I am working in Matlab, I want to make a little soft to post-processing information that I get using my picoscope 3424. Until now, I have used picoscope 6, but I want to make some jobs automatically in the future.
In this time I am "Calibrating" my new software, by comparing measurement previously processed with picoscope 6 with the same process made in Matlab and I have some problems with the FFT amplitude.
1) Does someone knows what equations Picoscope use to calculate FFT? Looking some books, the spectrum's amplitude is N/2 times the peak amplitud of the time signal, with N = FFT size. However, if I make this calculation, my final value in matlab is different from that of the picoscope.
2)What scale factor could have the spectrum value?
3) What does Picoscope do if the fft size is different from the total samples?
If the fft size is shorter than the total samples, does it use only the first N samples or does it zero padding to complete a integer number of fft sizes ? (e.g. if I have 2000 samples and I use a fft size of 1024, does picoscope use the first 1024 samples or does it add 48 zeros and perform two fft of 1024 samples and average them (2000+48 = 2*1024) ?
if the fft size is longer than the total samples, does it add zeros to complete the fft size or simply reduce the fft size?
Thanks in advice! Any information is usefull!