[Equipe Audition] [CNRS] [LPP Paris 5] [ENS] [DEC]

The Audition library for Pure Data: a platform for real-time auditory modelling

possible processing flows

The Audition library is package of tools that make it easy to implement simple auditory models in Pure Data. Pure Data is a free, open source, real-time programming environment. If you've never heard of pd, start here and here.

The approach of the library is to provide building blocks that perform signal processing informed by what we know of the peripheral auditory system. These blocks are useful to estimate perceptual attributes such as loudness or roughness, build time-frequency auditory representations, do some processing in the auditory time-frequency domain and then resynthesize other sounds... At the core of the library for now is a real-time implementation of the gammatone filterbank, along the lines of [Hohmann 2002]. An overview is given in this ICMC2005 paper.


The library is free and open-source.

Pure Data Audition Library V0.9 for Windows. To install, simply unpack the archive and place the files under the 'pd' root directory (see Readme.txt).

Source codes in C++ (requires flext) and VC.net project.

If you haven't got it already, Pure Data.

More info

The introduction patch 0-Help-intro.pd (screenshot) presents all the objects in the library available to date. To get help within pd, right-click on any of the objects.

More modules will hopefully be added by us and others in the future. All feedback welcome, of course. In the meantime, the library can be used in conjunction with all the sound synthesis and analysis tools available in pd.

There are a lot of very good auditory modelling platforms around. See for instance DSAM, from the University of Essex, or AIM from the CNBH in Cambridge.

The gammatone implementation we chose: Hohmann, V. (2002) Frequency analysis and synthesis using a Gammatone filterbank. Acta Acustica united with Acustica. 88: 433-442.

The library was designed by Daniel Pressnitzer and Dan Gnansia. Implementation by Dan Gnansia, contact Dan.Gnansia at ens.fr

[Equipe Audition] [CNRS] [LPP Paris 5] [ENS] [DEC]