Doing the final voice work today. Here's a clip with quite a bit going on:
The mid tom sound has every parameter randomized (noise level, click level, tone level, pitch EG amt, tuning, pan, and saturation). It's also being triggered by random notes, then sent to Dr. Device, which is set to the "Stereo Dubs (insert)" patch, kinetics on, and the grey knubbie driving filter freq and res.
The hi-hat, which has its decay time randomly sequenced, is getting 1/32 ghost notes via the random note generator, in addition to the 1/16th notes that are programmed in the sequencer, then sent out one of the mono outputs. The kick is unaltered in any way; it is just sent out another mono output.
I can now officially say that the voice programming is done. We've only got a couple little things left on the Windows side, then we can move to OSX and "optimization." It's a bit heavy on the CPU, but that's to be expected with roughly 50 sequencers running at once, plus a whole mountain of oscillators, envelopes, and filters.
It'll come down quite a bit as we move in to the final stages, of course. It's certainly usable now. Not on a Netbook or anything, though. ;-)
It would need 54 controllers just for the knobs in the drum synths. It would need a total of 83 controls if you wanted to get all the levels, pans, and the compressor. This doesn't take in to account the several hundred buttons, or the 2052 distinct controls that make up the modulation sequencing section.
In short, a dedicated controller for Tattoo would look about like the bridge of the Enterprise. And I'm not talking about no mamby-pamby Series One Enterprise either. I mean the whole hog 9th movie Enterprise.