EdytLab
ShippedDescribe an audio edit in English. It renders.
Drop in stems and ask for the mashup you want. An agent plans the operations, the DSP runs locally in Rust, and stem separation and transcription happen on your own machine.
- Rust
- Tauri 2
- symphonia
- fundsp
- ONNX
- Demucs
- Whisper
- Pure Rust DSP with ONNX running Demucs and Whisper on device.
- commits
- The agent plans a sequence over them; session state lives in a branchable graph.
- editing tools exposed to the agent
Audio editing has a vocabulary problem. You can hear the edit you want long before you can find the six menu operations that produce it, and that gap is where most people give up.
A conversational layer over a modular Rust DSP engine. Around twenty editing operations are exposed as tools, an agent plans a sequence, and session state lives in a branchable DAG so any plan can be undone or compared against an alternative.
Keeping it local. Stem separation via Demucs and transcription via Whisper both run through ONNX Runtime on device, so the only thing that leaves the machine is the language model request. Decode, routing, effects, resampling and IO are all pure Rust.
Phase one and two complete, single-track editing and mashups, with conversational mixing next. Branchable session graphs, A/B comparison and multi-provider model switching all working.