Plotting a variable in a for loop

I am struggling to plot the error in the second for loop.

X = audiorecorder(8000,16,1);

disp('Start speaking.');
recordblocking(X,5);
disp('End of Recording.');

M = 10; % M: Order of LPC model
seg = 30; % Segment or frame size is 240

if (size(x,2) == 1) % Check if x is a column vector
    x = x'; % If x is a column vector, convert it to a row vector
end
npts = length(x); % Find the number of samples
nseg = floor(npts/seg); % nseg: Number of segments in the signal

for each = 1:nseg
    xx = x((each - 1)*seg + [1:seg]); % pick 240 samples for each segment
    [a,G] = lpc(xx, M); % Compute lpc and variance
    e = filter(a,1,xx); % Compute residual error signal
    G = sqrt(G); % Calculate the standard deviation (gain) of the segment
    e = e/G; % Normalize the error signal

    parameter(each,:) = a;
    gain(each) = G;
    error((each - 1)*seg + [1:seg]) = e;
end

Answer 1

[Voss]

I only see one for loop.

X is your audiorecorder object, and x is apparently the recorded audio data, but the relation between X and x is not given in the code, so here I'll create random data x, process it using your code and plot error:

% X = audiorecorder(8000,16,1);
% 
% disp('Start speaking.');
% recordblocking(X,5);
% disp('End of Recording.');

x = randn(40000,1); % random data

M = 10; % M: Order of LPC model
seg = 30; % Segment or frame size is 240

if (size(x,2) == 1) % Check if x is a column vector
    x = x'; % If x is a column vector, convert it to a row vector
end
npts = length(x); % Find the number of samples
nseg = floor(npts/seg); % nseg: Number of segments in the signal

for each = 1:nseg
    xx = x((each - 1)*seg + [1:seg]); % pick 240 samples for each segment
    [a,G] = lpc(xx, M); % Compute lpc and variance
    e = filter(a,1,xx); % Compute residual error signal
    G = sqrt(G); % Calculate the standard deviation (gain) of the segment
    e = e/G; % Normalize the error signal

    parameter(each,:) = a;
    gain(each) = G;
    error((each - 1)*seg + [1:seg]) = e;
end

plot(error)

[Daniela Trevino]

Thank you!

[Voss]

You're welcome! If anything is not clear, let me know. Otherwise, if that solves the problem, please mark my answer as Accepted by clicking 'Accept this Answer'. Thanks!