an update on the organ console code!

Awesome! This is what I have been waiting for since you wired it to MIDI. I find it amusing that one note is going “Help meee…” and staying on. 😄😄

Wesley McIntire

thanks chris ill be loading it in a sec.

Look Mum No Computer

Please grab the very latest (just uploaded) It should make it all the way through with all of the stops pulled. Both of them should, I think! I tested with the SwellToGreat, SwellToPedal, GreatToPedal switches hard-coded to on (well, every single switch) and it got all the way through the file, so this might be a fully working version. I'm just not sure if the routing logic for the X-to-Y switches is inverted or not. I would also double check that the pins are being configured correctly in setup: StopStates[SwellToGreat] = digitalRead(SwellToGreat) == HIGH; //18 StopStates[SwellToPedal] = digitalRead(SwellToPedal) == HIGH; //17 StopStates[GreatToPedal] = digitalRead(GreatToPedal) == HIGH; //16

nice chris! ill be loading both of the latest codes and shooting a vid before lunch for you both.

Look Mum No Computer

I was able to test on my own hardware with a MIDI loopback and I can confirm that there isn't an issue with the output buffer anymore.... however, I'm still running into the stuck not issue. I think it's the input buffer at this point, so I can focus on that some more. Much faster now that I have a way to test on an actual device. Please feel free to test if you like, but it'll definitely run into stuck note issues with the stress test.

Another new version when you get a chance: https://gist.github.com/criggs/a647faa434e7814c63aaae4752d2ad40 * I implemented a ring buffer with effectively triples the size of the output cache (64 midi notes can be cached, which is 128 bytes of buffer space) * The ring buffer uses almost all available ram left on the little nano (2KB in total). If we use an Arduino with more ram, we could get MUCH more output buffer/cache! * Each loop iteration will throttle the number of MIDI output events. It's currently set to 24, which is less than a full output buffer (32 MIDI ON or OFFs). This should prevent the output buffer from overflowing, and can always be tuned down. * I added an automatic panic mode, so when the ring buffer is completely full, it will go into a panicked state for 10 seconds, clearing all of the input buffers and sending MIDI off to all channels. If it ever gets into a bad situation with stuck notes, just wait and it'll fix itself (this can be adjusted to a shorter time if desired) * No notes should be dropped without going into auto panic (assuming the send per loop limit is low enough), it should be lossless in that sense. * The only real CON for this update is that there might be a few loops of delay before all notes in the buffer are written out. I disabled the SwellToPedal and GreatToPedal switches for now, I haven't had a chance to debug that issue yet, though it might be a pin configuration. Feel free to uncomment those lines and see if the input pins are reading correctly. Good luck!