I'm trying to do some harmonic analysis on some wave files. I want to view each peak in a window only a few Hz wide and would like as high resolution as is practical. If I send the whole waveform into a regular broadband FFT (SVT FFT Spectrum (Mag-Phase)), the longer the file, the smaller df is, and the better resolution I get. For example, a 5 second sine wave give df = 0.2Hz and shows a triangular peak with a base of 0.8Hz centered at the right frequency. The smaller the wave file I use, the wider the triangular peak grows, which makes sense.
It seems like the SVT Zoom FFT would be more like what I'm trying to do, but first of all, it is much slower. Also, it looks very diff
erent. Instead of a wide peak I get lots of other peaks surrounding the main one, even with larger wave files. Ideally, the spectrum should be a single "infinitely thin" peak, so I don't know how to interpret this.
Why would someone choose to use a zoom FFT instead of the broadband FFT of a large input? What do they really do differently, and what are the extra peaks that I get in the zoom FFT? Thank you for any help.
Just in case, I'll attach some of the results I was describing.
