THE AUDITORY MODELING TOOLBOX

Applies to version: 1.1.0

View the help

Go to function

hohmann2002_filter - Construct a Gammatone filter within the HOHMANN2002 framework

Program code:

function filter = hohmann2002_filter(arg1,arg2,arg3,arg4,arg5)
%hohmann2002_filter   Construct a Gammatone filter within the HOHMANN2002 framework
%   Usage:  hohmann2002_filter(a_tilde,gamma_order);
%           hohmann2002_filter(fs, fc, gamma_order);
%           hohmann2002_filter(fs, fc, gamma_order, bandwidth_factor);
%           hohmann2002_filter(fs, fc, bw, attenuation_db, gamma_order);
%
%   Input parameters:
%      a_tilde             : Complex-valued filter coefficients
%      fs                  : Sampling rate (in Hz)
%      fc                  : Centre frequency (in Hz)
%      bw                  : Filter bandwidth (in Hz)
%      gamma_order         : Gammatone filter order
%      bandwidth_factor    : Bandwidth factor with respect to 1 ERB. Default: 1.0 
% 
%   HOHMANN2002_FILTER(a_tilde, gamma_order) specifies the complex-valued
%   filter coefficients directly as at the beginning of Sec. 2.3 in Hohmann
%   (2002). 
%
%   HOHMANN2002_FILTER(fs, fc, gamma_order) computes
%   filter coefficients for sampling rate fs, center frequency fc, and order of
%   the gammatone filter gamma_order. The filter bandwidth will be 1 ERB.  
%   Filter coefficients are computed according to Eq. 13 and 14 from Hohmann (2002).
%
%   HOHMANN2002_FILTER(fs, fc, gamma_order, bandwidth_factor) uses filter 
%   bandwidth of bandwidth_factor ERB (instead of 1 ERB).
%
%   HOHMANN2002_FILTER(fs, fc, bw, attenuation_db, gamma_order) designs
%   the filter such that for the bandwidth bw, the attenuation attenuation_db 
%   at the corner frequencies of the filter is reached. Filter coefficient 
%   are computed as in Eq. 11 and 12 from Hohmann (2002)
%
%   References:
%     V. Hohmann. Frequency analysis and synthesis using a gammatone
%     filterbank. Acta Acustica united with Acoustica, 88(3):433--442, 2002.
%     
%
%   Url: http://amtoolbox.org/amt-1.1.0/doc/modelstages/hohmann2002_filter.php

% Copyright (C) 2009-2021 Piotr Majdak, Clara Hollomey, and the AMT team.
% This file is part of Auditory Modeling Toolbox (AMT) version 1.1.0
%
% This program is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% This program is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with this program.  If not, see <http://www.gnu.org/licenses/>.

% AUTHOR: Universitaet Oldenburg, tp 
% Adapted to AMT (PM, Jan 2016) from function gfb_filter_new


filter.type = 'gfb_Filter';
switch nargin
  case 2 % a_tilde, gamma_order
    filter.coefficient = arg1;
    filter.gamma_order = arg2;
    
  case 3 % fs, fc, gamma_order
    fs = arg1;
    fc = arg2;
    filter.gamma_order  = arg3;
    filter.bandwidth_factor = 1.0;
    filter.L=24.7;  % Eq. 17 in [Hohmann 2002]
    filter.Q=9.265; % Eq. 17 in [Hohmann 2002]
    
    audiological_erb = (filter.L + fc / filter.Q) * filter.bandwidth_factor; % Eq. 13 in [Hohmann 2002]    
    go = filter.gamma_order;
    a_gamma = (pi * factorial(2*go - 2) * 2^-(2*go - 2) / factorial(go - 1)^2); % Eq. 14, line 3 [Hohmann 2002]
    b = audiological_erb / a_gamma; % Eq. 14, line 2 [Hohmann 2002]
    lambda = exp(-2 * pi * b / fs); % Eq. 14, line 1 [Hohmann 2002]
    beta = 2 * pi * fc / fs; % Eq. 10 [Hohmann 2002]
    filter.coefficient = lambda * exp(1i * beta); % Eq 1, line 2 [Hohmann 2002]
    
  case 4   % fs, fc, gamma_order, bandwidth_factor
    fs = arg1;
    fc = arg2;
    filter.gamma_order  = arg3;
    filter.bandwidth_factor  = arg4;
    filter.L=24.7;  % see equation (17) in [Hohmann 2002]
    filter.Q=9.265; % see equation (17) in [Hohmann 2002]
    
    audiological_erb = (filter.L + fc / filter.Q) * filter.bandwidth_factor; % Eq. 13 in [Hohmann 2002]
    go = filter.gamma_order;
    a_gamma = (pi * factorial(2*go-2) * 2^-(2*go-2) / factorial(go-1)^2); % Eq. 14, line 3 [Hohmann 2002]
    b = audiological_erb / a_gamma; % Eq. 14, line 2 [Hohmann 2002]
    lambda = exp(-2 * pi * b / fs); % Eq. 14, line 1 [Hohmann 2002]
    beta = 2 * pi * fc / fs;     % Eq. 10 [Hohmann 2002]
    filter.coefficient = lambda * exp(1i * beta); % Eq. 1, line 2 [Hohmann 2002]
    
  case 5   % fs, fc, bw, attenuation_db, gamma_order
    fs = arg1;
    fc = arg2;
    bw = arg3;
    attenuation_db = arg4;
    filter.gamma_order = arg5;

    phi =  pi * bw / fs; % Eq. 12, line 4 [Hohmann 2002]    
    u = -attenuation_db/filter.gamma_order; % Eq. 12, line 3 [Hohmann 2002]
    p =  (-2 + 2 * 10^(u/10) * cos(phi)) / (1 - 10^(u/10)); % Eq. 12, line 2 [Hohmann 2002]
    lambda = -p/2 - sqrt(p*p/4 - 1); % Eq. 12, line 1 [Hohmann 2002]
    beta   =  2*pi*fc/fs; % Eq. 10, [Hohmann 2002]
    filter.coefficient   = lambda * exp(1i*beta); % Eq. 1, line 2 [Hohmann 2002]
    
  otherwise
    error ('Number of required arguments: 2, 3, 4 or 5');
end

% normalization factor from section 2.2 (text) [Hohmann 2002]:
filter.normalization_factor = 2 * (1-abs(filter.coefficient)) ^ filter.gamma_order;

filter.state = zeros(1, filter.gamma_order);