I tried to do exactly this when I had a RYTM, but i found it useless as the track volumes were saved per pattern, so every time you change patterns the current state of the physical faders did not match the next pattern causing mayhem for a live set.
Im not sure if the digitakt is different? does it have a master volume for each track that is not part of the pattern?
This is not so much an issue if you were using encoders i guess…