Code covered by the BSD License
 celp16k(x,N,L,M,c,cb,Pidx)celp > 16000 bps CELP analyzer and synthesizer.
 celp9600(x,N,L,M,c,cb,Pidx)celp > 9600 bps CELP analyzer and synthesizer.
 celpana(x,L,M,c,cb,Pidx,b...
 celpexcit(x,cb,Pidx,ar,ac...celpexcit > CELP excitation sequence.
 celpsyn(cb,kappa,k,theta0...
 durbin(r,M)durbin > LevinsonDurbin Recursion.
 lpcana(x,M)lpcana > Linear prediction analysis.
 lpcrespitch(ehat,th,minla...lpcrespitch > Pitch estimation from prediction error sequence.
 lpcweight(ar,c)lpcweight > LPC based perceptual weighting filter.
 rf2lpc(kappa)rf2lpc > Convert reflection coefficients to prediction polynomial.
 CELP_RUN.m

View all files
CELP codec
by
Sourav Mondal
13 Nov 2012
This is a code to demonstrate CELP codecs of bitrate 9.6kbps and 16kbps.

lpcweight(ar,c) 
function ac = lpcweight(ar,c)
% lpcweight > LPC based perceptual weighting filter.
%
% ac = lpcweight(ar,c)
%
% The function takes the LP coefficients, ar = [1 a(1) ... a(M)],
% and the parameter, c, as inputs, and returns the coefficients of
% the filter function A(z/c) in the vector ac.
% Linear predictor order.
M = length(ar);
% The i'th coefficient of A(z/c) is given by ar(i)*c^(i1).
ac = ar;
ci = c;
for (i=2:M)
ac(i) = ar(i)*ci;
ci = ci*c;
end


Contact us