FFT function from raw sound pressure data

Hi,

I have some raw time vs sound pressure data that I would like to perform an FFT on and display as a frequency spectrum graph. I am struggling with how to get the correct frequency vector.

My data appears as: time = [0 1 2 3 4...] with corresponding sound pressure levels [5 34 23...]

Any help would be greatly appreciated.

Thanks for your time, Sam.

[Youssef Khmou]

compare the code below with this command :

 >>figure, psd(p) %  with p sound pressure

Answer 1

[Youssef Khmou]

hi Sam, try this code .

 % Given you vector p and time  .
 Fs=1; % sampling frequency.
 L=length(p);
 N=ceil(log2(L));
 fp=fft(p,2^N)/(L/2);
 Power=fp.*conj(fp);
 f=(Fs/2^N)*(0:2^(N-1)-1);
  plot(f,Power(1:2^(N-1)),'r'),  xlabel('  Frequency (Hz)'), ylabel(' Magnitude  (w)'),
 title('  Power Spectral Density'), grid on;

[Sam]

Thanks a lot mate. Perfect.

[Sam]

Oh just one thing. What is the units of the y axis? Should the w be replaced by dB or some other unit?

[Youssef Khmou]

in the first case the magnitude is simply in volts, or depends on what unis you are using , Pascal , Bar...but to plot in dB then :

 semilogy(f,Power(1:2^(N-1)),'r'),  xlabel('  Frequency (Hz)'), ylabel(' Magnitude  (dB)'),
title('  Power Spectral Density, logarithmic scale '), grid on;

[Sam]

Great cheers.