crema

convolutional and recurrent estimators for music analysis

Usage options

From the command-line, print to the screen in JAMS format:

python -m crema.analyze file.mp3

or save to a file:

python -m crema.analyze file.mp3 -o file.jams

From within python:

from crema.analyze import analyze

jam = analyze(filename='/path/to/file.mp3')

or if you have an audio buffer in memory, librosa-style:

jam = analyze(y=y, sr=sr)