It was a joke: jitterbug.  Well, I should say, it was *meant* to be a joke.

Darn I had the elbow grease on standby..

konradh
There is a well-documented jitter bug in Sonar.

Well played, Konrad.
Made me laugh.

Easy to test the core of midi like this:
Make a loopback cable from midi out to midi in(on whatever ports you've got.
Play back one midi clip - and record to another track.
 
#1.Compare what you've got if they line up - this is roundtrip latency.
#2. then nudge recorded clip till they line up - and compare of all notes are in the same spot on each clip
Any deviance is what we might call "jitter" or part of limitations in midi as such.
 
To my surprise I found the RME internal midi port as out was slower than my EMI Xmidi 2x2 on usb.

This is a very old thread. I remember many problems years ago, but surely midi timing issues are a thing of the past?

Well, I found this thread also recently and wonder if it's an issue I experience with an Akai lpk25 keyboard.  The timing when I play it back is atrocious so quit using that particular keyboard.

tenfoot
This is a very old thread. I remember many problems years ago, but surely midi timing issues are a thing of the past?

That's what I'm wondering.  At the time of this thread, PCI and Firewire seemed to offer better performance than USB.  However, I was this thread before the advent of USB2 audio interfaces?  I don't remember when those became more common.  Has anything changed on the OS side to improve midi performance?  I'm guessing not as midi for games is even less common now than it was in 2007.  Maybe developers such as Cakewalk have developed kernel level api's for midi?  Maybe the 1-2ms jitter level has just been acceptable with consumers and things haven't really changed since 2007?
Dan

I found jitter with the 1820m to be sub-millisecond at the time for discrete events. I bigger issue is the delay of messages that are intended to be sent simultaneously in due to serial transmission which has not changed for hardware though soft synths rendered from existing MIDI tracks may not be subject to that. I'd have to check.
 
 

Even though this is an old thread sadly enough the issues are still the same. It's still all about the hardware/ connection /drivers/ software. That remains a variable as there is thousands of combinations of the four. 
 
I think that most of us do not notice the tiny amount that may be present even on a top notch set up. And because we tend to quantize the crap out of midi it not an issue at all. It's more an issue for people playing soft synths live. 
 
Generally when someone is having issues with MIDI Latency it's because they have a plug in like the LP 64 EQ or Multiband running. 

Cactus Music
And because we tend to quantize the crap out of midi it not an issue at all.

I look at this other way around; because I tend not to quantize the crap out of MIDI, typical levels of jitter/delay in the system are masked by the natural variations of a live performance. Based on my own long experience of recording and analyzing my own performances and some performances of others, I'm fairly confident that the vast majority of users who believe that SONAR does not capture and/or playback their performances with sufficient precision are deluded about the precision of their playing. Or their environment has issues that are not typical.
 
If you have a keyboard synth, you can easily check this precision by recording simultaneous audio and MIDI from a hardware synth under Local Control and then re-recording audio from the synth being driven by MIDI playback. Audio timing can't be variable (clicks, pops and distortion would result), so if the MIDI timing matches the original audio track, and the two audio tracks match each other (after compensating for MIDI and audio latency), it's all good. Of course, this test assumes that the response of the keyboard synth itself is fast and consistent in both directions - not always a good bet - but you don't need to worry about that unless you find problems.