I'm working on a project where the purpose is to enhance FFT peaks in a signal (musical instruments), by other words, reduce non-harmonic components and enhance the identified FFT peaks. So far I'm doing the following:

  1. I'm analysing the file on a frame basis and
  2. for each frame, I calculate an AR(8) model,
  3. then reduce the bandwidth for each peak and
  4. apply this as a filter to the frame,
  5. combine everything using PSOLA

The results are good but they could be better. Any ideas for a different approach?

  • 2
    $\begingroup$ wanna do it with a time-domain approach, rather than FFT? $\endgroup$ Jan 5, 2020 at 1:00
  • $\begingroup$ or use a frequency domain that is dedicatedly combining harmonics into a single number? $\endgroup$ Jan 6, 2020 at 15:01


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.