How can I plot the frequency content of my sample?

Im trying to find the frequecny content of an audio signal and plot it. So far I have :

clear all %Clears the Workspace
close all %Closes all figue/plot windows opened by MATLAB
clc %Clears the command window

[Sample, Fs] = audioread('Snare_1.wav');

Sample = Sample(:,1);

period = 1/Fs;

t = 0:period:(length(Sample)*period)-period;

figure(1) 
plot(t,Sample);
grid
xlabel('Seconds (s)');
ylabel('Amplitude');
xlim([0 0.8])
ylim([-0.1 0.1])

But I just cant figure out how to analyse the frequency content, please help if you can

https://we.tl/t-LlzN3QxitD : This is the snare file

[Adam]

doc fft

is a good place to start

[Tiago Dias]

Hi, could you please attach the Snare_1.wav file?

[Jules Marrison]

https://we.tl/t-LlzN3QxitD This is the snare file

Answer 1

[Meg Noah]

Did you find a solution yet? Here's one. I wasn't sure if you wanted the frequencies identified or just plotted.

clear all %Clears the Workspace
close all %Closes all figue/plot windows opened by MATLAB
clc %Clears the command window
[X, Fs] = audioread('Snare_1.wav');
X = X(:,1);
N = numel(X);
T  = 1/Fs;      % [s] sampling period       
t  = (0:N-1)*T; % [s] Time vector
deltaF = Fs/N;  % [1/min]) frequency intervalue of discrete signal


figure(1) 
subplot(2,1,1);
plot(t,X);
grid
xlabel('Seconds (s)');
ylabel('Amplitude');
xlim([0 0.8])
ylim([-0.1 0.1])
title('Amplitude of Wave');

% compute the fast fourier transform
Y = fft(X);

% manually shifting the FFT
Y = abs(Y/N);
Amp = [Y(ceil(end/2)+1:end)' Y(1) Y(2:ceil(end/2))']';
if (mod(N,2) == 0)
sampleIndex = -N/2:1:N/2-1; %raw index for FFT plot
else
sampleIndex = -(N-1)/2:1:(N-1)/2; %raw index for FFT plot
end
AmpSquared = Amp.^2;

subplot(2,1,2);
plot(deltaF*sampleIndex, AmpSquared);
hold on;
idx = find(AmpSquared > 15);
idx(sampleIndex(idx) < 0) = [];
plot(deltaF*sampleIndex(idx), AmpSquared(idx), '+');
xlabel('Frequency [Hz]');
ylabel('|FFT(Wave)|^2');
title('Power of FFT of Wave (Audible Range = 20-20\times10^4 Hz)');
xlim([20 2000]);

[Jules Marrison]

This is extremly useful, thank you very much!

I do have one question, is the Y axis on your graph the equivilent to a dB scale or just a magnitude scale?

[Meg Noah]

On dB scale they usually do a 10*log10(abs(FFTValues)^2) or sometimes just 20*log10(abs(FFTValues))

[Jules Marrison]

So how can i impliment this onto my Y axis?

Sorry to pester im new to matlab

[Meg Noah]

It's a little complicated, because dB should have some sort of reference value to make them meaningful. I need to think about it a little more about whether or not .wav files have an implied reference that equates dB to energy in Watts or volume. There are some more complicated bandpass (frequency range) considerations for audio, because hearing ranges are limited.

Technically, this is 'dB' but audio community may have additional caveats. I'll think about it some more.

clear all %Clears the Workspace
close all %Closes all figue/plot windows opened by MATLAB
clc %Clears the command window
[X, Fs] = audioread('Snare_1.wav');
X = X(:,1);
N = numel(X);
T  = 1/Fs;      % [s] sampling period       
t  = (0:N-1)*T; % [s] Time vector
deltaF = Fs/N;  % [1/min]) frequency intervalue of discrete signal
figure(1) 
subplot(2,1,1);
plot(t,X);
grid
xlabel('Seconds (s)');
ylabel('Amplitude');
xlim([0 0.8])
ylim([-0.1 0.1])
title('Amplitude of Wave');
% compute the fast fourier transform
Y = fft(X);
% manually shifting the FFT
Y = abs(Y/N);
Amp = [Y(ceil(end/2)+1:end)' Y(1) Y(2:ceil(end/2))']';
if (mod(N,2) == 0)
sampleIndex = -N/2:1:N/2-1; %raw index for FFT plot
else
sampleIndex = -(N-1)/2:1:(N-1)/2; %raw index for FFT plot
end

subplot(2,1,2);
plot(deltaF*sampleIndex, 20*log10(Amp));
hold on;
idx = find(Amp.^2 > 15);
idx(sampleIndex(idx) < 0) = [];
plot(deltaF*sampleIndex(idx), 20*log10(Amp(idx)), '+');
xlabel('Frequency [Hz]');
ylabel('dB');
title('Power of FFT of Wave (Audible Range = 20-20\times10^4 Hz)');
xlim([0 2e4]);

It's basically some sort of noise, not pure tone content for this drum.