I have some audio recordings at 16KHz, which only contain music (BenSound Adventure: https://www.youtube.com/watch?v=0H8JTsG1Jtk). Since I know the music and have a separate wav file, I was wondering if I can remove the music from the original audio file. The original recording was a bit attenuated and mixed with other background noises. I am interested in retrieving those noises while removing the music completely. Is it possible to do that?

  • 1
    $\begingroup$ Sort of. Is there part where the music is playing but there is minimal other noise? The problem is that while you can try and "cancel out" the music, typically when it's played through speakers and recorded on a microphone, supposedly in a room, eliminating the material can be tricky. First you have to account for the frequency responses of the system playing the sound, the system recording the sound, and the "room response" of the environment. $\endgroup$
    – Keegs
    Oct 20, 2021 at 19:44
  • $\begingroup$ You've also got clock drift and things like that to deal with. If you cross-correlate the entire known signal with the entire recording, you should get a peak at the point where the music exists, and a sort of pseudo impulse response around it (since it's sort of a wideband signal) If it's a well-defined impulsive peak, it will be easier to remove $\endgroup$
    – endolith
    Oct 20, 2021 at 20:49

1 Answer 1


This is basically what adaptive filters do - identify a slowly varying impulse response convolved with one signal, added to another.

The success you wil have depends on stuff like clock drift, changes in acoustics and the relative power of those two signals.


Your Answer

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

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