Talking about "windows" when it comes to analog filtering doesn't make a whole lot of sense, because the goal of a "window" is to make a (mathematically) infinite impulse response into a FIR, but analog filters (and recursive digital filters) implement IIR filters as-is... with the caveat that due to the requirement that we need a causal filter, we can't have linear phase.The shape that you see on a analog spectrum analyzer is determined by the Q factor of the local oscillator of the spectrum analyzer, the response of the VBW filter and the RBW setting. So I think the window filter should be modelled by these factors.
Yet, the thing is, whatever "window" you pick, you're not going to exactly match the analog results with an FFT. You could technically design a FIR against the analog response, but that's inconvenient if we need to sweep the thing.
That said, I really don't see a reason why you couldn't just use the analog construction in digital "as-is" as well. One needs to mind aliasing, but especially if we use a complex signal path, that should not be a huge deal.
Statistics: Posted by mystran — Wed Dec 11, 2024 4:35 pm