But this can allow you to see signals buried in the scope time domain noise floor. They are also subject to aliasing since there's no... BW limiting. This can be a real nightmare and windowing do anything against this.
I agree that the FFT function on a scope is no match for a good spectrum analyzer. However, on a good scope with a comparatively high sampling rate, the FFT can be quite useful. It is the nature of the FFT that, provided certain assumptions are met, it can provide considerably more resolution and dynamic range than is implied by the number of bits. I believe that this subject was covered in an earlier thread. Most scopes do provide some BW limiting ranges, although the selection is very limited. Whether you use an SA or scope FFT, it's important to understand how windowing and sampling work, since otherwise you're likely to underestimate the energy in any given band. Paul Mathews
No. The FFT is always going to have a resolution determined by the sampling rate and the depth of the buffer. The left edge will always be DC. You could zoom in as much as you wanted at DC, but the most you can zoom in (the finest fft bin granularity) at 100MHz won't be nearly as good as a spectrum analyzer.