spiridonovpolytechnic/audio-pipe repository overview

⁠audio-pipe

A simple audio processing pipeline built around SoX and its python API.

Note: Only some of the SoX functionality is exposed here. If you want more functionality you can request it in Issues, or contribute. Adding a new transform is fairly easy: (1) copy an existing transform in the src/audio/transforms directory, (2) modify the copied class per pysox documentation (https://pysox.readthedocs.io/en/latest/api.html#⁠) for whatever transform you want to add, (3) add a hydration function to src/audio/audio_factory.py, (4) add a working test to test-workspace.

⁠How to use

1. Clone this repo and navigate into the project directory

2. To get a feel for how different transforms affect your sounds you can process the test samples by running the command below.

docker run \
    -v $(pwd)/test-workspace:/workspace \
    spiridonovpolytechnic/audio-pipe:latest

1. Put your input audio files in the workspace directory

2. Write your audio pipeline recipes (check test-workspace directory for some examples)

3. Run the container

docker run \
    -v $(pwd)/workspace:/workspace \
    spiridonovpolytechnic/audio-pipe:latest

1. Check the workspace/output directory for the processed files

⁠Local testing

docker build -t audio_pipe .

docker run -v $(pwd)/test-workspace:/workspace audio_pipe