For spectroscopic analysis is it better to use a fft, a spectrogram (short Fourier transform), or a discrete Fourier transform?

I want to use SWIR spectroscopy (940-2500nm) to determine elemental abundance in rock samples. Which is the best function to use when trying to take a complex sine wave and break it down into its individual components?

Answer 1

[Paul]

The FFT, implemented in Matlab by the fft() function, is just an algorithm for computing the Discrete Fourier Transform (DFT) of a finite duration, discrete-time signal. So there isn't a choice between FFT and DFT. As I understand it, the spectrogram is used if the frequency content of the signal changes with the the independent variable of the signal (typically time). So, if your SWIR signal fits that description, then perhaps the spectrogram is the right way to go.

doc spectrogram

Or try both the DFT and spectrogram and see what you get.