Multi-Track Audio Sampler

A simple Multi-Track Audio Sampler to splice up bars consistently across synchronised audio stems.

It takes up multiple synchronised .wav stem tracks (bass, drums, vocals, etc.) and slices them based on your selected musical segments (bars / milliseconds).

Exports the resulting samples into new files in the same directory.

REQUIREMENTS

ONLY WORKS UP TO PYTHON 3.11.x!

Python 3.13+ is not supported due to removed dependencies.

Dependencies:

• pydub
• ffmpeg

INSTALLATION

1. Clone the repo

git clone https://github.com/yourusername/multitrack-audio-slicer.git
cd multitrack-audio-slicer

2. Install Python 3.11

brew install [email protected]

if encountering filepath issues, run, otherwise ignore

brew link [email protected] --force --overwrite

3. Create and activate a virtual environment

/opt/homebrew/bin/python3.11 -m venv venv
source venv/bin/activate

4. Install dependencies

pip install pydub
brew install ffmpeg

USAGE

1. Place .wavfiles in same folder as the script.
2. Update the files dictionary with your filenames.
3. Define segments/regions to splice up using either of the two options:
   • Bar ranges: (start_bar, end_bar)
   • Milliseconds: (start_ms, end_ms)
   • Set USE_BARS to True/False respectively.
4. Run script

python AudioSampler.py

NOTES

• You can adjust SHIFT_MS - Optional manual offset (in milliseconds) to compensate for silence or pre-roll at the start of the audio file.
• If you re-run script with changes made, originally generated files get overwritten (fix by changing filenames!)
• I recommend using Demucs if you have your audiofile but the different stem tracks are not split up yet!