discrete physical units could ever achieve systemic output parity.
The objective was the realization of a continuously mutating effects engine capable of real-time architectural remodeling based entirely upon the player’s unique sonic topography. Rather than imposing a rigid, pre-determined DSP algorithm upon the signal path, the system dynamically derives its operational behavior from the specific spectral inputs introduced into the system.
By cross-referencing and integrating two distinct audio paradigms—Continuous Buffer Acquisition (Recording) and Formant-Matrix Synthesis (Vocoding)—the architecture yields a fully interactive, input-dependent processing matrix characterized by a fluid, non-static sonic identity.
Classified formally as a Spectral Morphing Workstation, the apparatus allows for the volatile capture of any arbitrary musical phrase or acoustic signal into its internal memory, instantly converting that dataset into a wholly unique, user-generated
signal-processing engine.
Upon the execution of a phrase capture cycle, the RECODER explicitly bypasses standard linear playback (looper) topologies. Instead, it executes a high-resolution Fast Fourier Transform (FFT) and temporal analysis of the captured donor signal—extracting its fundamental acoustic footprint with mathematical precision. This data is instantly compiled into a dynamic audio-processing template, which is then mapped directly onto the live incoming signal in real time.
Depending on the operational constraints and the structural complexity of the donor data, the RECODER may be classified under several technical designations:
a signal cross-pollinator array,
a dynamic spectral filter topology,
a refractive audio prism,
a stochastic formant resonator,
a formant resonator,
a granular echo/delay synthesizer,
an acoustic imprint/convolution engine…
…ultimately, a definitive empirical analysis of the hardware’s acoustic output remains impossible in a laboratory setting; the system architecture remains fundamentally incomplete until the user populates the volatile memory registers with their own unique musical data vectors….
The RECODER operates through the interplay between two signals: the Live Input from your instrument (INPUT), and a secondary MODULATOR signal – which can be any recorded PHRASE or a live
MODULATOR input (REC INPUT).
Internally, RECODER runs a real-time high-resolution Fast Fourier Transform (FFT) spectral analysis engine – measuring the energy distribution of both signals across
1024 isolated frequency bands.
The RESOLUTION control adjusts how often these “spectral snapshots”are refreshed – from rapid 5ms intervals to slower,
stepped windows (40ms, 80ms).
Once the spectral profiles of both signals are measured, the RECODER can apply precise, band-specific adjustments to the INPUT — performing surgical moves on every frequency band, and structuring the interaction between the two signals on a molecular level:
Transferring TIMBRE between signals, shaping the AMPLITUDE of one with the other, making your INPUT mimic the MODULATOR’s sound and respond to its every movement — and extracting SLICES, GRAINS, and RESONANCES from the overlapping spectral content, then pushing them through time as melodic feedbacks, reverbs, delays, and
harmonic blooms.
The operational matrix of the RECODER is sustained through the continuous, real-time interplay between two discrete signal vectors: the primary Live Input Path derived from the source instrument, and a secondary Modulator Vector. This modulating component is sourced either from a previously captured, volatile memory Phrase Block or from an active, real-time External Modulator Input via the dedicated REC Input bus.
To facilitate accurate cross-synthesis, the internal DSP architecture instantiates a high-resolution, real-time Fast Fourier Transform (FFT) spectral analysis engine. This subsystem continuously quantifies and maps the specific energy distribution and phase alignment of both concurrent signals across exactly 1024 isolated frequency bins (N=1024).
The frequency with which these discrete “spectral snapshots” are integrated and refreshed is governed entirely by the RESOLUTION telemetry control. Adjusting this parameter modifies the temporal window size (Δt) of the analysis stage:
High-Velocity Bounds (∼5 ms): Yields rapid, high-fidelity temporal tracking suitable for transient-heavy, volatile signal responses.
Low-Velocity/Stepped Bounds (40 ms≤Δt≤80 ms): Shifts the engine into macro-stepped integration windows, prioritizing high frequency-resolution over rapid temporal adaptation.
All RECODER sound engines are controlled
via 4 shared endless encoders
across three color-coded control LAYERS:
WHITE, YELLOW, and RED. Each control
LAYER gives the encoders a different
set of functions, letting you access
3 distinct processing domains:
INPUT MODIFICATION & FEEL
Use the spectral & dynamic profile of the MODULATOR to modify INPUT signal: Short 20-100ms PHRASES act like spectral “snapshots” that you can apply onto your input as timbral filters.
Longer PHRASES (or any LIVE signal routed into the REC Input) become a constantly moving dynamic modifier for your main instrument!
SPECTRALLY INDUCED INPUT MODIFICATION & DYNAMIC ADAPTATION
This operational mode executes the direct mapping and execution of the Modulator Vector’s complex spectral and dynamic contours onto the primary Live Input Path.
Through this cross-synthesis protocol, the user calibrates the specific systemic sensitivity profiles, regulating precisely how the internal DSP architecture responds to the incoming signal amplitude, envelope transients, and real-time playing dynamics of the source instrument.
SHAPE INPUT VOLUME FROM MODULATOR: RECODER tracks the MODULATOR’s main amplitude peaks in real-time and applies them to the INPUT, creating frequency-specific boost and cut zones.
You can use a recorded PHRASE or a secondary live REC INPUT as the MODULATOR!
Side-Chained Boost:
Boosts frequencies where MODULATOR signal is strongest. At maximum AMPLITUDE values, the dominant bands are boosted past breakup – creating a moving overdrive zone.
Side-Chained Cut:
Dips the frequency bands where the MODULATOR’s signal is strongest – producing a moving band cut
or a variety of custom-shaped tremolo effects.
FFT VOCODER EFFECT:
FFT analysis tracks the energy distribution of both INPUT & MODULATOR in real-time: The INPUT is then rebalanced across the full frequency spectrum to mirror and mimic the MODULATOR’s tonal profile. Inactive or silent MODULATOR bands do not produce dips in the INPUT signal.
This is called TIMBRE TRANSFER and you can think of it as a vocoder, operating at a significantly higher 1024-band FFT resolution.
Just like with a vocoder, TIMBRE Transfer allows you to alter the INPUT tone, while keeping your playing in tune!
Inactive or silent MODULATOR bands
do not produce dips in the INPUT signal.
The RECODER can also isolate and extract
the overlapping spectral content between
both signals – THE
MODULATOR & main
INPUT – generating entirely new sonic
structures – SLICES & RESONANCES.
Then – you can push them through time,
creating unique variations of melodic
feedback, echoes, delays and reverbs.
GENERATIVE & SPATIAL PHENOMENA
The RECODER continuously monitors both
active signal streams for regions
of spectral coincidence.
Whenever sufficient overlap is detected, the system may spontaneously generate SLICES and RESONANCES — autonomous acoustic structures derived from shared harmonic material present in both datasets.
These structures can subsequently be propagated through time, resulting in melodic feedback loops, recursive echo fields, temporal diffusion networks, artificial reverberation chambers, and occasional violations of local acoustic causality.
Extracts small audio fragments from the MODULATOR – triggered by matching or overlapping frequencies in the INPUT.
The amount & behaviour of the SLICES will always depend on each unique signal combination; SLICES always stay in tune with the INPUT signal!
Generates artificial sine-wave components from the strongest overlapping frequency peaks between INPUT & MODULATOR,
and sends them into a feedback network – creating responsive, melodic resonances that evolve with your playing.
Turn the knob to the left to create NEGATIVE RESONANCES: this produces a beautiful inverted harmony!
A flexible TIME & REGENERATION engine that processes both YELLOW Layer signals SLICES & RESONANCES simultaneously.
REPEATS: controls the direction and amount of regeneration;
DISTANCE: controls the spacing or diffusion density between REPEATS.
Together these controls allow you to create four distinct regeneration effects: Delay, Reverse Delay, Reverb & Feedback.
RESHAPING THE MODULATOR
The Red Layer lets you radically reshape
the personality and texture of
the MODULATOR, thus affecting all
further interactions downstream.
MODULATOR RECONFIGURATION
The Red Layer provides direct access to the structural properties of the MODULATOR dataset, allowing extensive modification of its spectral profile, dynamic behavior, and overall acoustic identity.
Since all subsequent processing stages derive information from the MODULATOR, alterations introduced here propagate throughout the entire signal-processing chain.
In LOOP or STEP ADVANCE mode – press both footswitches simultaneously to audition the MODULATOR with all RED LAYER adjustments via MAIN OUTPUT.
The RECODER offers three distinctive behavior modes – each offering a new way to interact with the MODULATOR:
LOOP Mode — Circular Playb
ck The default mode, and the one that inspired the RECODER concept in the first place.
The MODULATOR spins continuously inside the engine — a closed loop, playhead always advancing — driving every transformation in real time.
STEP Mode – Dynamic Playhead
RECODER analyzes the recorded phrase and splits it into segments based on transients and pitch changes — like tiny “phrase markers” or event points.
The playhead only advances when you trigger it with your playing.
Each time you strike a note or hit a transient — the engine jumps to the next slice, segment, or pitch zone in the recorded phrase.
This is the ideal mode for breaking out of a set rhythmical pattern!
* Advance Sensitivity & Threshold can be adjusted in Global Settings.
LIVE Mode – Real-time SPECTRAL MORPHING
Instead of relying on saved phrases as the signal source — the RECODER uses your live input as the spectral donor – essentially becoming a real-time spectral resynthesis and cross-morphing machine.
LOOP Mode — Continuous Buffer Cycling
Default operational mode.
The recorded phrase is stored as a circular buffer. A fixed-position playhead continuously cycles through the entire phrase in a closed loop, maintaining uninterrupted time progression.
All spectral, temporal, and harmonic transfer computations are driven by the current playhead position within the phrase.
STEP Mode — Event-Driven Playhead Advancement
The recorded phrase is segmented into discrete units based on detected transient peaks, spectral discontinuities, and pitch-zone boundaries. These become internal event markers. The playhead does not advance continuously — it advances only when triggered by the input signal.
When a new transient or note onset is detected at the input, the playhead jumps to the next indexed event marker (segment, slice, or pitch zone).
This yields phrase navigation that is directly mapped to user excitation.
Sensitivity, detection threshold, and segmentation density are adjustable in Global Settings.
LIVE Mode — Real-Time Side-Coding and Cross-Morph Resynthesis
In this mode, the spectral donor is no longer strictly the stored phrase. Instead, the engine dynamically acquires spectral frames from the live input and uses them for real-time convolution, harmonic transfer, and parameter resynthesis.
A stored phrase may still be used as a reference, but once the REC footswitch is unlatched, the system begins replacing donor frames continuously with incoming spectral data.
This enables live, continuously updated cross-morph processing — effectively converting the system into a real-time spectral resynthesis engine.
MATH SERIES is our new playground for digital invention — a space where sound isn’t just processed, but interpreted, reshaped, and reborn through code. We call it Creative Signal Processing.
We’re still the PLASMA people. We’re still the MOTOR people. We’re still the team that pulled reverb out of infrared reflections on a metal spring.
That mission continues — always. But we’re also venturing into a parallel universe, where we turn audio signal into Xeros & 0nes, and explore the infinite spaces that lie in between.
The RECODER is the very first creature to walk out of that laboratory…
The MATH SERIES represents our newly instantiated vector for digital architecture exploration—a dedicated research environment wherein acoustic data is not merely conditioned, but computationally interpreted, structurally remodeled, and computationally reborn via custom firmware topologies. This methodology is formally classified as Creative Signal Processing.
Our foundational engineering paradigms remain uninterrupted; we continue to operate as the pioneer developers of PLASMA high-voltage discharge synthesis and MOTOR electromechanical actuation, as well as the team that successfully extracted mechanical reverberation from infrared optical tracking of physical spring arrays.
While that physical research trajectory remains continuous, our R&D division has simultaneously penetrated a parallel computational domain. Within this ecosystem, continuous audio signals are digitized into discrete binary states (0s and 1s) to explore the infinite mathematical spaces contained within the discrete-time domain.
The RECODER signifies the initial physical hardware manifestation to emerge from this computational laboratory.
| Instrument Input | 1/4" mono jack |
| Main Output | 1/4" mono jack |
| REC Input | 1/4" mono jack, buffered |
| REC Output | 1/4" mono jack, buffered |
| Signal Path | Mono - all inputs, all outputs |
| Built-in Microphone | Electret condenser microphone |
| Power Requirements | 9V DC, center-negative, 500mA |
| MIDI In | 3.5mm TRS Type B |
| USB-C | Firmware updates & MIDI over USB |
| Spectral Processing Engine | High-resolution FFT-based vocoder |
| RESOLUTION Positions | High (1024) / Mid (512) / Low (256) |
| Max INPUT Latency | 40 ms FFT window |
| Total Internal Storage | 16 GB |
| Preset Slots | 234 (A1–Z9) |
| Maximum PHRASE Length | 10 minutes per slot |
| Recording Fidelity | 44100Hz, 16-bit |
| PHRASE Storage Quality | Full fidelity - no compression |
| Product Dimensions | LWH 140 x 125 x 50 mm (5.51 x 4.92 x 1.97 inch) |
| Product Weight | 600 g (1.3 lbs) |
The RECODER features a mono signal path for all inputs and all outputs.
Not quite, it features 234 total blank slots ranging from A1 to Z9 for you to fill in with your recordings.
It has four 1/4" mono jack connections: an Instrument Input, a Main Output, a buffered REC Input, and buffered REC Output. It also includes a built-in electret condenser microphone.
Yes. It offers full MIDI control over parameters using either MIDI over USB (Type-C) or via the dedicated 3.5mm TRS Type B MIDI Input. Please note it does not transmit a MIDI Output.
It features a spectral processing engine running a high-resolution, FFT-based vocoder with a maximum of 1024 frequency bins and a 40 ms FFT window for max input latency.
Yes, it features three resolution positions: High (1024), Mid (512), and Low (256), which can also be toggled via MIDI overrides.
It includes 16 GB of total internal storage. It records at full fidelity (44100Hz, 16-bit) with no compression, allowing for a maximum phrase length of 10 minutes per slot.
File access via USB is currently not supported. The USB-C port is strictly used for firmware updates and MIDI over USB.
The RECODER requires a 9V DC, center-negative power supply capable of delivering at least 500mA
The product dimensions are 125 x 140 x 50 mm (4.92×5.51×1.97 inches) and it weighs approximately 600 g (1.3 lbs).
RECODER was a limited Black Friday / End of Year 2025 release.
All units that were ordered during the pre-order time window
have been manufactured and are now shipping!
A small number of B-stock or Surplus units may become available.
Leave your email below and we’ll inform you about availability on
surplus RECODER units on a first-come first-serve basis…
The RECODER was instantiated as a finite operational event during the Q4 2025 Black Friday window. The fabrication cycle has concluded.
All verified units have entered the distribution network and are currently en route to their assigned human operators.
A small number of statistically anomalous B-stock variants may still exist within the system.
Establish a telemetry link below to receive notification should any additional units be detected, recovered, or spontaneously instantiate.
The idea for this project came out of an unusual creative challenge:
how can we build an effects pedal where no two units
will ever sound the same?
A continuously morphing effects engine that reshapes itself around
each player’s musical palette – producing radically different results
based on the sounds you feed into it, rather than imposing a fixed,
pre-designed algorithm.
A deceptively simple combination of two ideas – Recording & Vocoding –
resulting in a fully interactive, input-dependent effects pedal with
a fluid sonic identity.
A Spectral Morphing Workstation that lets you record any musical phrase
or audio signal into its internal memory, and turn it into an entirely new,
original sound-processing engine.
When you record a phrase, RECODER does not play it back like a looper.
Instead – it performs a detailed spectral analysis of the captured signal –
extracting its sonic DNA with extraordinary precision, and transforms it
into a dynamic audio processing template which is applied directly
onto your live playing in real-time.
The result is a continuous conversation between two signals –
where every new note you play pushes the character of the interaction,
and pulls the effect in new and unexpected directions.
You can think of the RECODER as:
a signal cross-pollinator,
a dynamic filter,
a moving audio prism,
a ghostly vocoder,
a formant resonator,
a granular echo creator, or even
a sound-imprint machine…
…but the truth is that even we don’t know
how your RECODER is going to sound, because
this pedal is only complete
when you fill it up with your own musical ideas…
The foundational architecture of this project emerged from a highly anomalous creative constraint: the engineering of an acoustic transformation node wherein no two discrete physical units could ever achieve systemic output parity.
The objective was the realization of a continuously mutating effects engine capable of real-time architectural remodeling based entirely upon the player’s unique sonic topography. Rather than imposing a rigid, pre-determined DSP algorithm upon the signal path, the system dynamically derives its operational behavior from the specific spectral inputs introduced into the system.
By cross-referencing and integrating two distinct audio paradigms—Continuous Buffer Acquisition (Recording) and Formant-Matrix Synthesis (Vocoding)—the architecture yields a fully interactive, input-dependent processing matrix characterized by a fluid, non-static sonic identity.
Classified formally as a Spectral Morphing Workstation, the apparatus allows for the volatile capture of any arbitrary musical phrase or acoustic signal into its internal memory, instantly converting that dataset into a wholly unique, user-generated signal-processing engine.
Upon the execution of a phrase capture cycle, the RECODER explicitly bypasses standard linear playback (looper) topologies. Instead, it executes a high-resolution Fast Fourier Transform (FFT) and temporal analysis of the captured donor signal—extracting its fundamental acoustic footprint with mathematical precision. This data is instantly compiled into a dynamic audio-processing template, which is then mapped directly onto the live incoming signal in real time.
Depending on the operational constraints and the structural complexity of the donor data, the RECODER may be classified under several technical designations:
a signal cross-pollinator array,
a dynamic spectral filter topology,
a refractive audio prism,
a stochastic formant resonator,
a formant resonator,
a granular echo/delay synthesizer,
an acoustic imprint/convolution engine…
…ultimately, a definitive empirical analysis of the hardware’s acoustic output remains impossible in a laboratory setting; the system architecture remains fundamentally incomplete until the user populates the volatile memory registers with their own unique musical data vectors….
The RECODER operates through the interplay between two signals: the Live Input from your instrument (INPUT), and a secondary MODULATOR signal – which can be any recorded PHRASE or a live MODULATOR input (REC INPUT).
Internally, RECODER runs a real-time high-resolution Fast Fourier Transform (FFT) spectral analysis engine – measuring the energy distribution of both signals across 1024 isolated frequency bands.
The RESOLUTION control adjusts how often these “spectral snapshots” are refreshed –
from rapid 5ms intervals to slower, stepped windows (40ms, 80ms).
Once the spectral profiles of both signals are measured, the RECODER can apply precise, band-specific adjustments to the INPUT — performing surgical moves on every frequency band, and structuring the interaction between the two signals on a molecular level:
Transferring TIMBRE between signals, shaping the AMPLITUDE of one with the other, making your INPUT mimic the MODULATOR’s sound and respond to its every movement — and extracting SLICES, GRAINS, and RESONANCES from the overlapping spectral content, then pushing them through time as melodic feedbacks, reverbs, delays, and harmonic blooms.
The operational matrix of the RECODER is sustained through the continuous, real-time interplay between two discrete signal vectors: the primary Live Input Path derived from the source instrument, and a secondary Modulator Vector. This modulating component is sourced either from a previously captured, volatile memory Phrase Block or from an active, real-time External Modulator Input via the dedicated REC Input bus.
To facilitate accurate cross-synthesis, the internal DSP architecture instantiates a high-resolution, real-time Fast Fourier Transform (FFT) spectral analysis engine. This subsystem continuously quantifies and maps the specific energy distribution and phase alignment of both concurrent signals across exactly 1024 isolated frequency bins (N=1024).
The frequency with which these discrete “spectral snapshots” are integrated and refreshed is governed entirely by the RESOLUTION telemetry control. Adjusting this parameter modifies the temporal window size (Δt) of the analysis stage:
High-Velocity Bounds (∼5 ms): Yields rapid, high-fidelity temporal tracking suitable for transient-heavy, volatile signal responses.
Low-Velocity/Stepped Bounds (40 ms≤Δt≤80 ms): Shifts the engine into macro-stepped integration windows, prioritizing high frequency-resolution over rapid temporal adaptation.
All RECODER sound engines are controlled via 4 shared endless encoders across three color-coded control LAYERS: WHITE, YELLOW, and RED.
Each control LAYER gives the encoders a different set of functions, letting you access 3 distinct processing domains:
INPUT MODIFICATION & FEEL
Use the spectral & dynamic profile of the MODULATOR to modify INPUT signal:
Short 20-100ms PHRASES act like spectral “snapshots” –
that you can apply onto your input as timbral filters.
Longer PHRASES (or any LIVE signal routed into the REC Input)
become a constantly moving dynamic modifier for your main instrument!
SPECTRALLY INDUCED INPUT MODIFICATION & DYNAMIC ADAPTATION
This operational mode executes the direct mapping and execution of the Modulator Vector’s complex spectral and dynamic contours onto the primary Live Input Path.
Through this cross-synthesis protocol, the user calibrates the specific systemic sensitivity profiles, regulating precisely how the internal DSP architecture responds to the incoming signal amplitude, envelope transients, and real-time playing dynamics of the source instrument.
SHAPE INPUT VOLUME FROM MODULATOR:
RECODER tracks the MODULATOR’s main amplitude peaks
in real-time and applies them to the INPUT,
creating frequency-specific boost and cut zones.
You can use a recorded PHRASE or
a secondary live REC INPUT as the MODULATOR!
Side-Chained Boost:
Boosts frequencies where MODULATOR signal is strongest.
At maximum AMPLITUDE values, the dominant bands are
boosted past breakup – creating a moving overdrive zone.
Side-Chained Cut:
Dips the frequency bands where the MODULATOR’s signal is
strongest – producing a moving band cut
or a variety of
custom-shaped tremolo effects.
FFT VOCODER EFFECT:
FFT analysis tracks the energy distribution of both
INPUT & MODULATOR in real-time: The INPUT is then
rebalanced across the full frequency spectrum
to mirror and mimic the MODULATOR’s tonal profile.
Inactive or silent MODULATOR bands do not
produce dips in the INPUT signal.
This is called TIMBRE TRANSFER and you can
think of it as a vocoder, operating at a significantly
higher 1024-band FFT resolution.
Just like with a vocoder, TIMBRE Transfer allows you to
alter the INPUT tone, while keeping your playing in tune!
Inactive or silent MODULATOR bands
do not
produce dips in the INPUT signal.
GENERATIVE & SPATIAL FX
The RECODER can also isolate and extract the overlapping
spectral content between both signals – THE
MODULATOR & main
INPUT – generating entirely new sonic structures – SLICES & RESONANCES.
Then – you can push them through time – creating unique variations
of melodic feedback, echoes, delays and reverbs.
GENERATIVE & SPATIAL PHENOMENA
The RECODER continuously monitors both active signal streams for regions of spectral coincidence.
Whenever sufficient overlap is detected, the system may spontaneously generate SLICES and RESONANCES — autonomous acoustic structures derived from shared harmonic material present in both datasets.
These structures can subsequently be propagated through time, resulting in melodic feedback loops, recursive echo fields, temporal diffusion networks, artificial reverberation chambers, and occasional violations of local acoustic causality.
Extracts small audio fragments from the MODULATOR –
triggered by matching or overlapping frequencies in the INPUT.
The amount & behaviour of the SLICES will always
depend on each unique signal combination;
SLICES always stay in tune with the INPUT signal!
Generates artificial sine-wave components from the strongest
overlapping frequency peaks between INPUT & MODULATOR,
and sends them into a feedback network – creating responsive,
melodic resonances that evolve with your playing.
this produces a beautiful inverted harmony!
A flexible TIME & REGENERATION engine
that processes both YELLOW Layer signals
SLICES & RESONANCES simultaneously.
REPEATS: controls the direction and amount of regeneration;
DISTANCE: controls the spacing or diffusion density between REPEATS.
Together these controls allow you to create
four distinct regeneration effects:
Delay, Reverse Delay, Reverb & Feedback.
RESHAPING THE MODULATOR
The Red Layer lets you radically reshape the personality
and texture of the MODULATOR – thus affecting all further
interactions downstream.
MODULATOR RECONFIGURATION
The Red Layer provides direct access to the structural properties of the MODULATOR dataset, allowing extensive modification of its spectral profile, dynamic behavior, and overall acoustic identity.
Since all subsequent processing stages derive information from the MODULATOR, alterations introduced here propagate throughout the entire signal-processing chain.
In LOOP or STEP ADVANCE mode – press both
footswitches simultaneously to audition the MODULATOR
with all RED LAYER adjustments via the MAIN OUTPUT.
At its core, the RECODER is built for single-channel phrase acquisition and deterministic processing.
Connect any excitation source to the Main Input — guitar, synth, voice, drum machine, acoustic pickup, modular — and initiate phrase capture:
1. Hold the REC footswitch to begin buffer acquisition (maximum phrase length = full length WAV sample).
2. Capture progress is visualized via the circular LED matrix (1 LED = 1 s).
3. Upon release of REC, the recorded phrase is written to volatile RAM and immediately enters loop-state.
4. Live input is streamed through that phrase as a reference dataset — where it is interpreted, filtered, time-sliced, convolved, resynthesized, and spectrally remapped using deterministic phrase-based convolution.
Each captured phrase can be permanently stored in non-volatile memory — assigned to one of 234 indexed slots — along with all mutation parameters, spectral transformations, layer states, and modulation configurations.
Micro-phrases (50–200 ms) function as spectral filters, resonant kernels, and timbral imprint matrices.
Macro-phrases (up to full length WAV samples) behave as rhythmic convolution maps, transient arrays, formant distribution engines, spectral architectures, and deterministic phrase-based excitation systems.

This is the default mode – the one we’ve focused on in all previous sections,
and the mode that inspired the RECODER idea initially!
In LOOP mode, the recorded phrase is constantly spinning inside the engine — like a circular buffer. The playhead continuously cycles through the phrase, over and over again.
STEP Mode – Dynamic Playhead
RECODER analyzes the recorded phrase and splits it into segments based on transients and pitch changes — like tiny “phrase markers” or event points.
The playhead only advances when you trigger it with your playing.
Each time you strike a note or hit a transient — the engine jumps to the next slice, segment, or pitch zone in the recorded phrase.
This is the ideal mode for breaking out of a set rhythmical pattern!
* Advance Sensitivity & Threshold can be adjusted in Global Settings.
LIVE Mode – Real-time SPECTRAL MORPHING
Instead of relying on saved phrases as the signal source — the RECODER uses your live input as the spectral donor – essentially becoming a real-time spectral resynthesis and cross-morphing machine.
LOOP Mode — Continuous Buffer Cycling
Default operational mode.
The recorded phrase is stored as a circular buffer. A fixed-position playhead continuously cycles through the entire phrase in a closed loop, maintaining uninterrupted time progression.
All spectral, temporal, and harmonic transfer computations are driven by the current playhead position within the phrase.
STEP Mode — Event-Driven Playhead Advancement
The recorded phrase is segmented into discrete units based on detected transient peaks, spectral discontinuities, and pitch-zone boundaries. These become internal event markers.
The playhead does not advance continuously — it advances only when triggered by the input signal.
When a new transient or note onset is detected at the input, the playhead jumps to the next indexed event marker (segment, slice, or pitch zone).
This yields phrase navigation that is directly mapped to user excitation.
Sensitivity, detection threshold, and segmentation density are adjustable in Global Settings.
LIVE Mode — Real-Time Side-Coding and Cross-Morph Resynthesis
In this mode, the spectral donor is no longer strictly the stored phrase. Instead, the engine dynamically acquires spectral frames from the live input and uses them for real-time convolution, harmonic transfer, and parameter resynthesis.
A stored phrase may still be used as a reference, but once the REC footswitch is unlatched, the system begins replacing donor frames continuously with incoming spectral data.
This enables live, continuously updated cross-morph processing — effectively converting the system into a real-time spectral resynthesis engine.
MATH SERIES is our new playground for digital invention — a space where sound isn’t just processed, but interpreted, reshaped, and reborn through code. We call it Creative Signal Processing.
We’re still the PLASMA people. We’re still the MOTOR people. We’re still the team that pulled reverb out of infrared reflections on a metal spring.
That mission continues — always. But we’re also venturing into a parallel universe where we turn audio signal into Xeros & 0nes, and explore the infinite spaces that lie in between.
The RECODER is the very first creature to walk out of that laboratory…
The MATH SERIES represents our newly instantiated vector for digital architecture exploration—a dedicated research environment wherein acoustic data is not merely conditioned, but computationally interpreted, structurally remodeled, and computationally reborn via custom firmware topologies. This methodology is formally classified as Creative Signal Processing.
Our foundational engineering paradigms remain uninterrupted; we continue to operate as the pioneer developers of PLASMA high-voltage discharge synthesis and MOTOR electromechanical actuation, as well as the team that successfully extracted mechanical reverberation from infrared optical tracking of physical spring arrays.
While that physical research trajectory remains continuous, our R&D division has simultaneously penetrated a parallel computational domain. Within this ecosystem, continuous audio signals are digitized into discrete binary states (0s and 1s) to explore the infinite mathematical spaces contained within the discrete-time domain.
The RECODER signifies the initial physical hardware manifestation to emerge from this computational laboratory.
The RECODER features a mono signal path for all inputs and all outputs.
Not quite, it features 234 total blank slots ranging from A1 to Z9 for you to fill in with your recordings.
It has four 1/4" mono jack connections: an Instrument Input, a Main Output, a buffered REC Input, and buffered REC Output. It also includes a built-in electret condenser microphone.
Yes. It offers full MIDI control over parameters using either MIDI over USB (Type-C) or via the dedicated 3.5mm TRS Type B MIDI Input. Please note it does not transmit a MIDI Output.
It features a spectral processing engine running a high-resolution, FFT-based vocoder with a maximum of 1024 frequency bins and a 40 ms FFT window for max input latency.
Yes, it features three resolution positions: High (1024), Mid (512), and Low (256), which can also be toggled via MIDI overrides.
It includes 16 GB of total internal storage. It records at full fidelity (44100Hz, 16-bit) with no compression, allowing for a maximum phrase length of 10 minutes per slot.
File access via USB is currently not supported. The USB-C port is strictly used for firmware updates and MIDI over USB.
The RECODER requires a 9V DC, center-negative power supply capable of delivering at least 500mA
The product dimensions are 125 x 140 x 50 mm (4.92×5.51×1.97 inches) and it weighs approximately 600 g (1.3 lbs).
| Instrument Input | 1/4" mono jack |
| Main Output | 1/4" mono jack |
| REC Input | 1/4" mono jack, buffered |
| REC Output | 1/4" mono jack, buffered |
| Signal Path | Mono - all inputs, all outputs |
| Built-in Microphone | Electret condenser microphone |
| Power Requirements | 9V DC, center-negative, 500mA |
| MIDI In | 3.5mm TRS Type B |
| USB-C | Firmware updates & MIDI over USB |
| Spectral Processing Engine | High-resolution FFT-based vocoder |
| RESOLUTION Positions | High (1024) / Mid (512) / Low (256) |
| Max INPUT Latency | 40 ms FFT window |
| Total Internal Storage | 16 GB |
| Preset Slots | 234 (A1–Z9) |
| Maximum PHRASE Length | 10 minutes per slot |
| Recording Fidelity | 44100Hz, 16-bit |
| PHRASE Storage Quality | Full fidelity - no compression |
| Product Dimensions | LWH 140 x 125 x 50 mm (5.51 x 4.92 x 1.97 inch) |
| Product Weight | 600 g (1.3 lbs) |