The modules

audiodsp is CircuitPython’s audio system for MicroPython and CPython: audiocore, audiomixer, synthio (including MidiTrack), the effects modules (audiofilters, audiodelays, audiofreeverb, audiospeed), audiomp3, and CircuitPython’s play(), stop(), pause() and resume() output semantics. Import names stay audiocore, synthio and the rest, matching CircuitPython, so source runs unchanged; the few behaviours that differ are in Where audiodsp differs from CircuitPython.

It adds modules CircuitPython doesn’t have (why each exists). Each has a page here or a guide:

Module

What it does

Where

audiobiquad

Biquad and all-pass filters whose tails reach exact zero

guide, reference

audiodynamics

Compression, limiting, expansion, gating, transient shaping

reference

audioecho

A delay with a filter, a soft-clip and a cross-feed in its feedback loop

reference

audioroute

Splitter fans a stream out; MidSide scales stereo width; Port is a re-pointable wire

MidSide, Port guide

audioshaper

A waveshaper whose curve is data, and an exact-ratio sample-and-hold

reference

audioladder

A transistor ladder filter

reference

audioverb

A reverberation tank whose line lengths and taps come from Python

reference

audiomath

Multiply one stream by another, divide one down in frequency, and cast_into

reasons

audioconvolve

Apply an impulse response by partitioned FFT

reasons

audiooscillator

A band-limited oscillator read sample by sample

reference

audiomodal

A bank of resonators

reference

audiograph

A whole patch rendered by one C call per block

reference, how it works

audiosampler

Recorded notes played at any pitch, looped, under an envelope

guide

audiometer

Band levels for a spectrum meter

guide

audiofeatures

Speech features and voice activity for a wake word model

guide

audiovorbis

Ogg Vorbis files

guide, C interface

audiopump

The audio pull, off the interpreter thread

guide, reference

Beside the core

Two pure-Python modules sit beside the CircuitPython-compatible core:

  • lib/audiodsp_util/ — float32(value) and float32_bits(value), struct and nothing else. A Python float is the interpreter’s own width: a double here, a single on every board. So a setting written in Python — node.mix = 0.35 — is two different numbers, and the node renders different bytes on a board than on a desktop before its kernel is reached. Put a derived setting through float32 and both targets hold the number the board would have held. See docs/correctness-standard.md.

  • lib/audiorender/ — renders a whole composition offline: tracks, tempo map, notes and automation in, a mixed stereo master and a level report out. This is the one part of the repository written for a desktop rather than a board — numpy throughout, a whole song in memory — so it ships in the wheel and is never frozen into firmware. It has its own README.

Instruments and effects built on these modules are in their own libraries, audioinstruments and audioeffects, which also define the component contract they share. audiorender drives any component that speaks that contract and doesn’t need either library installed. What stays the same across releases, for audiodsp and for them, is in Sound stability.