I noticed my Bass Pan was off center in Trilian in a certain project.  When I try to center the pan and restart the project at different points in time, the pan changed back to off center.  After a some investigation I tracked it down to certain midi clips.  If I started the project on these midi clips, the pan changed.  After I bounced the entire midi track, creating a single midi clip, the problem went away.  There are NO events in these clips other than midi notes.  And I have no Panning Envelopes.   I don't know if this is X2 or not.  I've had similar issues with the Pan setting in Trilian in X1 that I couldn't figure out.  How can a midi clip change the Pan setting inside the Trilain mixer?  
  
I also found a spot in the track where there are no midi clips but the pan changes.  So maybe it is not a midi clip. 

I'd say that the problem needs deeper inspection, because there's just so much that needs to be eliminated before it can be said it's X2.

I'd start of with no automation lanes at all, so that automation can be immediately counted out as a source of the problem. Then, I'd try muting all tracks except the the one that's *supposed* to be going to channel one, play that and observe if the pan moves. If that works ok, enable the other tracks one at a time to see which one causes the pan.

-Gene-

Gene -  Thanks for the advice.  I actually did more than that to troubleshoot it.  I deleted every track and every buss and every envelope and every instrument except Trilian and it's audio and midi tracks.  The midi track has nothing but midi notes.  No controller events.  

Then I cloned the midi track without events and archived the original track.  Archiving the track made the problem go away.  Then I tried copying over the clips in question and isolated the clips that were responsible for the issue.  It took a good hour to get to the point where I could tell exactly what was causing the issue.  I filed a problem report and uploaded the stripped down project with directions on how to re-create the problem.  It is totally repeatable.

I checked every other project and it seems this particular project is the only one that exhibits the issue.  This project is one of my most complex from a time signature aspect.  It's got about 50 time signature changes.  Don't know if that has anything to do with it.

Bouncing the midi track to one big clip makes the problem go away. 

Sounds like corrupt MIDI data, although one would expect the corruption to a) show up in the event list and b) be propagated to the new clip after bouncing. I hope you'll report CW's findings when you've heard back from them.

Did the problem go away when you disabled automation?

Just curious, is it possible you used Trilian's automation while recording those particular clips?  That does confuse Sonar going back to 7.0 when I first used Omnisphere.
 
But if you don't tweak on the fly when recording, I have no idea, but no problems here...

bitflipper


Sounds like corrupt MIDI data, although one would expect the corruption to a) show up in the event list and b) be propagated to the new clip after bouncing. I hope you'll report CW's findings when you've heard back from them.

Did the problem go away when you disabled automation?

Bit - I had no automation for trilian on this track.  But I deleted every other track in the project so at the end it was just the midi track and trilian audio track - no automation at all, but the problem still persisted.  


There are a couple really weird things:  1) there were two points in the project where if I started exactly on those points or after them the pan would change.  But if I started before the first point with the pan centered, once it reached theses points the pan would stay centered.  

2) the first point seems to be tied to a clip where if I deleted the clip the problem would go away.  The second point was in an area in the track where there was no midi clip at all.


3) the first point changed the pan left  about a quarter turn.  The second point changed the pan to the right.


I agree - must be corrupt midi data.  There was a point during the X1 madness where all sorts of weird things were happening.  One thing was this: some FX chains would crash Sonar if I tried to delete them.  I eventually just turned these chains off and left them in the project because I couldn't delete them.  I still have some of these corrupted FX chains today and to this day they crash X2 when i try to delete.  What I'm thinking is that something like this occurred with the midi data.

declan


Just curious, is it possible you used Trilian's automation while recording those particular clips?  That does confuse Sonar going back to 7.0 when I first used Omnisphere.
 
But if you don't tweak on the fly when recording, I have no idea, but no problems here...

Declan - it is quite possible that I may have hit a slider or mod wheel when I recorded the midi.  When I looked at the PRV there were two controller lanes cc:1 and cc:10.  But they were both totally empty.   Strange that they even showed up in the PRV.  And nothing in the event list.

 When I looked at the PRV there were two controller lanes cc:1 and cc:10.  But they were both totally empty.

I discovered through trial and error that Sonar will not display codes before the first note (in some cases?). I cannot say whether this is all the time or what certain conditions are necessary - I didn't bother to parameterize the issue any further.. Specifically, I noticed the same thing - the track NOTES view indicates a couple of CC's that you could choose to display and that appear to be in use, but nothing shows up in the event lists for that track. On a hunch, I opened the same midi file in Acid and looked at its event list there. Sure enough, there were those suspect CC's lurking in the event list, plain as day.