This code also implements the roex auditory filters.
Please read the article located on the debian file system file:///usr/shar/doc/libaudiomask-1.0/Flax.2000.Improved.Auditory.Masking.Models.pdf
It may also be found on the web http://www.assta.org/sst/SST-00/cache/SST-00-Chapter9-p2.pdf
This example shows how to use Dr M.R. Flax's (2000) hybrid simultaneous audio masking class to find the masking threshold of a time domain signal.
The compilation of this file is demonstrated in Makefile. Run this file : ./AudioMaskerExample View the results of this file using www.octave.org by running the script view.m
The input audio should be stored in the file INPUTFILENAME in text format - each sample seperated by a white space.
========================= HOWTO ===============================
// First find the masking threshold AudioMasker masker(sampleFreq, count); // Create the audio masker class using fs=sampleFreq and count filters masker.excite(input, sampleCount); // find the mask for the array of input data which has sampleCount time samples. // Now do something with the masking threshold ... // The frequency domain mask is now located here for (int j=0; j<count;j++) masker.mask[j]; // This is the mask at each of the count frequencies of interest // A more sophisticated example - find the threshold for each Fourier bin double fact=(double)sampleFreq/((double)sampleCount-1.0); // convert from an index to the equivalent * Fourier bin frequency for (int j=0; j<halfSampleCount;j++){ cout<<"finding for freq "<<j*fact<<'\t'; // The frequency we are inspecting double threshold=masker.findThreshold(j*fact); // The masking threshold 20*log10(threshold); // The threshold in decibels (dB) }