The reason is because the 1ms intervals do not necessarily land evenly on 960PPQN tick intervals. In fact it completely depends on the tempo. This does not mean that anyone is getting sub-millisecond timing. Nobody is. The best you can expect is 1ms. But, because of the MM timer and the midi clock timer not necessarily being synchronized, errors can result there too which might cause a little jitter right around 1ms.

Yes. Makes sense to me.

I just switched my USB controller off and am using the midi-out into my parallel-port MOTU interface and can feel a noticeable difference while playing piano. I really think avoiding USB midi is key for discerning people. Given that Sonar always uses the 1ms timer, I see zero advantage to using lower PPQN than 960.

I think that's pretty telling -- at least for your particular system

I'd like to sum up some conclusions I think we've arrived at, through the course of this 13 page (!) thread:

• Sonar's performance, as far as MIDI jitter is concerned -- is at least as good as any Windows app we know of.
  
• If you're concerned about MIDI jitter, a parallel-port or PCI-based MIDI interface should give the best results.
  Avoid USB MIDI interfaces**. (The verdict on Firewire interfaces is not yet in)
  ** Some USB MIDI interfaces may have relatively good timing - but look for clear test results that show it.
  
• Soft-synths, if written poorly, can have really lousy MIDI jitter. That is not Sonar's fault.
  
• Soft-synths, if written well, can have sample-accurate MIDI timing. Sonar gives those soft-synths what they need in order to deliver the goods.
  
• 960 PPQN resolution is a good thing. Use it.
  
• A mis-configured system (or one with a number of different MIDI interface drivers) can produce driver conflicts that do bad things to MIDI timing. This one is based on personal experience, and also on reports from at least one user with 5! MIDI interfaces.
  
• The built-in metronome, for some people, has been associated with timing problems.
  For details and workarounds, see http://forum.cakewalk.com/fb.asp?m=1191679, below (post #379, by dewdman42).

I'm offering these conclusions partly so that the key points we've concluded are all in one place -- and also so that, hopefully, nobody will use this thread for Sonar-bashing purposes. I think Sonar is great. I've been a bit concerned this thread might have raised some eyebrows, given that a new Sonar release (with tons of MIDI improvements) has just shipped. I think we now have a better handle on the kind of performance Sonar can deliver, and things people can try if their system is not doing what it should.

Just my 2 bits -- YMMV, as always.

Jim

Edited to add metronome bullet.

ORIGINAL: dewdman42
If you have 2-4ms of slop when recording your midi parts, then it does not matter what the resolution is set to..it only means it will record the slop more precisely... more precisely recorded slop! heh heh.. Its futile to think its giving any more accurate precision than that. and there is the possibility that the increased work load going to the computer to handle more ticks per second will make the slop worse.

Point well taken, but there are still benefits to using 960ppqn resolution, especially given the current proliferation of softsynths. Granted, MIDI data coming in from, or going out to, external devices won't benefit from the increased resolution, but softsynths rendered internally to the application will... and that means that 960ppqn resolution gives more precise editing options. It means that if I'm tweaking a BFD drum performance, whether that performance originated on a set o V-Drums, a keyboard or a prerecorded MIDI groove, I have that much finer control over where the drum hits will ultimately sound, and therefore the way the performance feels.

ORIGINAL: RTGraham
Actually, I'm thinking even more broadly. Microsoft should build an API that incorporates a similar feature, and then that should become a standard for all MIDI interface manufacturers to use. I don't want to be limited to just MOTU products - their Windows support is still sub-par, in my experience. In theory, there may be manufacturers whose interfaces already have the internal capability to support hardware timestamping, and those interfaces would be "upgradable" to the new standard through a driver update.

There is nothing limiting anyone right now from building hardware interfaces that are compatible with Apple's CoreMidi (which is MOTU's MTS). What I'm suggesting is that if the same technology is used for windows, then midi manufactures can build midi interfaces that would work with both CoreMidi and whatever MS builds...anyone could build it...and there would already be some MOTU interfaces in existence that would work out of the gate.

ORIGINAL: dewdman42

There is nothing limiting anyone right now from building hardware interfaces that are compatible with Apple's coreMidi (which is MOTU's MTS). What i'm suggesting is that if the same technology is used for windows, then midi manufactures can build midi interfaces taht would work with both CoreMidi and whatever MS builds...anyone could build it...and there would already be some MOTU interfaces in existence that would work out of the gate.

Got it.

ORIGINAL: Jim Wright

I'd like to sum up some conclusions I think we've arrived at, through the course of this 13 page (!) thread...
...
...
I'm offering these conclusions partly so that the key points we've concluded are all in one place -- and also so that, hopefully, nobody will use this thread for Sonar-bashing purposes. I think Sonar is great. I've been a bit concerned this thread might have raised some eyebrows, given that a new Sonar release (with tons of MIDI improvements) has just shipped. I think we now have a better handle on the kind of performance Sonar can deliver, and things people can try if their system is not doing what it should.

Nice summary, Jim.
And a good point, that we don't want to portray SONAR as having problems - quite the opposite, rather: we're pinpointing system-wide, industry-wide limitations.

And thankfully, it seems that we're all managing to get on the same "page." (after 13 of them! )

I agree, nice summary. I can't think of anything else to add, except for perhaps a comment about some known audio metronome problems which sometimes afflict people in Sonar in ways that can sound like midi timing errors.

ORIGINAL: RTGraham

ORIGINAL: dewdman42

Ideally, Microsoft would mimic Apple and license MOTU's MTS. That way, all the MOTU hardware would instantly become useable under Windows sequencers that use the new model.

Actually, I'm thinking even more broadly. Microsoft should build an API that incorporates a similar feature, and then that should become a standard for all MIDI interface manufacturers to use. I don't want to be limited to just MOTU products - their Windows support is still sub-par, in my experience. In theory, there may be manufacturers whose interfaces already have the internal capability to support hardware timestamping, and those interfaces would be "upgradable" to the new standard through a driver update.

Ideally, we would develop some new technology for MIDI interfaces, that would provide far better timing for MIDI-DIN input and output without overly burdening the host computer (or costing an arm, leg and torso). I know I'm being kind of a tease here - but if someone with some hardware chops (preferably FPGA design) wants to collaborate, let me know (jim period wright aaattt acm period org should reach me). I had a brainstorm a few years back, and it's itchin' something fierce

- Jim

ORIGINAL: dewdman42

I agree, nice summary. I can't think of anything else to add, except for perhaps a comment about some known audio metronome problems which sometimes afflict people in Sonar in ways that can sound like midi timing errors.

Good catch! Please suggest some concise language (you're more familiar with metronome does and don'ts than I am) and I will edit my post accordingly.

- Jim

Just that there are some scattered reports from people having wierd timing problems related to the audio metronome. Including me. This year I had a project, only a few tracks and it would start out fine but after working a while the timing started getting all messed up. I couldn't figure it out. Then I looked around on this forum and found other people with similar issues and it was related to the audio metronome. When I routed metronome to midi instead of using Audio, the problem disappeared. It seems this is some kind of wierd bug that comes up for some people. I have heard a few possible work around suggestions:

- Route the metronome to midi and use a soundsource like Session Drummer 2

- Record a quantized one bar metronome midi pattern, convert it to groove, put it on a midi track, playing through Session Drummer2.

- In the metronome setting (project options), send the audio metronome to a buss instead of directly to the hardware audio port which is the default.

- Here is a suggestion that looks very good too, which involves setting up a metronome audio track or set of audio tracks that output sample accurate metronome output: http://forum.cakewalk.com/fb.asp?m=1192305

I can't remember what I did to move foward at the time, I think I just turned off the metronome and did what I need to do. Never had the problem come up again. I only mention it because someone trying to troubleshoot their midi timing errors needs to be aware of this possibility and rule it out from their scenario just in case.

Dewdman42,

RE: doubling (or more) tempo to simulate increased resolution, Yep Roland recommended that way back in 1987 for the MC500mkII.


This is has been a good learning experience for me. You guys seem to have a real understanding of the challenges in making it work correctly,reliably, accurately (choose one) for critical listeners and discerning users, on a pc.


At the risk of a total flameout, (and to summarize what is starting to gel now in my twisted mind).

It seems reasonable (to me) to assume, that at the present time, it's as good as it's gonna get.

IF that is the case, it seems to me that based on the joyful noises, glee and downright exuberation exibited by untold numbers of users everywhere, that there must be many zillions of happy campers out there that are not even aware that "Houston we sorta have a problem". OR, perhaps they don't expect anything better ? OR even, (and I wince) they can't tell ?

Gulp! Dang.
Although I hesitate to really believe that, some part must ring true because I have seen quite a few concerned individuals shark attacked by many of them thar happy campers in the past for their efforts at trying to discuss and understand just what the heck is going on.

On a side note, if most of this is true, that must explain why I don't listen to the radio much any more.


Also thank you Jim for sharing some of your exceptional knowledge on this subject.