prologue

What is Signals & Sorcery?

Signals & Sorcery is a Generative Audio Workstation (GAW) — a new kind of music tool built around contract-based composing. Define a musical contract (key, chords, tempo, bars), then let AI generate MIDI via Gemini and audio via Lyria within those constraints. Preview privately in headphones, refine, and push approved material to your audience in real-time.

The platform features hardware-level audio routing: separate headphone (cue) and main outputs let you audition AI-generated clips privately before your audience hears them.

Who is this for?

  • Generative Music Performers who want to create and perform AI-generated music live
  • Electronic Musicians looking for a new approach to live performance with AI assistance
  • Live Coders & Experimental Artists who want intuitive AI-driven music generation
  • Streamers & Content Creators who want to generate music in real-time while broadcasting
  • Anyone who finds traditional live performance setups limiting and wants to explore generative approaches

How does it work?

Core Technology

  • Native Audio Engine: Built on Tracktion Engine (C++) with JSON-RPC communication
  • AI-Powered Generation: Uses Gemini for MIDI generation and Lyria for audio generation
  • Contract-Based Composing: Define key, chords, BPM, and bars — AI composes within those constraints
  • Dual-Deck Architecture: Loop A (composition/preview) and Loop B (performance/audience) with independent routing
  • Plugin SDK: All generators (synths, samples, audio textures) are built on the extensible Plugin SDK
  • Custom Instrument Support: Load any VST3/AU instrument plugin on synth tracks

The Performance Workflow

  1. Define - Set up a musical contract (key, chords, tempo, structure)
  2. Generate - AI composes MIDI and audio that fits your contract
  3. Preview - Generated clips play in your headphones (cue output)
  4. Refine - Iterate on the generation until you're satisfied
  5. Perform - Push approved clips to the audience (main output)

Key Features

  • Contract-Based Generation: Musical contracts ensure coherent compositions across tracks
  • Gemini MIDI + Lyria Audio: Purpose-built AI models for music generation
  • Private Preview: Audition generated content in headphones before the audience hears it
  • Deck Transfer: Move material from composition to performance deck
  • Plugin SDK: Built-in synth, sample, and audio texture generators — with upcoming integrations for Splice, ElevenLabs, live coding, and agentic prompting
  • Custom Instruments: Load any VST3/AU synth plugin alongside or instead of the default Surge XT

Current Limitations

  • macOS Only: Tested on M-series & Intel chips (Windows/Linux not supported)
  • Surge XT Default: Ships with Surge XT as the default synth, but any VST3/AU instrument can be loaded
Last Updated:
Contributors: Steve Hiehn, shiehn, stevehiehn, Steve Hiehn