[Gardner Raymond]

Thanks, both!

I generally haven’t had any issues at all using the linked comps, even in projects with many of them. However, I will keep that in mind going forward.

[Gardner Raymond]

After a bit more sleuthing and watching through the entire video with QT player, I think I figured it out. At around 19:23, it started getting little video hits/glitches occasionally, but throughout until the end of the video. That’s about the time that PPro starts crapping out, so everything is pointing to a poor export from the videographer.

Sorry for taking up your time with the longer first post, though I’m still open to suggestions/thoughts if you have any.

Thanks all!

[Gardner Raymond]

Awesome! Thanks so much

For mac, it seems to be Cmd+Shift instead of Ctrl, but still, it works!
Thanks again

Gardner

[Gardner Raymond]

Thanks Oliver – two very good options!

Still gathering thoughts though, so anyone else please feel free to chime in as well

[Gardner Raymond]

Good advice it looks like. I’ll add it to the list!

Thanks again

[Gardner Raymond]

Hi Dave,
Thanks for the feedback. That’s odd, because I know I’ve imported other mp4 files this way on different machines.
I just did a little test, and this is what worked and what didn’t:

-exported my clip to ProRes 422 and imported to MochaAe – no problem
-Tried other mp4 files that were shot with the same camera (Sony a7S, 24p, 50mbps) – error – unsupported format (though like i said, I know i’ve used clips from this camera in the past)
-Tried other mp4 files shot with a different camera (Canon C100) – opened & cached no problem.
-Tired other .mov files shot with a Canon 5D – opened & cached w no problem.

I find it odd that some flavors of mp4 work and others don’t. I’ll also have to run these same tests on my home system (I’m at the office now) where – as I said – I’m pretty sure I haven’t run into any issues.

Anyhow, thanks again. Annoying workaround for the time being, but at least it’s only a few shots w this project. Cheers!

[Gardner Raymond]

oh jeeze…what’s wrong with me? How do I not see that? I’m an idiot sometimes.
Thank you.

[Gardner Raymond]

Roland,
You are correct – there is no anomaly outside of the KF rage, only within.

I guess my confusion comes from the way I expect the KFs to correct the motion, and they way they actually do. I understand that the positions in between the good KFs have a higher probability of being “bad”, as that’s where the drift in the original track occurred, correct?
When it comes down to it, I just want to get the adjusted track to be correct with as few keyframes as possible. Do you (or anyone else reading this) have any suggestions for this besides trial and error?

I’ll post a screen cap in a bit when I get another few free moments. In the meantime, to further clarify (or confuse) things, here’s the most accurate way I can think to describe what’s happening:

-Say that the whole clip is 30 frames
-The tracking point of corner 1 of the surface is correct from frames 0-10 at XY position (0,0), and is correct again from frame 21-30 at (0,0);
-it drifts up and right between frames 11 and 20, with the “worst” position being at frame 16, having drifted to (10,10);
-I set a “good” kf at frame 10, and another at frame 21;
-I go to the worst drift frame of 16, and adjust the point to be back where it should be at (0,0);
-Now, the point is adjusted correctly between frames 16 and 21, and the drift is corrected. However, if I back up in the timeline one frame from my adjusted one to frame 15, the point immediately jumps out of location, back to (10,10), and then slowly goes back to it’s correct location as I continue backwards to frame 10.
-many times when this happens, the only way I can get it to say in the correct location is to set a new kf on (nearly) every frame in between, and I know this isn’t the best method (and defeats the point of tracking!)

Thanks again for your help, and especially at such an hour! If anyone else has any thoughts, they’ll also be very appreciated.

Cheers, all!

[Gardner Raymond]

Roland,
Again, I appreciate your input. I don’t mean to sound like a jerk here, but I completely understand what you suggested before, and that’s how I work when I’m using keyframed data linked to tracking data – in Mocha or doing roto work elsewhere. You use the “good” frames as anchor points on either side, and adjust the most drifted/”worst” frame in the middle, and cut down your problem areas by essentially half each time. Once you get the frames in enough good places, the program should interpolate the in-between motion correctly.
BUT, this isn’t what is happening for me. Even if there is still a drift problem in between two good keyframes, the point should drift back to the “bad” point somewhere in between the good ones, not immediately jump back to the “bad” point on the very next frame.
I understand how this could be confusing reading my description without actually seeing it happen, so please let me know if a screen cap would help, or if you have any other ideas.

Thanks

[Gardner Raymond]

Thanks for the quick reply. Perhaps I’m just confused, but what happens when the point in Adjust Track drifts off the point permanently and not back to a good point? Where do the “anchor” keys get set? In this particular corner pin track, the point stays in the correct position for about half the shot, then drifts to a new spot, and more or less stays there for the remainder.
Using the described method, I would think to
-set my Master Frame at the beginning point of the shot, continue to the frame just before the drift begins, and set another
-continue to the point where the drift more or less stops, and set a new frame with a new adjusted point
-work in between the last two set keyframes to fix any additional drift

My problem occurs in between these last two keyframes. Inevitably there is a little more drift to get fixed, and when I start adding additional keyframes (always first at the “worst” drift point), this issue of “jumping around” crops up. This morning after your reply, I re-tracked the footage and really paid attention to my keyframe strategy, but I’m still having the issue, though this time, it’s a different corner.