I had to film a kids Ballet recital and the video is fine, but the audio has that cheap "hollow" auditorium echo/reverb to it.
I have no problems ADDING echo/reverb... there are enough delay filters out there to sink a battle ship. But how does one REMOVE it. I can dress it up by increasing the highs and lows but it still sounds a bit cheap.
Duplicate your audio track and invert it. This should give you silence if you've got it right. Now apply a compressor to one of the audio tracks and adjust the knee, ratio, attack and decay times, probably best in that order.
This works because the compressor will leave part of the signal untouched causing partial cancellation from the two summed tracks. You want to cancel the quieter part of the tracks on the trailing edge of the sounds, that's where the echo is noticeable.
I've had this work quite well with speech from a very live large space. Hasn't worked so well for me with music, you might fare better.
I wish but not my idea, got the idea from someone who used to do it using reel to reel tape. I think you can also enhance the effect by adding a tiny amount of delay between the two tracks, which is probably where making duplicate tapes came in to what he was describing.
I should warn anyone trying this you need to do quite a bit of experimenting as each situation is different. From what I've achieved I never got perfection so it was a matter of deciding the point where I was doing too much harm to the wanted signal while removing more of the unwanted.
Basically, this trick simulates a "downward expander" instead of a hard cut-off noise gate. Under certain circumstances, it can work pretty well.
If you use the Track Compressor in Vegas, which works perfectly well for this task, turn off "Auto Gain Compensation", set "Amount" to "2.0" to start, adjust "Threshold" to taste. Set the attack time at or very close to zero in order to not cut off the first part of words and play with the release time to get rid of the reverb, maybe start at 100 to 250 ms. It's going to take some messing.
A few people have taken it a bit farther by refining the attack / release delays and playing with the EQ's selectively on the feedback track; it's more of an art than a science to get the de-effect right. As I caution every time it gets mentioned -- use it sparingly.
The acoustic mirror idea is something I tried a while back, and never got a verifiable result from. The reason is phase chasing, which acoustic mirror cannot do, afaik. The inverted image approach Bob and I described takes care of this without a whimper. It's actually the same theory on which early high-fidelity recordings and amplifiers relied.
In the old days of studio recording we had someone with coke-bottle thick glasses chasing the phase on a dual-trace oscilloscope with a big knob on an analog LC delay / phase shifter (whoops, was that a capstan slip that just crashed our remix?).
Well, some masterpieces like Dark Side Of The Moon, Jefferson Starship, Abbey Road, and Karn Evil Nine were the result of that early technology (no, unfortunately I was not involved in any of those projects).
No, blinky, it doesn't mean your not the brightest bulb on the tree, it just means you're not the oldest one . . .