For me, it depends on how I intend to record my output. When I was working just on the DT or with one thing through the inputs, I’d do all the mixing and compression on the DT itself and record it out to a Zoom R8. Easy.
Now I have a larger setup and I work in Ableton, and I use Overbridge to take the separate tracks from my DT and Analog Keys over USB as well as the inputs. This takes the audio before the effects, panning and the compressor. It also substantially reduces the volume and bypasses the internal mixer.
For the FX and panning, I do that on the fly now, because I monitor through Ableton at all times, and I use a Novation Launch Control XL to control my sends in Ableton in the same way I would use them on the DT itself.
For the ‘mixdown’ proper, I come back to the song after I’ve recorded everything and check all my levels, set up compression, do any adjustments to FX and panning that I feel are required, do some EQ if I feel like there is any sort of frequency issues. At this point I’ll listen critically at reasonable volume through my monitors and my headphones to check for anything that doesn’t feel right, and make adjustments if necessary.
I’m a novice at this, so hopefully some more experienced people comment. I keep it quite simple.