How can I calculate the energy of an audio file (audio signal) that I imported in Matlab?

Question

Aziz Alharbi 2016-10-28

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/309648-how-can-i-calculate-the-energy-of-an-audio-file-audio-signal-that-i-imported-in-matlab

编辑： Walter Roberson 2017-2-22

I'm trying to calculate the energy of an audio file in matlab. I wnat to calculate its energy before and after filtering. Here's my matlab script:

clear('all');
close('all');
[f,fs] = audioread('test.wav');
N = size(f,1); % Determine total number of samples in audio file
figure;
subplot(2,1,1);
plot(1:N, f(:,1));
title('Left Channel');
subplot(2,1,2);
plot(1:N, f(:,1));
title('Right Channel');
n = 2;
fprintf('The sum of the absoulte values (Before filtering) is: \n');
fff = fft(f);
h = sum(abs(fff));
max(f)
disp(h)
beginFreq = 700 / (fs/2);
endFreq = 1600 / (fs/2);
[b,a] = butter(n, [beginFreq, endFreq], 'bandpass');
fOut = filter(b, a, f);
% Construct audioplayer object and play
%p = audioplayer(fOut, fs);
%p.play;
fprintf('The sum of the absoulte values (After filtering) is: \n');
figure;
plot(filter(b, a, f))
ggg = fft(f);
h = sum(abs(ggg));
disp(h)
fprintf('Sampling Frequency is: \n');
disp(fs)

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Image Analyst 2016-10-28

Wouldn't you use pwelch()?

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Frank Macias-Escriva 2017-2-22

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/309648-how-can-i-calculate-the-energy-of-an-audio-file-audio-signal-that-i-imported-in-matlab#answer_255820

在 MATLAB Online 中打开

In your code you will start approaching to your solution just fixing the mistyping on the calculation of "ggg", I mean, almost at the end of your script you should:

ggg = fft(fOut);

instead of using "fft(f)". (in the last plot I would also use "plot(fOut)")

Anyway, you should check if you really want the energy of the discrete signal then you should look for:

sum(abs(fff).^2)/numel(f)

That Parseval's theorem states will be equal to "sum(f.^2)".

Consequently, for the filtered signal it will be:

sum(abs(ggg).^2)/numel(fOut)

again, it will be equal to sum(fOut.^2).

Hope this help.

Frank