home/ modules/ performing-a-live-synth-rig

Performing a live synth rig: real-time granulation, scripted modulation, and free-rhythm sets

  • learner can perform a live set that captures incoming audio into a looping buffer and granulates it in real time with a trailing pointer
  • learner can drive expressive real-time modulation using scripted (Lua Formula) modulators and audio-rate texture generators (pulsar, waveset)
  • learner can shape set-level pacing by automating master tempo for free-rhythm feel and by organizing granular material across meso and macro scales
  • learner can exercise economy of selection and expert heuristic judgment under time pressure, overriding formal rules when perception demands it
  • learner can deploy an analytically-designed drum palette (iconic 808/909-character voices) with intentional, DSP-literate signal-processing choices

Perform and record a continuous 8-12 minute live synth-rig set: capture a live audio source into a looping buffer and granulate it in real time, layer pulsar/waveset textures, drive modulation from a scripted Lua Formula modulator, back it with an analytically-designed iconic-drum-machine palette, and steer the whole set's pacing with automated master tempo (free rhythm) while making live economy-of-selection and heuristic composition decisions.

This module builds toward the thing granular synthesis was always meant for: standing in front of an audience with a hybrid rig — SuperCollider for microsound, Surge XT for voices, a DAW as the set’s spine — and improvising an 8-12 minute textural set in the lineage of Objekt or Roads-school electroacoustics, where nothing is safely pre-sequenced. Live input becomes material the moment it arrives; pacing comes from tempo you bend by hand, not a grid.

The arc starts supported. First, wire up the circular-buffer architecture from “Live granulation requires recording to a looping buffer first” in a studio setting, learning by ear why the grain pointer must trail the record head. In parallel, script a first “Formula modulator” in Lua that breathes life into one Surge patch, and rehearse “pulsar synthesis” and “waveset manipulation” as texture generators until switching between pitched, buzzy, and glitched regimes is a reflex — these three are the part-task drills, because on stage there is no time to think about pointer offsets or duty cycles. Next, design the drum palette offline: the 808/909 waveform-analysis principle plus DSP-literate before/after listening make those voices intentional rather than preset-shaped. A dress rehearsal adds the automated master-tempo lane, practicing free-rhythm pacing before the unsupported capstone recording.

The required atoms are exactly what the capstone cannot survive without: the granulation plumbing, the modulation and texture engines, the tempo and multiscale-organization moves, and the judgment principles — economy of selection, heuristics over formalism — that govern every live decision. The supporting atoms (physical-modeling voices, Ambisonic spatialization, envelope internals, spectral scale design) enrich the rig’s palette and deepen the DSP literacy the set draws on, but the performance stands without them.

Runnable examples

Generated from the context/ instrument corpus by concept (redistributable idioms only). Do not edit — regenerate with gen-module-examples.mjs.

feedback-loop

{ CombN.ar(Impulse.ar(2) * 0.3, 0.2, 0.2, 3) }.play

supercollider-0015 · CC0

s("cp").delay(0.5).delaytime(0.166).delayfeedback(0.7)

strudel-0018 · CC0

physical-modeling

Mandolin m => dac; 0.9 => m.pluck; 220 => m.freq;

chuck-0043 · MIT

Atoms in this module

Required — these gate the capstone

Live granulation requires recording to a looping buffer first, then running GrainBuf on the buffer with a trailing pointer
Procedure L4 Performance BF
Surge XT's Formula modulator runs Lua scripts that read voice/MIDI state and output arbitrary modulation curves
Concept L4 Performance B
Automating master tempo in a DAW removes the fixed rhythmic grid, creating felt compression and expansion rather than locked-in groove
Procedure L4 Performance B
Economy of selection — picking the salient few from a vast field — is the irreducibly human element in generative composition
Principle L4 Performance BF
The character of iconic drum machines like the 808 and 909 lies in fine waveform details that are easy to state in principle but hard to replicate exactly
Principle L4 Performance B
Developing technical DSP literacy transfers directly to more analytical and intentional sound design and production
Principle L4 Performance BM
Pulsar synthesis generates pitched textures by combining a waveform table with a duty-cycle envelope repeated at audio rate
Concept L4 Performance BF
Waveset manipulation treats individual waveform cycles as grains for extreme time-stretching and morphing in SuperCollider
Concept L4 Performance BF
The unsolved problem of granular composition is coherent structure at meso and macro scales, not just generating clouds
Principle L4 Performance B
Heuristic algorithms pair computational power with expert judgment to guide composition, unlike borrowed formal models
Principle L4 Performance BF
Elegant formal rules do not guarantee compelling music because music works through perception, not logical consistency
Principle L4 Performance BF

Supporting — enrichment, not gating

Constant-rate ADSR release keeps the fall rate fixed, so releasing from a lower sustain takes less time
Principle L4 Performance B
A single targetRatio parameter morphs an ADSR segment's shape from near-exponential to near-linear
Principle L4 Performance B
All-pass filters provide delay without altering the spectrum but with frequency-dependent phase response
Concept L4 Performance BE
Ambisonic A-format (tetrahedral microphone raw output) is converted to B-format via a matrix transcoder with correct orientation
Procedure L4 Performance B
Correct Ambisonic format conversion requires matching four parameters: order, component ordering, normalization, and reference radius
Principle L4 Performance BF
Encoding a source at a small radius adds proximity bass boost but distorts dangerously near zero, so clamp the radius and high-pass first
Principle L4 Performance B
Coupled strings interact through a shared non-rigid bridge modeled as a single shared loss filter
Concept L4 Performance BE
Fixed-length variable-step delay lines keep decay time constant across pitch, unlike variable-length lines
Principle L4 Performance BE
Fractional delay times require interpolation, which trades tuning accuracy for frequency-dependent attenuation
Principle L4 Performance B
Dictionary-based pursuit decomposes any sound into grains, enabling analysis-driven transformations
Concept L4 Performance B
A digital waveguide section models wave propagation with two delay lines, six attenuators, and two summers per section
Concept L4 Performance BE
Higher-order Ambisonics adds degree-n spherical harmonics to increase soundfield spatial resolution at the cost of (n+1)² channels
Concept L4 Performance B
Guitar feedback can be simulated by feeding Karplus-Strong output through a nonlinear shaper back into the delay line
Procedure L4 Performance BE
First-order Ambisonics encodes a 3D soundfield as four channels (W, X, Y, Z) decoded for any speaker layout
Concept L4 Performance BF
Designing a scale for an inharmonic instrument requires FFT analysis followed by computing dissonance curve minima
Procedure L4 Performance BA
Spectral mapping transforms the partials of one sound to a target spectrum while preserving the overall tonal character
Procedure L4 Performance B
Inverting the sign of a waveguide's reflected wave drops the pitch an octave and cancels DC buildup
Procedure L4 Performance BE
A multi-section string waveguide simplifies to a single delay line with one lowpass filter by exploiting linearity
Principle L4 Performance BE
Woodwind synthesis requires a nonlinear reed model to provide gain and generate harmonics
Concept L4 Performance BE
Woodwind pitch is changed by tone holes that selectively reflect low frequencies and transmit high frequencies
Concept L4 Performance BE