Would this process work?

[egxdomain]

I'm aware of vocal removal by cancelling left and right to keep center.

But I was wondering if the reverse is done (ie. retain vocals, remove background music), perhaps the mono result can be subtracted from each channel, separately, perhaps cleaning the mono vocal version by suppressing any instruments that were originally in the center, then doing the subtraction. If this is done independently, instead on the original stereo file, the stereo effect will remain.

Another theory I had was if somehow the original stereo image is expanded, (I've noticed the vocals are supressed a little in most cases), the previous process would yield better results! Is this possible?

I DO have a question, though. Is there a way to capture de reverb effect from both channels, perhaps by canceling the resulting channels if the previous did work? I believe if this was possible, a BETTER vocal removal would be possible, since you wind up with the center vocals and any reverb added to them.

Thanks, in advance, for your advice.

P.S. I heard somewhere about CSP vocal removal EVEN with a mono source. Is this just bunch of hype?