Split stems and transcribe notes from audio files
An audio processing application that combines stem separation and music transcription. Built to analyze audio files by splitting them into individual instrument tracks (stems) and transcribing the musical notes.
Features
- Stem Separation: Uses Demucs to split audio files into individual instrument tracks
- Note Transcription: Extracts musical notes from audio
- Web Interface: Modern web UI built with TypeScript, JavaScript, and HTML/CSS
- Python Backend: Robust backend with comprehensive testing and type checking
Technical Stack
- Python 3.9 (required for Demucs compatibility)
- Poetry for dependency management
- Demucs for stem separation
- TypeScript/JavaScript frontend
- pytest for testing with coverage
- Ruff for linting
- mypy for type checking
- tox for comprehensive test automation
Development
The project follows best practices with:
- Comprehensive test suite
- Type checking
- Linting and code quality tools
- Automated testing with tox