I too started down that Drum Map path, but found it too limiting.
My fantasy solution, and I am getting closer, is the ability to physically trigger articulation changes with key switches, but transform them to CC messages on the way in and send that CC message to Kontakt or Vienna Instrument or whatever, and record it as such.
Two primary reasons:
- CC messages can be chased, as just mentioned above, which is a big benefit for me.
- CC messages do not appear in the keyboard range of the PRV, nor do they clutter up notation (I use Finale for my notation needs.)
As an added benefit, while CC switches can be triggered from an external device with enough buttons (the Kore 2 has 8 buttons, which is proving to be insufficient in some cases), it is easier for me to use the keyboard keys, for now.
There are literally dozens of Kontakt scripts that provide conversion between CC and MIDI Note Number, but that limits one to Kontakt. Transmidifier is a stand-alone tool that can do all sorts of filtering and conversion, I've tried to get it to work as a single solution, but I'm not quite clever enough yet. Bidule is another candidate, and there are scripts available that do the magic.
Mostly it comes down to integrating one of these solutions, and thus far that has proven to be non-trivial. It becomes even more "interesting" if you use Spitfire UACC<G>!
I have experimented with a couple different iPad controller apps, and this might prove to be pretty cool, but right now I end up spending more time focusing on the iPad than the music, so that's not working<G>.
Of course that only addresses articulation switching - a major component of working with virtual instruments, but not everything. And in this case the CC messages are used as switches, that is, some small range of values represents each articulation.
The other thing I'd dearly love to see is better drawing tools for adding continuous controller data to the PRV. The current approach makes it difficult to draw something like Expression, or even volume in. I can record the controller itself, and often resort to that, but that isn't always as effective as just drawing the curve in COULD be.
Neither of which requires all the functions in Expression Maps - and I too have looked longingly at that other sequencer. But the cost, in $ and time, still outweighs the benefits for me.