A real-time audio time-scale modification library¶
AudioTSM is a python library for real-time audio time-scale modification procedures, i.e. algorithms that change the speed of an audio signal without changing its pitch.
- Source code repository and issue tracker:
- Python Package Index:
- MIT – see the file
Audiotsm should work with python 2.7 and python 3.4+.
You can install the latest version of audiotsm with pip:
pip install audiotsm
pip install audiotsm[gstreamer]
If you want to play the output of the TSM procedures in real time, or to use the examples, you should install audiotsm as follow:
pip install audiotsm[stream]
The audiotsm package implements several time-scale modification procedures:
- OLA (Overlap-Add);
- WSOLA (Waveform Similarity-based Overlap-Add);
- Phase Vocoder.
The OLA procedure should only be used on percussive audio signals. The WSOLA and the Phase Vocoder procedures are improvements of the OLA procedure, and should both give good results in most cases.
If you are unsure which procedure to choose, the Phase Vocoder should sound best in most cases. You can listen to the output of the different procedures on various audio files and at various speeds on the examples page.
Below is a basic example showing how to reduce the speed of a wav file by half using the WSOLA procedure:
from audiotsm import phasevocoder from audiotsm.io.wav import WavReader, WavWriter with WavReader(input_filename) as reader: with WavWriter(output_filename, reader.channels, reader.samplerate) as writer: tsm = phasevocoder(reader.channels, speed=0.5) tsm.run(reader, writer)
If you are interested in time-scale modification procedures, I highly recommend reading A Review of Time-Scale Modification of Music Signals by Jonathan Driedger and Meinard Müller.