This is not a reference document that walks through every parameter of every device end to end. It is short on purpose. With instruments built this way, an hour spent turning knobs and listening will almost always take you further than an hour spent reading about them.
Treat everything below as a starting point, not an instruction to follow. Open a device, move a few controls at random, and listen to what changes. The parameter explanations are here so you know roughly what you are reaching for, not to tell you where to set them. There is no correct setting. There is only what you want to hear right now.
Nothing in this bundle synchronizes its LFOs to Live's tempo. This is by design. Every device in the series runs on its own internal clock: random timers, drifting cycles, processes that never lock to a grid.
That refusal to quantize is exactly what gives the instruments the quality of a living organism rather than a programmed sequence. Tying the LFOs to the project's BPM would have introduced one mechanical, predictable note in the middle of a bundle built specifically not to have one. it would have broken the coherence of everything around it.
In practice: don't expect any of these devices to lock to your grid, your metronome, or each other. Let them drift. The drift is the instrument.
Every device in this bundle ships at full output gain. This is not an oversight.
Only a basic limiter was applied — no loudness normalization, no inter-sample peak control, no dynamic shaping. The decision was deliberate: noise music and death-ambient have never operated under the assumption that a signal needs to be contained. Clipping, saturation, and inter-sample excess are not problems to correct here — they are part of the timbral vocabulary the bundle was built around.
Pull the output controls down to wherever they need to be for your setup. There is no target level to hit. Trust your ears, not a meter.
Minimum requirements: Ableton Live 12.4 (latest version) with Max for Live.
Every device is a standard Max for Live instrument or audio effect. Drop it on a track like anything else from Live's own library. Every parameter is automatable and mapped for Ableton Push out of the box. there is nothing to configure before you can reach a control with your hands.
A practical way to work: load the device, hit play on whatever is feeding it (or just let it run if it's self-generating), and start moving controls slowly, one at a time. Stop and listen for a few seconds after each move before changing anything else. Most of these instruments respond on a slow timescale. judging a parameter after one second tells you very little.
When something interesting happens, record it. These are not the kind of instruments that return to the same place twice. A configuration you like now is worth capturing now, not "later, once it sounds right." It may never produce that exact moment again, and that's the point, not a flaw to work around.
The smallest instrument in the bundle, reduced to six controls. There is no melody here and nothing to play in the conventional sense: Atrax produces a slow, oppressive low-end mass that you shape rather than perform.
Feedback is the engine: raise it slowly and the drone starts to feed on itself. Rust adds corrosion and electrical instability on top of that. Drift keeps the whole thing from settling into a static tone.
A deep root tone, shaped by a slow rhythmic pulse and run through tape-style degradation until it sits somewhere between a frozen drone and a distant bell.
Pulse_rate sets the speed of the internal throb (remember, it is never tied to Live's tempo). Glacier is the freeze control: pushed up, it stretches the material until it stops moving entirely. Metal adds inharmonic shimmer on top of the root tone.
Erasehead is a fully self-contained instrument. It does not accept external audio — there is no audio input, no sidechain, no way to route material from Ableton into the decks. Everything it works with is generated internally: a growling voice synthesizer (Mouth), an unstable filtered feedback network, and a detuned chord layer.
The two decks (A and B) capture and loop segments of what those internal generators are producing at the moment you press PRINT. That is the only source they have access to. Once something is printed, the deck loops it independently — you can reverse it, change its speed, and degrade it with wow and wear, but you are always working with material that originated inside the device.
Takes a sound source and breaks it into grains, scattering and re-reading them until the original material becomes texture instead of a recognizable sound.
Scatter and readJitter control how far the grains stray from the original timeline. low values keep things close to the source; high values turn it into a cloud with no clear point of origin.
Of all the modules in the bundle, this one rewards a closer look. The underlying process goes beyond traditional grain-based synthesis: instead of simply slicing audio into grains, it searches for the smallest energy fluctuations in the signal, picking out hidden transients, micro-events, and barely perceptible spikes that a conventional granular engine would average over. The effect is a kind of sonic magnification, bringing forward detail that normally stays buried in the background. It performs especially well on low-level or near-static material, surfacing subtle content in a way closer to spectral upward compression than to grain manipulation, emphasizing not what is already obvious, but what would otherwise go unnoticed.
Getting this behaviour right took months of benchmarking against the limitations of more conventional granular tools. A short explanation of what sets it apart: youtube.com/shorts/_hxX1hkWhjM.
Two detuned oscillators and a noise layer carved into a pulsing rhythm by a second internal gain stage, then smeared through a dark, feedback-heavy echo.
RhythmRate and rhythmShape are what carve the sustained tone into a pulse. as with everything in the bundle, that pulse runs on its own clock, not on Live's. tune it by ear against what's happening elsewhere in the track, not by the numbers.
Independent voice clusters fade in and out on their own schedule, occasionally surfacing a buried melodic fragment. Fully self-generating: it doesn't need an input signal to run.
New clusters of voices appear at irregular intervals. that interval is intentionally long and unpredictable, so judging this device in the first thirty seconds will give you the wrong impression of it.
What makes this one behave differently from the rest of the bundle is how it builds: every 18 to 44 seconds (controlled by density) a fresh wave of 10 to 44 voices enters, each with its own multi-second fade-in, sustain and fade-out. New waves arrive before old ones finish, so the layers genuinely accumulate rather than alternate — over 3 to 5 minutes the texture settles into a dynamic plateau. Lowering density does not undo this quickly: voices already alive keep living out their full span, so a crescendo started a few minutes ago takes roughly as long to recede as it took to build.
Because of that same accumulation, loudness here is not fixed by where you set output at the start — it keeps building underneath you if density is high. Start with output low (roughly the first tenth of the dial's range) and bring it up gradually once you can hear where the texture is heading, rather than setting a level in the first few seconds and walking away.
A fragile melodic cell resurfaces irregularly, repeating and decaying over a slow independent drone layer, as if being recalled from a damaged recording rather than played.
Memory controls how intact the melodic cell stays each time it returns. low values fragment it further with each repetition. Transpose shifts the whole register down, useful for moving the instrument out of the way of other material.
Built from layered noise beds rather than tonal material: mist, swarm, and electricity textures sit under sparse pressure tones and occasional call phrases. Closer to a field recording instrument than a synthesizer.
Swarm and electricity add dense, high-frequency noise events. Distance pushes everything further back and adds grain, useful when you want the texture to sit behind other elements rather than in front.
Slow, melancholic harmonic clusters wrapped in tape character, stretched by reverb into a cavernous, mournful space.
Out of a long list of parameters, two are worth living with longer than the rest: chaos and darkness. between them they cover most of the emotional range of the device, from solemn and clear to grainy and unstable.
The most complete voice in the bundle. dual oscillators with sub and noise layers, ring modulation, two LFOs, a resonant filter with envelope and formant shaping, drive, bit-crush, tape delay, a three-band EQ, and full ADSR — 39 continuous parameters in total, played live from a 13-key on-screen keyboard (or MIDI) rather than sequenced.
It's monophonic, last note has priority, and includes a hold/latch mode for sustaining a note indefinitely without holding a key, plus a panic trigger that cuts to silence in a fixed 30ms regardless of the release setting — useful since notes here don't pass through Live's native MIDI engine, so a missed note-off can otherwise hang. Six presets are included as starting points, alongside full randomization across all 39 parameters.
The timbral reference point is the cold, clinical electronics of Close To A Corpse by Atrax Morgue (originally on Slaughter Productions) — not an emulation, but the territory Tanatoprassi's drive/bitcrush/tape-delay chain was built to sit in. One of the factory presets, Corpse-Synthesizer, takes its name directly from a track on that record.
Tape Reamp takes a loaded sample and runs it through a degraded playback chain. The source doesn't get processed so much as it gets mishandled — slowed, looped with unstable memory, filtered through cabinet coloration, layered with hiss, and subjected to wow instability that compounds across repetitions. At lower speed values the material barely moves. At higher memory settings it starts to forget where it's been.
The six parameters cover the full arc from relatively functional playback to a machine that is clearly failing. speed and memory together define the loop character; cab and hiss set the physical environment the tape appears to be stored in; wow introduces slow pitch drift independent of speed.
Post Rave is a granular engine built around a loaded sample. Grains are scattered in time and pitch, biased toward low frequencies, and pushed through a reverb and waveshaping chain that removes any suggestion of the original material being intact. The name describes the register: not a club sound, but what's left after one — worn textures, smeared low-end, the acoustic residue of a space that was recently very loud.
Ten parameters cover both micro-structure and macro-character. density and stretch govern how many grains are active and how long each lasts; low biases pitch transposition downward; ghost adds a displaced echo copy of the grain cloud; smear shifts band positions horizontally; fog controls reverb depth; rust drives the waveshaper and high-pass filter simultaneously; width and drift control stereo spread and pitch instability; out sets the master level.
1991 does not scatter grains across a buffer. It locks onto a single fragment — between one and twenty-two milliseconds — and repeats it obsessively with almost no movement, until repetition itself replaces the original sound. The source material becomes irrelevant. What remains is the mechanism.
The engine includes an internal generative source that requires no external sample: a six-second synthetic field of electrical interference, sub drone, metallic resonance and tape dropout, built specifically for this kind of locked obsession. Load your own material or use the internal source — the behaviour is the same either way.
density and window define the rate and size of each grain; lock sets the position in the buffer where the engine anchors; drift allows micro-movement around that position; fault introduces small random jumps; scatter introduces large violent ones. reverse sets the probability that any individual grain plays backwards; pitch spreads intonation across the grain field; cut controls the lowpass filter; click sharpens the attack envelope; reverb adds a cavernous convolution tail with feedback delay.
Three instruments in this bundle use granular synthesis as their core mechanism. Each one uses it differently enough that the connection is not immediately obvious — but taken together they form a complete map of what the technique can produce.
Mesostruttura operates at maximum density. Thousands of grains per second are scattered across the entire buffer, treating the source as a spectrum to be magnified rather than a sound to be played. It functions as a sonic microscope — the interest is in what is hidden inside the recording, not the recording itself.
Post Rave operates at medium density with longer grains, suspending material in space until it loses its origin and becomes texture. The interest is not what the source was but what it becomes after the engine has stretched, displaced and eroded it across time and frequency.
1991 operates at the opposite extreme. It locks onto a single fragment and repeats it until repetition itself becomes the instrument. Not a cloud, not a texture — a single point held under pressure until it changes character on its own. The internal source means it does not even require a sample: the engine is already carrying its own material.
Spectral microscopy, spatial erosion, locked obsessive repetition. Three results from one mechanism. Use them in sequence on the same source to hear the full range.
A fourth device works on loaded audio without granular processing: Tape Reamp runs any source through a degraded playback chain — speed instability, cabinet coloration, wow and flutter, hiss and magnetic decay — treating the sample not as material to fragment but as a physical object that is slowly failing. Together with the three granular engines it completes a full toolkit for transforming recorded sound into something that no longer quite belongs to the present.
A note on Push mapping: Panacousticon is built for Ableton Push and requires a Live set to run — it is not a standalone application. All original devices in the bundle are fully mapped for Push. The granular arc instruments (Mesostruttura, Post-Rave Soundscape and 1991) do not yet have a dedicated Push mapping; this may be added in a future update. In the meantime all parameters remain accessible from the device view in Live.
If you are looking for a bundle compatible with Push 3 Standalone — instruments that run without a computer — the Brain Recordings Bundle was built specifically for that context.