clear all; close all; Fs=8000; [y,Fs] = audioread( 'C:\Program Files\MATLAB\MATLAB Production Server\R2015a\bin\work\data\1-11.wav'); y=y./(1.01*abs(max(y))); %normalizing the signal amplitude from-1.0 to 1.0 y=y(241:400); % select 20msec segment of the speech signal N=160; w=hamming(length(N)); y=y.*w; %multiplying the signal with hamming windoW P=10; %autocorrelation ycorr=xcorr(y,y); ycorr=ycorr./(abs(max(ycorr))); A=ycorr(1:P); %pth order autocorrelation sequence r=ycorr(2:(P+1)); %r=abs(r); figure; subplot(3,1,1); plot(y); title('voiced speech segment'); subplot(3,1,2); plot(ycorr); title('Autocorrelation of voiced speech'); A=toeplitz(A); %toeplitz autocorrelation matrix %A=abs(A); L=-inv(A)*r;%direct matrix solving method %LP coefficients disp('LP Coefficients are'); disp(L); subplot(3,1,3); plot(L); title('Linear prediction coefficients');