2025-05-30

parametric composer

this document series develops a "super instrument" (a unified additive-synthesis engine spanning distinct spectral behaviors) and a "super composer" (mapping over a breadth of distinct compositional behaviors).

all structure and sound arises from static, hierarchical domains. these domains do not merely replicate known musical idioms, but instead span a multidimensional space of sonic and temporal forms - enabling the construction of both familiar and previously inaccessible behaviors.

general structural rules, design philosophy, and template logic that apply across all layers, domains, and forms.

minimal viable song engine required domain generators

timbre

defines how spectral energy is shaped, transformed, and distributed over time.

amplitudes defines how energy is distributed across partials in the spectrum. each form produces a distinct spectral shape - flat, tilted, peaked, or chaotic - suitable for everything from simple tones to turbulent noise-like bodies.

decorators defines partial-level augmentations such as detuning, sideband FM, or noise injections. forms also control how these effects are applied across the partial spectrum — randomly, periodically, or by region.

dynamics defines time-varying transformations applied within a note's lifespan, including spectral morphing, gated filtering, and stochastic grain skipping. each follows a structured temporal contour.

envelope defines the shared amplitude contour for a notes duration using parametric spline shapes (linear, exponential, power, bezier arc). forms express diverse attack and release behaviors without conditionals or gating.

envelopes defines how the envelopes properties vary by partial index - allowing individualized attack, decay, sustain, and delay across the spectrum. used for sculpting spectral motion and micro-timbral events.

frequencies defines the frequency spacing of partials. forms range from harmonic, stretched, and inharmonic to modal and stochastic, each describing a different structural view of the frequency domain.

onset

defines when things happen - in both regular, musical ways and irregular, naturalistic or chaotic ways.

density statically allocates onset slots over a given time base, independent of content. usable with both metrical (`grid`) and non-metrical (`field`) onset domains. forms define uniform spreads, clusters, and deterministic distributions.

field defines non-metrical onset distributions: stochastic clouds, chaotic intervals, continuous-time drifts. used for naturalistic textures (rain, fire, random impulses) and non-cyclic materials.

grid defines a metrical pulse lattice: repeating time structures based on time signatures, subdivisions, or polymeters. ideal for structured music with cycles, measures, and groove.

offset introduces structured timing deviations to incoming triggers: swing, jitter, or timing curves. transforms onset positions without altering their identity. operates atop both metrical (`grid`) and non-metrical (`field`) sources, with some forms only meaningful in grid contexts.

pattern defines reusable rhythmic motifs mapped onto a stream of triggers. patterns impose local structure and phrasing atop incoming onset lists, regardless of whether they are generated by `grid` (metrical) or `field` (non-metrical) domains.

voicing assigns incoming triggers to playback voices under static polyphony constraints. determines which voice plays each event using rules like round-robin, priority, or weighted selection.

interplay

defines relationships between voices, tracks, or repeated events - emphasizing time-based interaction rather than isolated behavior.

interaction defines inter-voice relationships such as call & response, phasing, canon, and hocketing - all pre-resolved and not runtime-reactive.

mutation defines evolving modifications to sequences over time through accumulation, reset, and template-based operations. enables generative variation, progressive transformation, and non-repetitive behaviors - all fully deterministic.

repetition defines how motifs or sequences repeat: as copies, decaying echoes, nested variants, or time-shifted mirrors.

variation defines systematic, stateless departures from a base behavior using transformation rules, pseudo-random perturbations, and probabilistic filters. adds local or global diversity without memory, accumulation, or temporal evolution.

pitch

defines how frequency identities are selected, generated, and mapped.

atonal non-tonal pitch collections: chromatic continua, micro-grids, fixed serial rows, and spectral-derived sets-none rely on scalar centricity or cyclic resolution.

generative algorithmic pitch sequencers - deterministic processes that unfold rule-based, evolving patterns from compact parameter seeds.

tonal cyclic, bounded pitch structures-finite sets with internal order, repetition, or symmetry, regardless of musical tradition.

tuning defines how pitches are mapped to actual frequencies. supports tempered grids to ratio-based or freely defined arrays.

structure

defines large-scale temporal form - how sections, cycles, or phrases are divided and arranged.

hierarchy nests or layers segments recursively to create deep structural repetition and variation - supporting fractal forms, mosaics, and branching trees.

ordering determines the sequence in which segments are played - linear, deterministic "random," palindromic, or looped - defining macro-temporal navigation over a pre-segmented timeline.

segmentation divides time into discrete regions via equal durations, proportional ratios, or custom breakpoints-these segments anchor change and transformation.

selection defines how to generate, analyze, and curate sets of sequences using combinatorial enumeration and deterministic statistical grouping. enables structured exploration of pattern space and mapping of chosen sequences to control domains.

spatial

defines the position, movement, and spread of voices in an abstract or physical sound field.

diffusion defines the apparent size or spread of a source: from tight point sources to fully spatialized textures. may be mapped per-frequency or per-partial.

motion continuous spatial movement: circular motion, LFO-driven panning, chaotic wander, or path-following.

position static placement: center, stereo spread, frequency-dependent bias, or randomized positions.