indeed Spectral Conquest can send OSC data via
sendOSC( host:port , path, dataformat.. , { values,.. } )
but this has a lot latency because i would have to translate back to midi via in example OSCulator
my idea is to script an Audio2Midi in lua with just an
upper threshold for midi note on
lower threshold for midi note off
or could i implement "luamidi" in my scripts even if i dont know how this moment?

and is it possible to setup the GUI's Parameter Names from my extra script too without involving "gui.lua" ?

OSC is the only way to do this right now. The latency should be very slow, if the OSC-receiver is on the same machine.

GUI scripts and MIDI/OSC behaviour scripts are entirely seprate.

Well,
latency should be slow, even with specialized software it would be so,
because recognizing audio streams into midi notes and not just triggering start signals needs to watch until the end of a sound frequency and calculation when to stop a midi note even if there could be the next note in the same frequency spectrum. I decided to experiment a bit. So maybe it is easier to record the note on and off in a lua object and when called from midi to send them out completely as package to osc.

for realtime recognition i have Widi audio to Midi Au. but it is a bit to raw, to mich notes or to less, i think they use just one trigger level and not two.

And now i'd like to say, that GUI improvement just for names of parameters would be on my wish list for spectral conquest. so i can remember what kind of parameter i thought to setup with this 4 wheels.

will see what i find out how to..

thanks