Publisher review:GammaTone Tool Kit takes input sound signal and transforms into multichannel basilar membrane motion output. All filter coefficient derivation and filter application is the work of Stanley (1993) and can be found in the freely available document ?An Efficient Implementation of the Patterson-Holdsworth Auditory Filter Bank? included in this .zip as ?PattersonsEar.pdf?. The source files uploaded are just modified versions of those found in the appendix of this document. All other references to related auditory material can be found in the acknowledgments of that document.IntroductionThis is a small collection of .m files which can be used to implement the Patterson-Holdwsworth auditory filterbank in order to observe corresponding basilar membrane motion (BMM) to an input signal across a number of channels, which is a commonly used pre-processing stage in auditory models. Until now I have had to invoke 3rd party software from MATLAB to accomplish this relatively simple task which is time consuming. The files included to make life easier are described as follows.GammaToneMake.m:The user inputs sampling frequency, number of channels required and the range of frequencies to be covered and this function generates filter poles and zeros for each auditory filter accordingly. It also gives useful information such as centre frequency and equivalent rectangular bandwidth (ERB) of each filter. This function has been modified slightly from Stanley?s original with the inclusion of being able to switch the spacing method more easily between Moore/Greenwood etc. A highest frequency input variable was added to give more control (rather than just fs/2 to low frequency).GammaToneApply.m:Very simple function, nearly identical to Stanley?s original which just takes an input array and filters it with the filter weights generated in GammaToneMake.GammaTone.m:This is just the above two functions combined so that the filters are generated and applied to the signal in one step. This is generally the easiest way to apply the filters in ?one shot? situations. If you are planning on batch processing many signals with the same set of filters then I recommend using the above separate functions as the filters only have to be generated once which in turn saves processing time.BMMplot.m:This is a simple plotting routine which I put together to mimic the BMM plot types seen frequently in ?Journal Of The Acoustical Society Of America? (JASA) articles among others, as well as software such as AIM and AMS. This allows data from each channel to be viewed as stacked line graphs.DemoGT.m:Contains sample script to show how everything works together. Put this in the working directory and run from the command line or editor. Requirements: ยท MATLAB Release: R14SP1
GammaTone Tool Kit 1.0 is a Matlab script for Signal Processing scripts design by Nick Clark.
It runs on following operating system: Windows / Linux / Mac OS / BSD / Solaris.
Operating system:Windows / Linux / Mac OS / BSD / Solaris
GammaTone Tool Kit 1.0 script details 
