2025-06-01

noise

generates sound using structured broadband energy distributions — filtered, clustered, and animated. forms shape the density, spread, and modulation of energy across time and spectrum.

overview

  • 0 static band

    • description: fixed frequency band with flat spectral energy

    • real-world analogy: pink/brown/white noise filtered by bandpass

    • notes: a sets center, b sets width

  • 1 animated band

    • description: slow modulation of center and width

    • real-world analogy: wind, unstable airflow

    • notes: a sets movement speed, b sets range

  • 2 clustered grains

    • description: bursts of noise in narrow frequency regions

    • real-world analogy: fire crackle, sand

    • notes: a = cluster density, b = duration spread

  • 3 chaotic spread

    • description: deterministic chaos across spectrum

    • real-world analogy: insects, swarm, burning wire

    • notes: based on logistic/baker maps

parameter behavior summary

  • 0 static band

    • a: band center (0 = low, 1 = high)

    • b: bandwidth (0 = narrow, 1 = wide)

  • 1 animated band

    • a: modulation speed

    • b: modulation range

  • 2 clustered grains

    • a: cluster density

    • b: cluster time spread

  • 3 chaotic spread

    • a: chaos intensity

    • b: temporal correlation

parameters are bounded: a, b ∈ [0, 1].

generator structure

all noise is built from additive filtered bands — no procedural or pseudo-random generation. behaviors are repeatable and precomputed. filters and cluster placements are static functions of (a, b) and form id.

why these were chosen

  • banded noise (0–1) covers all steady and modulating natural broadband sources
  • clustered behavior (2) enables bursty microtexture (e.g., rain, fire)
  • chaotic forms (3) introduce dense irregularity without randomness
  • all forms emphasize density, structure, and repeatability

what’s not included

  • spectral noise modulation (handled by dynamics or envelopes)
  • cross-synthesis or ring modulation with noise (use decorators)
  • procedural noise (e.g., Perlin, simplex): non-deterministic by nature

conclusion

the noise domain brings full structural modeling of broadband, non-tonal sources into the system. integrated as a core generator class, it enables:

  • additive, shaped representations of rain, wind, fire, insects, static, and more
  • parametric design of noise textures that are scalable, composable, and repeatable
  • precise shaping and interplay with all other sound and control domains

with sinusoidal and noise, the engine achieves complete coverage of deterministic sonic generators.