extract common part of two audios

[mathmax]

The problem is that although we may be able to "hear" what the best parts are, human hearing is incredibly sophisticated, and trying to emulate that process in software is extremely complex. If that were not enough, the processing must not change the voice part at all, or it will not cancel out the voice in the mix track, so that rules out using spectral processing tools.

ok.. so I'll just go with the untouched sample and use the denoise tool after deleting tags. Which settings do you use to remove the noise?

A much easier option is to find a second copy of the song and simply edit the two songs together, missing out the parts that have the unwanted vocal. Or easier still, let the artist benefit a little and buy the CD.

You know, the problem is that this track is unreleased and leaked this way with the tags on it. The artist (Michael Jackson) is dead and I'm afraid he'll not benefit anything now... Sony released an album last month, they could have put the song on it but they preferred to add some fakes sung by an impersonator...

[steve]

Which settings do you use to remove the noise?

I made a guess as to which voice sample you used, then made a noise sample from that, discarding (deleting) the voice.
I then amplified the noise sample by about 8 dB, and used that to create the noise profile.

I then made a duplicate of the section that I wanted to de-noise, and applied Noise Removal to the duplicate.
This made the copy quite a bit quieter than the original, so I use "Fast Lookahead Limiter" to amplify the duplicate whilst preventing clipping.
This was then a "patch" that I used to replace the original noisy section, and cross-faded from the original, to the patch and back to the original (using the fade-in, fade-out effects, and "Silence Audio" from the Edit menu).