Help! - MIDI information processed incorrectly (w/ video)

[devinb]

On the first pass of a 4 bar loop, all notes sound correctly, on the second pass beat one of the fourth bar is dropped entirely. When the loop is then shortened to just the fourth measure, the missing chord from the previous playback is dropped the first time before then loop plays normally four times, after that, only the highest note of the chord in question will sound.

 

The problem is always present in bounces, and even if I delete the region and copy and paste it in from elsewhere in the song where it plays correctly, or recreate it on another track or in an entirely different project by right clicking it in or playing it in with a keyboard that was just returned to factor settings the problem persists...I bounced the track at least five times and the issue always showed up.

 

Not in the video, but curiously, if I switch the piano sound from the stock Logic Steinway sound to a Native Instruments piano plugin the same chord is affected, but in an entirely different way - instead of being dropped, the chord becomes roughly a 16th note, and the key velocity seems to spike to 127 even though the list editor still shows the chord being a quarter note with the velocity steady at 80.

 

If I drag the region in question to another point in the song it's glitchy in a slightly different way, even if I don't move automation (there's no automation on the track put there by me, or displaying, not that I know of a way to make automation work differently on different passes of the same exact section), if I copy and paste, it does bring the issue with it, even if I don't copy automation.

 

Unfortunately, it seems that the issue is not limited to note length, key velocity or level, all editable parameters of any virtual instrument or effect seem subject to the changes. I have recreated affected projects from scratch on a second computer and had the same issue, even though the only thing that the two computers had in common was that I had to log into my App Store account on the second computer to download Logic.

 

Any help would be greatly appreciated.

 

[YouTube]

[/YouTube]

 

The video is not embedding for me, but I'm new to the forum, so perhaps I'm not doing something right, or approved to add a video...at any rate, here's the link:

 

[devinb]

Since my initial post I made the discovery that sometimes (I have yet been able to identify the conditions that cause the problem consistently) when two notes of identical pitch are overlapped (or sometimes as in the case of the loop in the video, simply absolutely adjacent to each other) it seems as though the information regarding when to end the second note becomes confused..specifically, the second note to sound is cut short to the endpoint of the first note - I now suspect that the aforementioned differences I experienced with Logic and Native Instruments pianos have less to do with any sort of differing architecture but because it seems as though Logic's piano samples are offset by about 5 milliseconds when compared to NI's.

 

I have a couple of questions...

 

First, could this be a clocking issue, and if so, is there any way to address it? Most of what I do is in the box and I haven't been the least bit concerned with buying an external word clock so I can't say I know much about them, but it seems from briefly checking out a few that one would only help at the A/D conversion stage, and wouldn't help the computer sync things for playback? If that is the case, is there a way to improve the computer's internal clock, or supersede it with some other external unit? - there must be a reason why an external clock can easily cost as much as my entire iMac, right? I've had this issue on a MacBook Pro on which it drove me so crazy I put in a new Logicboard, several new hard drives with clean installs of the OS and Logic, and finally replaced with an iMac which to this day has the same issue.

 

I've also had similar issues in time based effects or parameters of instruments that involve time (ADSR, Glide and so on...)...