Hi, we have been using the mocha demo for the last few days at the studio i work for.
ive watched the few tutorials that are on the website and am slightly
confused by whats happening. when i try to recreate the tutorials it never is
as easy as what they make it seem. alot of times when the object im tracking
moves off screen my matte has a hissy fit and goes out of control, making it
impossible to manual track. i also would like the know the difference
between the large motion and small motion option. if anyone knows about
private lessons or anything like it please tell me.
I posted a similar question some time ago…
…here is the link to the thread and the reply.
As you might notice, it doesn’t really put an applicable answer to a real life situation. Exactly, how to use the “small/large” motion issue seems to only be learned through experience.
As for any suggestions, first and foremost… Save Frequently… and even that may not work. I have had a few crashes that has corrupted both the working copy and the saved copies (corrupting copies is …in my book…strange). Most of the time, the work was not completely lost, but many hours of the most recent work would be lost. But in all instances this has been when I used long video clips with many layers and worked on for many hours/days. So, it is best to not work on large projects at one time…nor, try to develop too many layers in conjuntion with long video segments.
What is too long? Well, that is tough to say exactly since it seems that the number of layers seems to fit into the equation as well. However, segments that are in the 2+ min. range with 10 or so layers definately fits the high probably of problem category. But, short segement (under 1 min.) with only a 3-4 layers seem to work fine.
It’s a VERY good tracking and roto-ing tool, but definately has some saving file issues when dealing with large files (or, sometimes when you work for many extented hrs. on the same project).
Finally, to answer your question about how to add stability to your project. Based on what you posted… I am guessing you are trying to track too many points on too large of an area. You will have much more success if you take the object and add multiple tracking/roto-ed layers. these smaller layers will track more closely and more accurately. For example, if you have a head turning, you might use 3 or more sections of the head to get a good roto-ed layer. If you are trying to track legs moving… well, my the lord have mercy on your soul … tracking leg movement is tough… to say the least. This is why tracking and roto-ed layer are best suited for inanimate objects with sharp edges and more cohesive linear-like movement.