THE AUDITORY MODELING TOOLBOX
BISCHOF2023_FILTERBANK - filterbank used in model bischof2023
Usage
[sig_out,gt_out,fc] = bischof2023_filterbank(sig_in,N,len,fs,f_range)
Input parameters
| sig_in | Input time signal to be filtered |
| N | Filter order of the gammatone filters (default: 4) |
| len | Filter length in samples (default: 512) |
| fs | Sampling frequency in Hz(default: 44100) |
| f_range | (optional) Defines the upper and lower frequencies for the bark filterbank and chooses the right channel |
Output parameters
| sig_out | Matrix with filtered input signal. The matrix dimensions are: size(sig_out,1): length(signal) size(sig_out,2): numer of used bark bands(either defined by fs of f_range) size(sig_out,3): number of channels of sig_in |
| gt_out | Matrix with all gammatone filters |
| fc | Center frequencies of used bark filters |
Description
BISCHOF2023_FILTERBANK creates BARK-scaled gammatone filters and filters the input signal. Corner frequencies of Bark Scale according to Zwicker (1961).
References:
E. Zwicker. Subdivision of the audible frequency range into critical bands (frequenzgruppen). J. Acoust. Soc. Am., 33(2):248--248, 1961. [ DOI ]
T. Gunawan and E. Ambikairajah. Speech enhancement using temporal masking and fractional bark gammatone filters. In 10th International Conference on Speech Science & Technology, pages 420--425, Sydney, Australia, 2004.
Build with Bootstrap