nussl 0.1.0

A flexible sound source separation library.

nussl (pronounced like “nuzzle”) is a flexible python source separation library created by the ![Interactive Audio Lab] (http://music.cs.northwestern.edu/) at Northwestern University.

Features

At its core, nussl contains implementations of the following source separation algorithms: * DUET * REPET * NMF * KAM * RPCA

It provides an easy framework for adding new algorithms to experiment with and has some basic functionality for interacting with audio files in both the time and frequency domains. At the moment, nussl only works with .wav files. Please see the issues page before contacting the authors.

Installation

We recommend getting installing anaconda prior to using this package.

Installation is easy if you have pip (recommended): ` $ pip install nussl `

–or–

Run setup.py like this: ` $ python setup.py install `

You can also download and copy the nussl folder into your working directory.

nussl requires:

numpy, version >= 1.8.0 scipy, version >= 0.13.0 matplotlib, version >= 1.3.1

Documentation

Full documentation is 

Note: This package has been tested with python 2.7, but not 3.*. Use 3.* at your own peril!

Status

KAM, RPCA (and corresponding demos) need to be migrated to the framework.

Licence

nussl 0.1.0 is under an Apache License 2.0 (http://choosealicense.com/licenses/apache-2.0/)

Contributors

Corey Grief (), Ethan Manilow (), Fatemah Pishdadian (),

Contact

Contact Ethan Manilow (ethanmanilow2015 [at] u.northwestern.edu) with any questions or issues. Please look at the “issues” page before reporting problems.