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
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)
Note that audioread needs at least one of the programs to work properly.
librosa uses audioread to load audio files.
To fuel audioread with more audio-decoding power (e. g. for reading MP3 files),
you can either install ffmpeg or GStreamer.
If you are using Anaconda, install ffmpeg by calling
conda install -c conda-forge ffmpeg
If you are not using Anaconda, here are some common commands for different operating systems:
- Linux (apt-get):
apt-get install ffmpegorapt-get install gstreamer1.0-plugins-base gstreamer1.0-plugins-ugly - Linux (yum):
yum install ffmpegoryum install gstreamer1.0-plugins-base gstreamer1.0-plugins-ugly - Mac:
brew install ffmpegorbrew install gstreamer - Windows: download binaries from the website
For GStreamer, you also need to install the Python bindings with
pip install pygobject
Please direct non-development questions and discussion topics to our web forum at https://groups.google.com/forum/#!forum/librosa
If you want to cite librosa in a scholarly work, there are two ways to do it.
-
If you wish to cite librosa for its design, motivation etc., please cite the paper published at SciPy 2015:
McFee, Brian, Colin Raffel, Dawen Liang, Daniel PW Ellis, Matt McVicar, Eric Battenberg, and Oriol Nieto. "librosa: Audio and music signal analysis in python." In Proceedings of the 14th python in science conference, pp. 18-25. 2015.
-
If you are using the library for your work, for the sake of reproducibility, please cite the version you used as indexed at Zenodo:
