I'm in the middle of mixing a tune ( in PT HD ) where the percs and drums were recorded in the same room. The Percs mics has a good dose of drum leakage wich i would like to keep down to a minimum...what are your suggestions ?? BTW, i thought of the SPL or Sony Transient Designer, unfortunately, i don't own either...
If it is very jazzy, or folk/ethnic sound, I'm afraid can't help much.
What I would do (if enough time and the case important), I would load the perc tracks to the wave editor and process one after the other. I would use any comp/expander that can be set up as a soft gate, every good editor has one.
I would monitor it in headphones, one channel is original, the other is gated. so I can check where a percussion hit would be cancelled in error by the common setup I choose. For the problems that I have found, I can set cue ranges and apply a softer gate locally or silence the drum sound manually. I would use the cue list to save all special handling ranges, then apply these individually, and then process the rest of the track with my standard gate.
It depends on the type of music. It might be possible to use the expander to shorten the perc sounds, and later in the mix get them longer, with an ambience or room plugin.
I would also use a highpass and a low pass to let only the bandwith come through, that the instrument/track actually needs.
It may be also possible to find typical frequencies of the drum set (is there "ringing"?) and block them with a notch filter.
Anyway, you will have a tradeoff between the sound of the drums and the sound of percussion.
of course, you would try the Transient Designer first, as the pro's below are telling, if you are lucky to have such a device or a similar one
sometimes difficult things can be done by hand, or mimicking a delicate plugin by combining some processes you already have in your gear. you might approach a transient designer device somewhat, if you chain a compressor and an expander, with very ideal characteristics (at the start switch off all tube effects and the like, no kee) and give them inverted curves so the output is the same as the input. then you begin changing time constants between the two devices, and see what happens..