Build and share

Use audio models creatively, build your own instruments and tools, and bring them to the people who'll play them.

Audio machine learning and creative sensing

What models do with sound, where they shine and where they struggle, and how to use them in your own tools.

  1. What an audio model learnsTraining, inference and why a model's output is an estimateComing soon
  2. How models see audioWaveforms, spectrograms, latents and tokensComing soon
  3. Stem separationHow models pull drums, bass and vocals out of a mix, and where they struggleEarly draft
  4. Audio generationHow models make new sounds from noise, tokens or promptsComing soon
  5. Neural instruments and timbre transferInstruments whose sound comes from a trained modelComing soon
  6. Camera and gesture controlTurning movement into musical controlComing soon
  7. Adding a model to your appCapability checks, downloads, workers, progress and cancelComing soon

Build your own tools

For developers: from a musical idea to an effect, instrument or MIDI tool that runs in the browser and beyond.

  1. Signals and systemsLinearity, memory and why EQ and distortion behave so differentlyComing soon
  2. Anatomy of a Maru moduleDeclaration, ports, parameters, processor and faceComing soon
  3. Build an effectFrom a musical promise to a working processorComing soon
  4. Build an instrumentVoices, notes, envelopes and stuck-note safetyComing soon
  5. Build a MIDI toolEvent timing, routing and deterministic randomnessComing soon
  6. Realtime and numerical safetyDenormals, allocation, smoothing and block boundariesComing soon
  7. Testing sound and behaviourSilence, clipping, DC, extreme settings and listeningComing soon
  8. Presets, state and automationMaking saved sessions come back exactly as they wereComing soon
  9. Shipping to browser and pluginsTargets, formats and what each one has been tested onComing soon
  10. Building with an AI agentBriefs, reviewing changes and catching unsupported claimsComing soon

Instrument design and publishing

What makes an instrument inspiring, and how to bring one to the people who'll love it.

  1. What makes an instrument inspiringImmediacy, response, identity and depthComing soon
  2. Controls, macros and rangesMaking every knob worth turningComing soon
  3. Presets as product designA default that sells it and a set with distinct jobsComing soon
  4. Interface and first useLabels, hierarchy, accessibility and the first minuteComing soon
  5. Finding an audienceWho it's for and how to test whether they careComing soon
  6. Identity and storyTurning an artist's sound into something playableComing soon
  7. Demos, pricing and launchHonest demos, costs, channels and a first launch experimentComing soon
  8. Rights and maintenanceLicences, credits, updates and supportComing soon