Hi,
Having some weirdness with Transcribe receiving spurious midi messages - I’m sending on channel 1!
<-cmd
254<-note (midiParam1)
14<-channel (midiChannel)
240<-command (midiCommand)
254<-value (midiParam2)
Whether from the macbook or Korg SQ1, it gets a seemingly tempo based message on channel 14 or 6. Tempo based as, if you raise the speed on the SQ1 the messages arrive faster. No (actual) note messages are carried, when pressing a key, the pro-micro led blinks but nothing is seen (in the serial monitor of VS Code) except the above messages over and over.
Any ideas? Reacting as normal to the Nanokontrol on USB midi.
thanks,
Robin
Perhaps should add I reflashed the pro-micro a few times with the stock code with the same resulte
not sure exactly what this is - but could it be some kind of midi sync message ? you might be able to turn this off on your device and see if the messages stop ?
from a little bit of googing i found this:
254 (0xFE): This is a System Real-Time “Active Sensing” message
Active Sensing is sent periodically (often every 300 ms) by some MIDI devices to ensure the connection is still active. If no Active Sensing messages are received within a certain period, the receiving device may stop processing MIDI data
so it sounds like this could be part of it - based on those print outs it seems that transcribe is not correctly decoding these messages (since midiChannel and midiCommand are outside of the range you would expect here)
i think the best next step would be to add a new Serial.print() above this step in the code which dumps the raw midi message before transcribe tries to decode. then you can google this message to figure out what it is (or post it back here!)
Aha thank you so much that is invaluable! I did’nt think to look up the messages though I have done that before when trying to work out how to change NRPN messages.
I’ll try and filter out active sensing and see what happens then.
I’ll post back with how I got on!
So I got a better response from an Arturia keystep, which transcribe receives notes on the correct channel when sequencer is stopped - however when the sequence is running, everything goes strange again and the incorrect channel messages are received.
I’ve not figured out how to add the serial print step in the code. I will try next to filter out everything except notes using a hardware midi wrangling device.