AudioSampleForge is a Python-based audio processing toolchain that allows you to download, extract, clean, and edit audio samples through a simple command-line interface and an intuitive waveform editor.
- Download audio from YouTube
- Extract and normalize audio segments
- Clean raw samples using deep learning models
- Visually edit waveform segments with a modern UI
- Export curated samples ready for use
The waveform editor allows you to select, move, delete, or insert silences between segments. It provides real-time playback and duration control for fine-tuning your samples.
Below is a snapshot of the editor, launched via the
audiosampleforge-servecommand:
To install the project locally using uv:
uv sync
uv tool install . -eHere's the basic workflow:
uv run audiosampleforge-download --url https://www.youtube.com/watch?v=6VAF1YThcbc --out data/0_raw_audiouv run audiosampleforge-extract --input data/0_raw_audio/segment.wav --start 0.0 --dur 10000.0 --out data/1_normalized_audio/segment.wavuv run audiosampleforge-serve --input data/1_normalized_audio/segment.wavuv run audiosampleforge-clean --input data/2_raw_sample/segment.wav --out data/3_clean_sample/segment.wavuv run audiosampleforge-cleansr --model speech --input data/output/clean.wav --out data/output/clean2.wavdata/
├── 0_raw_audio/
├── 1_normalized_audio/
├── 2_raw_sample/
├── 3_clean_sample/
└── output/
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change or improve.