FFT

原文

Fourier transform

The Discrete Fourier Transform (FFT) is a method that takes a complex waveform and decomposes it into a set of component sine waves. There's a clear explanation of Matlab's FFT transform here: The FFT requires a dataset that is 2N elements long, i.e. 2, 4, 8, 16, 32, etc., so the transform uses the value of N that is closest to the signal length you have (NFFT). This is computed below for illustration, but not used.

%-----------------------------------------------------------------------
% do_FFT.m
%-----------------------------------------------------------------------
% assumes you have already created mysig.
L = length(mysig);
myFFT = fft(mysig,S);
myFFT=myFFT/L; %scale the output to 1
freq = S/2*linspace(0,1,S/2);%create a range with all frequency values
% Plot single-sided amplitude spectrum.
figure; 
stem(freq,abs(myFFT(1:length(freq))));
xlabel('Frequency (Hz)');
ylabel('Amplitude ');

If you look at the myFFT variable, you will see it has a real and imaginary part. These correspond to the amplitude and phase of the signal. We select just the real part by using the abs() command.