"Diffing" Channels

[Norm]

This could be a stupid question...

I have taped a show recently and it has some pretty heavy talking in mostly one channel.
I know the general conscenus is this cannot be reduced easily.
But, is there any tools for getting the difference between two channels, and some how reducing the difference in the offending channel? It sounds a bit like set theory!
I have been able to reduce the volume of the offending channel by around 50% and mx the other channel over it. This seems to remove audibly the talking, but i think it loses too much stereo seperation too.

Cheers

Norm

[Steve J]

I haven't tried this myself; but for those who have: what software are you using to edit the recording?

[bhtoque]

I'll do that type of channel swap with SF 7 to hide small flaws (especially works well with split omni's) but you will definately lose most of your stereo image when you do it across an entire show.

What I would suggest is using the good channel as the M and the noisy side as S and just decode to stereo that way. You'll get less noise and a better image than a blend of L & R.

mid/side mixing takes some practice though. Try it a few times and listen to the results to get a feel for how it works.

JAson

[Norm]

I'll do that type of channel swap with SF 7 to hide small flaws (especially works well with split omni's) but you will definately lose most of your stereo image when you do it across an entire show.

What I would suggest is using the good channel as the M and the noisy side as S and just decode to stereo that way. You'll get less noise and a better image than a blend of L & R.

mid/side mixing takes some practice though. Try it a few times and listen to the results to get a feel for how it works.

JAson

Thanks, could you just explain that a little better for my simple ears.
Cheers

Norm

[RebelRebel]

Mid/Side : two mics...a cardiod and a fig of 8.
First, aim the cardiod directly at the source.. as if you were recording in mono. Take that in to a track you call "mid" in your DAW.

Second, use the bottom element(fig 8   ), which is now 90 degrees off-axis from the source, and bring that in on a channel called "side".

To mix, duplicate the side track, and flip the polarity (phase). Lock the two side channels together and bring that into the mix to taste so that the stereo image begins to appear.

I'll do that type of channel swap with SF 7 to hide small flaws (especially works well with split omni's) but you will definately lose most of your stereo image when you do it across an entire show.

What I would suggest is using the good channel as the M and the noisy side as S and just decode to stereo that way. You'll get less noise and a better image than a blend of L & R.

mid/side mixing takes some practice though. Try it a few times and listen to the results to get a feel for how it works.

JAson

Thanks, could you just explain that a little better for my simple ears.
Cheers

Norm

[zowie]

Diamond Cut, and no doubt other programs, lets you subtract  L - R or R - L, and you could reverse phase and add it back in.  But I doubt what you're trying to do will accomplish what you want - because of time delays between mics (depending upon your setup), the difference between the channels will probably be pretty much everything.

[Norm]

Thanks lads going to look into that, even if its a meaningless excercise.
I just cannot bear to listen to the tape, the music seems average quality, or at least listenable. +Ts all round.