Procedural audio and creative sound-design practice
Learning objectives
- learner can build sounds procedurally from parameterized generators (layered filtered noise, coloured noise via DFT) rather than recordings
- learner can adopt creative-practice heuristics — happy accidents, uncontrollability, economy of selection, felt qualities — as a sound-design method
- learner can distinguish deliberate glitch/failure aesthetics and use them intentionally
Capstone — one whole task that evidences the objectives
Create a procedural texture generator (e.g. layered filtered noise for wind or fire), then produce three variations by seeding it from an uncontrollable process — at least one variation must apply a deliberate, engineered glitch/failure treatment as intentional material — and select the most compelling one, documenting the creative decisions including how the deliberate glitch differs from mere accident.
Prerequisite modules
This module is where synthesis stops being about presets and starts being a design practice. The whole task: build a texture — wind, fire, rain — entirely from algorithms and parameters, the way a game-audio designer or live coder must when no sample library fits and sounds have to respond to a running system in real time. On a live-coding rig this is the difference between triggering the same WAV every set and owning a generator whose parameters you can play.
The arc starts supported. First exercises rebuild the two core mechanics under guidance: summing independent filtered-noise generators so each band contributes a distinct character (the layered-composition principle), and colouring white noise by scaling DFT bins with a 1/f^B tendency — these two are drilled repeatedly inside small whole-texture builds until they are automatic. From there the scaffolding shifts from technique to judgment: the atoms on happy accidents and seeding from uncontrollable processes serve as just-in-time pointers when the learner first lets an unfamiliar process drive parameter choices, and the economy-of-selection principle guides the final act of picking one variation from many. The felt-qualities atom gives vocabulary — crunchy, glued, open — for documenting why a variation wins, which the capstone demands in writing.
The required atoms gate the capstone directly: without the procedural-audio framing, the layering and noise-colouring procedures, and the accident/selection heuristics, the three-variations-plus-selection task cannot be completed honestly; the glitch cluster supplies the deliberate-versus-accidental distinction the learner must exercise when engineering the required glitch-treated variation and defending it in the documentation. Supporting atoms enrich rather than gate — spectromorphology names how textures move, extended acoustic techniques and one-off deployment strategies widen the palette, analog imprecision and minimal-craft stories widen the aesthetic frame, and the generative-art definitions situate this practice in its lineage.
Atoms in this module
Required — these gate the capstone
Supporting — enrichment, not gating
Part of curricula
- Synthesist / Sound Designer — deep DSP to a performed live synth rig — Deep DSP — advanced operators, spectral, physical, formant, procedural recommended