Conversation
This seems to works for me for adding extra mididevices to the show. Need to do some extended testing, but in short,. I needed to connect multiple midi devices, anmd this way, I can give setupMidi a midiobjectname. Not sure if this is the best iomplementation, but it seems to work and it should not break existing code. _still testing_
|
test look fine, but it might complicate the API a bit? Although everything works as intended (and as it used to, with the midi5 object) if you do not add the extra parameter |
|
Heyhey - cool idea, so it's not locked down to a single variable/scope – i need to test a bit before merging. i did a quicko update release yesterday to include the latest p5.js + some bugs within the demos file that got messed up during a data-loss.. so can plan for it in the next version. In general, curious what your midi setup is that you needed multiple instances? I guess having multiple separate gears plugged in on their own midi port, rather than chained and using diff channels? I haven't tested much outside of a single gear, so mostly made sure channels/notes/control-changes worked (testing with a midi keyboard + elektron model:cycles). |
|
you would not want to see a screenshot of my connected mididevices :-) It comes and goes... but mostly I have mappings ready for separate devices, If I can attach them to a separate object, I can quickly swap controls in software. Using mainly selfbuilt hardware, it gets messy real quick. I adhere no standard mapping, I adapt to the device Example this one is for a more standard device I have (OG OP-1) , adapted to use the above patch. (it still uses some global vars and I still merged it with another controller (tidal) here so might not be the cleanest code, but you get the idea) |
This seems to works for me for adding extra mididevices to the show. Need to do some extended testing, but in short,. I needed to connect multiple midi devices, anmd this way, I can give setupMidi a midiobjectname. Not sure if this is the best iomplementation, but it seems to work and it should not break existing code.
still testing