Well, if you do not want to Bounce per microphone, pre-mixer, you can only Bounce per Mixer output.
The logic is pretty simple; e.g. your standard mixer setup would be more than 1 microphone for Kick and Snare, right?
These mic faders, with perhaps EQ and Comp on each, would most likely be Bussed to a Sub group Bus with additional processing before being routed to the Master Fader and output. The Master Fader perhaps also has some kind of processing to add to the final sound. This Master Fader processing will react differently if there is a lot of instruments hitting it than if you solo e.g. the Snare.
So, when you Bounce and want your EQ, Filters, Compression, Saturation, etc. on an instrument, the most logical thing would be to Bounce at the end of the chain, right?
Bouncing per fader or even at the Sub group Bus would not yield the same audible result as Bouncing the Output in this scenario. Perhaps you are also Sending to an FX Bus, like a Reverb?
Then you either route this Reverb Bus to the Master or a separate output.
So, route all faders and/or Busses you wish to have as separate Audio files to their own Output in the SD3 Mixer before you Bounce.
I would also rename the Output Faders to what is Bussed there, e.g. Kick, Snare, etc. This will also make it easier to know what's what when you have Bounced, since the file name then will be something like 'Out_Kick', 'Out_Snare', etc.
I hope this clarifies,
John