Earlier I asked a question here about performing FFT at lower frequencies but still at high sample frequencies. I was under the impression that the FFT was inherently calculated at every frequency 0->(Sampling Frequency)/2 distributed in bins of width Fs/(2*N). But, looking at some of the answers I received, it appears that it is possible that Fs is only useful in determining the maximum frequency, which is Fs/2, but that it is entirely possible to, say, sample at 640 Hz but to only sample frequencies 0-64 Hz.
Yet, from what I've read so far on FFT implementations, it appears that the FFT in the examples I'm seeing, is being performed on the entire range 0->Fs/2, so I'm not entirely sure how to go about performing the FFT on only a portion of the maximum frequency.
Now, just to pre-empt this, I realize that somewhere along the line there is a high chance of some sort of misunderstanding on my part. I'm not entirely sure what I'm missing, but that's why I laid out my current understanding here and maybe somebody can properly explain why I'm perceiving that the FFT is only being performed on a part of the range, which I have the suspicion, due to the contradiction outlined above, is just an incorrect interpretation by me. Or maybe oversampling requires some sort of different logic that I was unaware of. In either case, I'm left a little confused so any help is appreciated.