I have seen in a few papers, for instance in this one, the following method :

  1. Find an initial "frequency of interest" (FOI)
  2. Filter the input signal with a Notch filter around the FOI
  3. Subtract the result from the initial signal
  4. The result is the "interesting part" of the signal, with mainly frequencies around the FOI

Why would one do that rather than applying directly a bandpass filter of similar order and bandwidth ?


The original use appears to be in this paper, which is referenced in your linked paper.

The part that's missing in your explanation is explained in that paper:

SCreen capture from linked paper.

Specifically, the notch-filtered signal is not removed from the original signal, but from another processing of it: the ANC-de-noised version of it.

Otherwise, you are correct, you might just as well produce the filtering directly.

  • $\begingroup$ I get it for you article you're quoting, thanks. In mine, they say "The PPG signal is then filtered by this notch filter, and the result is subtracted from the PPG signal.", so it looks like they subtract it from the same signal they used a notch filter on, so it's equivalent to using a bandpass, right ? $\endgroup$
    – gdelab
    Nov 3 '21 at 16:29
  • 1
    $\begingroup$ @gdelab Yes, if that's what they're doing, it's different from my reference and, in effect, is the same as what you say. $\endgroup$
    – Peter K.
    Nov 3 '21 at 17:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.