This repository has the source code for computing spectrogram with torch. When processing audio signals, it can be done quickly in batch processing.
STFT
: same with librosa.stftMelSpectrogram
: same with librosa.feature.melspectrogramLMLFSpectrogram
: same with madmom.spectorgram.LogarithmicFilteredSpectrogramProcessor
If you run torch_spec.py
, you can see the result as shown below.
- Jongwook Kim's https://github.com/jongwook/onsets-and-frames
- Prem Seetharaman's https://github.com/pseeth/pytorch-stft
madmom
: https://madmom.readthedocs.io/en/latest/index.htmllibrosa
: https://librosa.github.io/librosa/index.html