Stereo outs on the tempest are great for things like toms and certain cymbal or effects. I typically use voices 1kick 2snare 3 hats (all mono recorded), 4 toms (stereo) 5 fx (stereo), leaving voice 6 (unplugged)for basses. Allocating each voice to an output and using one of the mains to catch click and using voice 6 a bassline I'll replace with a voyager later, works best for me. 12 inputs sounds like a lot, but tis always better to have extra interface ins, than not enough. I use a qu-16 thats 22 inputs and its constantly full up, once i insert a tempest, an effects send/return pedal, and a couple mics for guitars. Sending audio as a send FROM the daw, uses channels. Still, my fav workflow is catching each voice into the daw, processing it as I want, sending a stereo out to my timefactor and catching it too. This enables me to play as I normally would, while capturing everything raw for post process, but still hearing the mix as I expect post to be. The allen and heath stuff may not be creme of the crop, but its a low noise floor, reliable pres and after I mangle each sound horribly with plugins, I'm betting the quality is less aparent. Fwiw, Ive yet to design a drum rack in ableton, or use a plugin that inspires me as much as the tempest. Not as easy as just using a push ITB, but WAY more gratifying. Even better once you get used to NOT using quantatize. My pearl kit hasnt such a feature and its epic too.