rogeriodec
Thanks for the reply, but no.
The example I sent attached in post 1 was only because TTS-1 is a standard installation of Sonar.
I use Kontakt, with very recent version (x64 5.5.2).

 
I tested your file with Kontakt. Same perfect results as all the other instruments I tried (other than the TTS-1). However, I was not using Kontakt multitimbrally, nor was I using patches that had extensive sync-to-tempo processing built in.
 

At the same time my hardware specifications are updated as well, as you can see in my signature.

 
How a soft synth responds to the data depends on their coding, what they're doing at any given moment, which mode of operation they're using, etc. However if anything, a faster computer might actually make what I think could be the problem worse because SONAR could push the now time ahead in the timeline faster than a slower computer in response to a tempo change, thus increasing the odds a soft synth that's slower to respond will miss a simultaneous Note-On event.
 

What about your argument on simultaneous processing, the problem occurs in a slow BPM, with only one channel, ie it gives more response time for processing the events.

 
You're thinking in terms of sequential events, not simultaneous ones. If three events happen simultaneously and an instrument has to parse those at the exact same instant, it doesn't matter what the tempo is. Events don't happen any less simultaneously at a slower or faster tempo.

And also this did not occur in older versions of Sonar.

 
Were you running them on a slower computer (see above)?
 
Your issue is not occurring here in Platinum, except with your specific file that has multiple events (including tempo changes at extremely slow tempos) happening simultaneously, and only with the TTS-1. Furthermore, note that by running at a really slow tempo, changing by 2 BPM from (for example) 10 to 12 BPM is a huge percentage of change you're trying to execute compared to changing by 2 BPM from, let's say, 120 to 122 BPM. I don't know whether that might make a difference or not but you're asking the now time to move way ahead, on a percentage basis, compared to where it was.
 
Now, maybe something can be done with SONAR to compensate for instruments with particular characteristics (e.g., if a tempo change command and note-on occur simultaneously, the note-on is offset forward somewhat to compensate), but that would not be considered fixing a bug. The bottom line is that excluding the TTS-1, in all my tests (including Kontakt) SONAR worked exactly as expected with the file you provided.
 
If someone has a different explanation I'm all ears. However I don't think the explanation is SONAR can't play back the notes, because it can.
 

Could this be related to the PPQN setting? I've never had occasion to do anything below about 40 BPM, but it would seem that extremely slow tempos might warrant higher PPQN values to counter the reduced timing resolution. I'm just guessing, as I have no DAW and don't remember what SONAR's default resolution is or if it can be increased.
 
 

To prove that this is indeed a Sonar bug, I exported this project as a MID file, I opened it in Sibelius, which was configured to use the same Kontakt VST, with the same instrument I used in Sonar.

And guess what? In Sibelius all the same notes sounded perfectly with the same Kontakt, and same instrument.
 
In Sonar, with Kontakt, I tested it with various instruments, and with all of them, the notes stop playing from the measure 1:02:360 (tempo 9.32), ie, the first 11 notes sounds; from the twelfth note, nothing sounds. 
 
So I tried switching to my midi standard output "Microsoft GS Wavetable Synth" and in this case all the notes play perfectly, regardless of the tempo.
 
Conclusion: If you use VST in Sonar, never use tempos below 10 BPM (preferably never use tempo below 20 BPM), or they will not sound.
 
Well, is it a Sonar bug or not?
 

rogeriodec
Conclusion: If you use VST in Sonar, never use tempos below 10 BPM (preferably never use tempo below 20 BPM), or they will not sound.

 
I would be happy to shoot a video that shows otherwise. It's irrelevant that Kontakt worked as expected for you in Sibelius; when I tested your file with Kontakt in SONAR Platinum, the combination worked as expected. What would be relevant is if it worked for you and didn't work for me. The only way I have of knowing Kontakt doesn't work in your system with SONAR is because you say so, and of course I assume you're reporting your experience accurately.
 
However - Kontakt played back the file you provided exactly as expected in my system, so the only provable conclusion I can come up based on my experience with at least a dozen instruments is: when driving the TTS-1, you don't want to mix tempo changes with multiple simultaneous MIDI events when creating extreme tempo changes (on a percentage basis) in the lowest possible tempo range allowed by SONAR. (If I remove any of those conditions - e.g., don't create extreme tempo changes, change some event start times so they're not simultaneous, etc. - then the TTS-1 works, too.)
 
As in these tests we both used SONAR Platinum, played the same file, and tested with the same instrument - yet you had problems and I didn't - that would certainly tend to indicate there's a variable other than the file, the instrument, or SONAR. What that variable is, I don't know because once we get to the hardware, driver, and system level, there are likely hundreds of variables.
 
Again, if you need a video, I'll gladly provide it. The only downside is that will preclude my doing a Friday's tip of the week, because I will need to spend the tip time on doing the video instead. But if you believe your conclusion is universally applicable, I can show very easily that it is not.
 
Perhaps I'm the only person who experiences flawless performance under the conditions you specify, but I know enough about computers to realize it's often naive to extrapolate an individual's experience to all users. I can only report on my own experience...just as you can report only on yours.
 
If others commonly mix tempo changes with multiple simultaneous MIDI events when creating extreme tempo changes (on a percentage basis) in the lowest possible tempo range allowed by SONAR, perhaps they could chime in with they've found so we can have a bigger data set than two people with different experiences.
 

No problem, Anderton, and again, I appreciate your efforts to check this problem.
If this problem does not occur with you (and we both using Platinum) possibly is a specific problem within some point of my hardware/software configuration.
Thank you.

A couple more questions...
 

• Are you using Kontakt in multi-timbral mode, or as a single instrument?
• Were you running your older versions of SONAR where you didn't experience this problem on a slower computer?

 
I think a major difference is I tested this on a laptop that doesn't have the power of your machine. This is why I think perhaps that with SONAR's recent optimizations and a fast computer, the tempo is processing the tempo information so rapidly that SONAR has already jumped ahead on the timeline before the note-on gets read. This is just speculation, I don't have the ability to test this. I did notice with the TTS-1 that nudging notes slightly ahead of the beat (so it wasn't simultaneous with the tempo change command) seemed to solve the problem with it, which would make sense if the issue was SONAR needing to read the note-on first, before the tempo command told it to jump ahead on the timeline.
 
This is why I also wonder about multi-timbral instruments. I know "computer MIDI" is much faster than physical MIDI, but isn't MIDI still a serial protocol, even inside the computer? Hmm...
 
It may seem silly that I'm taking an interest in anomalous behavior that I can't reproduce, but sometimes pursuing something will uncover other unanticipated issues you'd never find any other way. 

10BPM...I don't think I could handle the count in...
 
Dan

Anderton

• Are you using Kontakt in multi-timbral mode, or as a single instrument?

 
I use Kontakt in both cases.
 

Anderton
 

• Were you running your older versions of SONAR where you didn't experience this problem on a slower computer?

 

 
Exact. As I said before, I use BPMs below 10 to simulate a Fermata, and allow the MIDI stay intact for later use in Sibelius.I have old Sonar projects containing BPMs below 10, it means that they worked many years ago, of course with much slower computers than my current computer.

I don't see any reference to the MIDI Prepare Using buffer in the thread. This setting would clearly be relevant to dropped MIDI notes at very low tempos. In most cases, I would recommend setting this to the old default of 500ms, which seems to work well in most cases, but it might need to be quite a bit higher to accommodate tempos of 10BPM (6000ms between beats!).

brundlefly
I don't see any reference to the MIDI Prepare Using buffer in the thread. This setting would clearly be relevant to dropped MIDI notes at very low tempos. In most cases, I would recommend setting this to the old default of 500ms, which seems to work well in most cases, but it might need to be quite a bit higher to accommodate tempos of 10BPM (6000ms between beats!).

Hi, I tested here with 500ms, 600ms and up to 9000ms but apparently had no effect.