Skip to main content
7 events
when toggle format what by license comment
Mar 3, 2021 at 13:34 comment added Dan Boschen This is true if you are creating the signal. If you are analyzing the signal then the total duration is critical to the frequency resolution achieved (and even in that case replicating the signal will not improve the resolution (ability to see closely spaced signals) for the example I gave.
Mar 3, 2021 at 13:32 comment added Ariane Understood: the narrower spectra come from the longer windowing, but we do not need these smaller spectra: If we know that N is the period of the total signal, then we will be able to resolve the two tones that the signal consists of, as soon as we observe at least N samples!
Mar 3, 2021 at 13:11 history edited Dan Boschen CC BY-SA 4.0
added 67 characters in body
Mar 3, 2021 at 13:10 comment added Dan Boschen Here is another good example: Compare the DFT of a zero padded waveform to the cut and paste waveform (the zero padding provides more samples on the DTFT, and in the DTFT the frequency resolution of both cases is clearly visible).
Mar 3, 2021 at 13:06 comment added Dan Boschen So in other words for the example you would need to copy and paste 10 seconds to get the full number of periods, right? Then you are also saying there is nothing else within that higher frequency resolution for you to see--- that doesn't mean the resolution isn't there, it just means you already know the other spectrum that you would possibly see isn't! I am looking at it as if I received the waveform from someone else and didn't know if they copy and pasted or not.
Mar 3, 2021 at 13:04 comment added Ariane Thank you, I agree with what you are saying, but in that case we are not copy-pasting a full (or an integer number of) period(s) of the signal that we are interested in! :) I am specifically interested in that case.
Mar 3, 2021 at 12:57 history answered Dan Boschen CC BY-SA 4.0