You are right…apologies
We work with live recorded drums, sometimes with whole bands, sometimes just tracked by themselves
The main reason Im looking to use SL is because (genre dependent) for drums where we are using a main “kit mic”, it is for very open captures where that kit mic presents the instrument as a whole and the close mics reinforce (not always). It also means that best practice is to have cymbal mics but in 40 years…I have never heard this work so well
So
The idea is to mic the same
1. MS stereo kit mic (AEA r88 into transformerless tube pre) for whole picture and cymbal control
2. Close mics (D112) on kick, toms (atm450 + nt6) etc
Scenario A requires the kit mic to be unmixed and combined with the close mics using gonio alignment. Unmixing this track is the most challenging as it requires unbled tracks as the source elements
Unmixing leaves a lot of artifacts and bleed is mixed, often even cymbal attacks contain the snare attacks and it requires so much edit work I just go back to old school processes
Scenario B is where its jazz etc and we use the 1 kit mic for the whole capture as the mic is capable of serious low end…this should just be an unmix but its like what we are capturing is not really matching the model. It does often contain alternate snares etc which can be tuned very high…eg Pearl Firecracker 12” often between C/D/E so around 350hz although the shell still puts out a 250hz. Any close mic integration is really noticeable because phase is obvious in the open space and no matter how you do it…it can be heard ie phase alignment is only ever a best compromise
Scenario C is where we are doing fairly isolated stuff eg metal/goth where the close mics dominate but the kit mic/rooms are really just about cymbals and once again, the unmix really does do well. The close mics are even spectrally sidechained into the rooms but bleed in this case is really bad.
Anyway, The very essence of the debleed seemed quite obvious, grab a source and select the other tracks etc but its just not great because of the artifacts and fluffing around. Then I find out that there are other factors…like FFT windows etc
Im a mixer and producer…SL sounds like a better job than I am currently doing but it would help if there were pointed tutorials for the more complex aspects. I was originally inspired by one of the drum producers raving about it, then I see Andrew Scheps talking about it…but yeah…maybe Im drowning in siltation from a long time of doing it the hard way but sitting in front of a screen is no longer any joy…thats why I have gone to hybrid…its gotta be about music, the spontaneous joy of people and capturing, curating that but also always looking for ways to improve.
I thought the instrument capture now available might help so I tried doing 10 secs of dry hits (each element) as a reference for that session but it seemed to have made no difference for drums and there is practically nothing else in the manual to augment that?
Training your own models sounds like a great feature and solution
HTH