View Single Post
Old 8th November 2007
  #185
Quote:
Originally Posted by pianodano View Post
Sure, here you go. Be prepared to spend sometime reading this to fully understand what's going on.
MIDI "Jitter" - It Does Exist=

BTW, here's my midi i/o setup.midi ports reordering - still not fixed
Much obliged!



[update]

pianodano -- I'm going to share my thoughts briefly off the top of my head -- and then I'm going to give this jitter thread a proper reading.

But it seems to me an excellent test for WIPINWIG (What I Play Is Not What I Get) in general is recording the audio you're monitoring as you track your MIDI and then comparing. There would be two main classes of error, I'm thinking: 1) a uniform shift forward or backward in the timeline and 2) irregular MIDI playback of MIDI-recorded information (as documented in the audio recording) which would suggest a form of MIDI jitter (irregular timing of the MIDI clock).

If it's the former and it's consistent, Sonar has an adjustment similar shift the MIDI by a given number of ticks.

If it's the latter, you'd need to isolate the problem.


I began using this MIDI vs audio test when I started syncing my ADATs to the old MasterTracks Pro in Windows 3.x around 1993.

(I switched to Cakewalk in '96 when Windows added a multichannel audio layer to the OS, CW jumped on it and MasterTracks took a wait and see attitude to see if it would catch on, only coming out with an audio sequencer a few years later.)

I did have occasional problems in Windows 3. But they weren't nearly as often as one would expect from that benighted OS.

Now -- mind you -- my criteria was fairly generous... if the audio from tracking and the MIDI playback of the same performance were indistinguishable (a bit rare) or minimally phase shifted, I considered it acceptable.

But if there was anything approaching distinct iterations (flamming) I considered that straight up unacceptable. Of course, to test I used transient-laden percussion. I also often panned MIDI to one side and audio to the other to get a better idea of what was going on those mostly rare times when there were problems.


I'll go dig into that thread, now...