A python package for music and audio analysis.

See http://librosa.github.io/librosa/ for a complete reference manual and introductory tutorials.

What does librosa do? Here are some quick demonstrations:

• Introduction notebook: a brief introduction to some commonly used features.
• Decomposition and IPython integration: an intermediate demonstration, illustrating how to process and play back sound

The latest stable release is available on PyPI, and you can install it by saying

pip install librosa

Anaconda users can install using conda-forge:

conda install -c conda-forge librosa

If you use Anaconda on Windows, we recommend installing the gstreamer and/or ffmpeg libraries separately, as they are not (yet) available as conda packages.

To build librosa from source, say python setup.py build. Then, to install librosa, say python setup.py install. If all went well, you should be able to execute the demo scripts under examples/ (OS X users should follow the installation guide given below).

Alternatively, you can download or clone the repository and use pip to handle dependencies:

unzip librosa.zip
pip install -e librosa

or

git clone https://github.com/librosa/librosa.git
pip install -e librosa

By calling pip list you should see librosa now as an installed pacakge:

librosa (0.x.x, /path/to/librosa)

To fuel audioread with more audio-decoding power, you can install ffmpeg which ships with many audio decoders. (Note: if you are using the conda package for audioread, this will be done automatically.)

You can use homebrew to install the program by calling brew install ffmpeg or get a binary version from their website https://www.ffmpeg.org.

Please direct non-development questions and discussion topics to our web forum at https://groups.google.com/forum/#!forum/librosa

Please refer to the Zenodo link below for citation information.