Split audio based on silence and min/max chunk lengths

1. FFmpeg (instructions for Mac, Windows)
2. pip install -r requirements.txt

Windows: replace python3 with python

python3 split_audio.py --inp <input_audio> [--min <min_time> --max <max_time> --sil <min_silence> --osr <osrate>]

or

python3 split_audio.py -i <input_audio> [-m <min_time> -M <max_time> -s <min_silence> -S <osrate>]

Options

• input_audio: path to input audio file (with extension)
• min_time: minimum length of segment in s (default: 60.0)
• max_time: maximum length of segment in s (default: 180.0)
• min_silence: min silence duration to split on in s (default: 1.0)
• osrate: output sample rate in whole Hz (default: 22050)

Windows: replace python3 with python

python3 split_audio.py --inp example.wav --max 600.0 --min 60.0 --sil 1.0 --osr 22050

The script outputs the chunks to the <input_audio>_split directory

python3 split_audio.py --help

or

python3 split_audio.py -h