I have three recordings of the same live gig/concert/performance. One excellent and almost crowd free recording of the left speaker. One audio track from a video taken mostly on the right side of stage and one other crowd filled recording around the center of stage.

I want to somehow merge these three recordings so that only the audio that is in all recordings remains and all the other audio is dropped.

I'd also like to experiment mixing in such a way that the excellent crowd recording is most prominant but mix in the right speaker recording for a stereo feel and add some center recording, again for stereo and a bit of crowd feel.

Each recording has crowd noise (people talking etc.) that I don't want, but there are different people around each microphone so in theory this merge should drop all sets of voices and leave only the cheers. Is this possible?

I could just set up the recordings, synch them up and the resulting volume boost where the audio matches up would mask a lot of non-matching audio but I'm worried about introducing too much extra crowd.

Any help? Articles, software to point me at?

I also have another gig where we have a soundboard recording which is mostly vocals and bass with a little drums and guitar and a good quality audience recording. I wouldn't mind mixing the two to up the quality and clarity of the audience recording.

(the recordings are of of my band if you're wondering)