[Sorry if this double posts]

I’ve been studying Brian Maffitt’s 3D camera tracker video  http://www.youtube.com/watch?v=M9ZeG7-QdSo  and comparing my own results with a clip I shot with a Glidecam system rotating around a basket sitting on a countertop.  I get entirely different results than those shown in Brian’s video.

After analyzing/solving my clip by way of “Track Camera”, then picking the tracking points that create a target plane of the correct orientation, then right-clicking that target and choosing Create Text and Camera, things go incredibly different than as shown in Brian’s video at 3:30. In my clip’s case, after applying text and camera, if I then scrub through the comp my video clip layer wildly moves around and goes off screen, revealing the black background, yet with all the tracking points and my new text layer correctly in place.  I find myself needing to replace the original video layer from which the tracking points were obtained because the original layer took off, flying around as if it was constrained to some imaginary null that is now the 3D camera’s point of interest.

Very annoying.  I’m doing the exact same things as shown in Brian’s video, but Brian’s layer of the tree limb does not start moving around after adding text and camera.  In Brian’s video, when he adds Create Text and Camera, his text is immediately visible in the scene, In my case, I see a text box, but no text. The text is actually laying behind my video layer. Very puzzling.

Any explanation?

Thanks,

John