FFT Overlaps for feature extraction

Question

Fft question: Do I need overlaps if I’m only grabbing some features like mel or bark? I don’t even necessarily need every frame of a STFT even one every second or so works. How I understand fft overlapping is that each overlap sums together after the final window after the inverse fft. However, for my needs I do not even do an inverse fft and I only extract mel and similar features in the frequency domain. So do I need to do overlapping for my use case? It doesn’t seem like the feature values i extract even consider the overlap.

Unless I’m supposed to combine overlapping fft values for the feature extraction itself??? Actually this makes me wonder if I should be doing that when extracting feature because some info is lost from the windowing to prevent spectral leaking. If that’s the case then do I just sum the window overlaps together after the fft and before feature extraction??

Because I don’t do that for spectral audio processing (again, i only sum after the final window after the ifft..) Thanks!

Answer 1

The most common way of doing an STFT is probably using a 50% overlap. The reason for this is quite simple: this roughly gives "equal" weight to all time domain samples and you are not missing anything important. Let's consider a signal that consists mainly out of a short large impulse: if you don't overlap you will get VERY different results depending how the impulse lines up with the window boundary.

If you signal is reasonably steady state and/or if the features you want to extract are not particularly sensitive to impulsive signals you can indeed do with less frames. However that depends a LOT on the property of your specific signals and the requirements of your application.

FWIW: Personally, I'm not a fan of not overlapping. Most "normal" audio signals do have prominent features that can be quite short and the incremental work of processing all samples using a 50% overlap seems minimal (for most applications).