Did first, simple test - it works so far.
- played white noise sample and recorded it from my position with t-bone MM-1 microphone
- used EQuivocate with sidechain to match the original file
- eq "correction match curve" generated.
I must go right now, back in some time with some pics.