As to Beepster, I agree he's often a good fellow, I've seen helpful posts frequently.   That doesn't mean he is always right or always nice, however.
 

williamcopper
That doesn't mean he is always right or always nice, however.
 

Now THAT I can agree with.

Beepster
 
 
Dood... you do not understand what is going on (or are being intentionally obtuse). There is nothing left I can say that will make you understand WHY the export engine needs RAW, CONISTENT audio for these types of tests.
 
So I'll just say this... your entire thread premise is faulty. There is nothing wrong with Sonar export and you have absolutely NOT proven anything to the contrary.
 
 

 
DOOD:  this is not about sonar export, my friend.  It is about how Sonar renders midi run through Kontakt to make an audio track.   Can I say it more clearly?  I don't know how.

williamcopper

Beepster
 
 
Dood... you do not understand what is going on (or are being intentionally obtuse). There is nothing left I can say that will make you understand WHY the export engine needs RAW, CONISTENT audio for these types of tests.
 
So I'll just say this... your entire thread premise is faulty. There is nothing wrong with Sonar export and you have absolutely NOT proven anything to the contrary.
 
 

 
DOOD:  this is not about sonar export, my friend.  It is about how Sonar renders midi run through Kontakt to make an audio track.   Can I say it more clearly?  I don't know how.

I understand what you are doing. You are not going to get a consistent result. The wild card is the synth and how it interprets/plays back the MIDI on each pass.
 
Perhaps you should go yell at the NI boards about it. I'm sure they'll have just as much fun as we have here.
 
Based on the data here you are accusing the Sonar export process for the discrepancies. That is completely flawed logic.

And you have altered your OP which could be construed as "moving the goalposts". Still though the wildcard would be the synth. There is still also the possibility of hardware shenanigans in play. Your refusal to test this properly means there is NO way to find out what is going on...
 
so around and around weee goooo...
 
This is the type of thing that makes Beepster "not nice".

Another thing to examine (ye testers of testosterone) is that "tail effect" of using "Bounce to Tracks" with a selection of midi data and VST output tracks.   You bounce ... silently .. and at the end there's that little explosion of sound.     I've always ignored it, but parenthetically, as mentioned in one of the many posts above, sometimes it can surprisingly show up at the beginning of the next Bounce to Track for a different selection, different VST. 

williamcopper
 
DOOD:  this is not about sonar export, my friend.  It is about how Sonar renders midi run through Kontakt to make an audio track.  

In this case you need to test Kontakt as well, in as many DAWs as possible.
1. Take (make) a MIDI file.
2. Import it into DAW#1
3. Route imported MIDI to Kontakt within that DAW.
4. Load desired instruments, libraries.
5. Render and export audio using different methods, that are being questioned (real-time and offline in this case).
6. Repeat steps 2-5 with a different DAW.
7. Compare exported audio
a) per method;
b) per DAW.
8. Post detailed reports with screenshots and brief descriptions.
 
 

For any one's fun: here are two audio samples, the first mixed from 5 audio tracks made with "Fast Bounce", as above Midi and Kontakt.   The second mixed from the same time range, the same source tracks, and made with "Fast Bounce, 64 Bit Engine".     They are different, they sound different. 
 
Download of the wav file should be enabled.   These are very brief, about 9 seconds.
 
https://soundcloud.com/williamcopper/0513-2-fastbounce-test
 
and
 
https://soundcloud.com/williamcopper/0513-2-64bit-test
 
For me, the 64bit engine option, whatever it does, sounds very nearly like the original, while the ordinary fast bounce option, which I have usually used, does NOT sound quite the same.   (edit) And THAT is why I posted here!!!
 

Soundwise, yes your prescription is a good one.   Just that it's a lot of work, and mostly geared toward fingering Kontakt as the bad guy here.   I just don't think this is a Kontakt issue, but I may get to it.   I have Cubase and samplitude, both could do this test.   All the loading and routing for this project is easily several hours of work ... it's not just the midi data, but you have to have the same midi going to the same instance of Kontakt loaded the same way.  

williamcopper
Just that it's a lot of work, and mostly geared toward fingering Kontakt as the bad guy here.  

Yes, it'll take some effort, but it's well worth it. And no, Kontakt is not assumed to be the bad guy. You need to find  out whether reported inconsistency happens only with Sonar or with other DAWs as well.