Midi interface

gistfile1.txt

// in sc..
MIDIClient.init;
~midiOut = MIDIOut.newByName("X18/XR18", "X18/XR18 MIDI 1");

// prototype: addMIDI { |name, device, channel=0| ... }
addMIDI("vbass", midiOut, 1)

// in tidal
d1 $ n "c a f4 e" # s "vbass"

// since channel is only default send something to tenth channel on same device as the vbass like this:
d1 $ n "40" # s "vbass" # channel 9

// these could all do the same, i.e. set pan (which is general midi control 10):
d1 $ cc "10:60 10:122" # s "vbass"
d1 $ ccv "63 127" # cc "10" # s "vbass"
d1 $ pan "0.5 1" # s "vbass"

// could also be paired with 'n' to both set the control change and set the note
d1 $ cc "10:60 10:122" # n "c7 a7*2" # s "vbass"

// The mapping from the 'pan' parameter to cc 10 would be best in superdirt I think, because tidal wouldn't necessarily know that 'vbass' was a midi synth.

// Starting midi clock could be as simple as this, the logic for this could be done in tidal:
d10 $ midiclock "vbass"

ok, good. Next is a list of parameter names we want to use in tidal. Should we prefix by 'midi' or is that too wordy?

e.g.

midichan
midion
midioff
midicc

etc.

cc could be an exception, because it is clear

you wouldn't name devices, but just link names and default parameters to them.