CompuServe Thread

#Copy Tracks ?

4 messages in this thread
#26242From: alec jasonNov 5, 1992 12:37 AM
Yann, Perhaps you can answer my question: After having "copied tracks" from one object to another using the "absolute" mode, I find that the two objects do not move and rotate the same. I have checked the XYZ values in the Track Info boxes and they are identical for both objects at each keyframe; yet the two objects are actually offset from each other in position and counter to each other in rotation. Are the XYZ values relative or absolute? I thought those values represented absolute positions and rotations. Could the fact that the two objects have different axis cause the problem? Thanks, Alec Jason
#26252From: Yost GroupNov 5, 1992 9:26 AM
Hi Alec, I'm sorry that I missed your earlier question about this. Without seeing your file, it's impossible to say what's going on. This is because quaternions _are_ relative. Plus, object positions are subordinate to the hierarchy. If you've got a simple example of what you're talking about, zip it up and email it over to me to take a look at and I'll let you know what's up with it. – G
#26429From: alec jasonNov 7, 1992 12:51 PM
Gary, Thanks for the reply. Could you please provide a definition for "quaternions." (Not found in my dictionary or the 3DS docs index.) I assume that quaternions are the XYZ values. Your message said that they are relative — but to what, the object's axis? (BTW, so the "G-" stands for "God" <g>) Alec Jason
#26431From: Yost GroupNov 7, 1992 1:12 PM
hmmm…. Quaternions are incredibly complex equations that provide answers to rotational problems that occur when rotational axes pass through the "north and south poles" of the world coordinate system. They were developed in the early 1800's to deal with gimble lock problems on large sailing vessels, and then Ken Shoemake rediscovered them about 15 years ago (at Xerox Parc I believe) and wrote some great papers about how to use them with 3D animation to provide smooth rotations in all cases. The books that Dan Silva used to develop our routines are about 700 pages long (each, there are two of them), and every page is filled with bizarre mult-dimensional geometry equations. If you want specific references to the book titles, or the ACM Siggraph papers that Shoemake wrote, I can dig them up for you. The rotations are all relative to the objects inherent transformation matrix. This is the orientation that the object is in when you look at it in Object Pivot mode. (again, if you want me to look at that file of yours, zip up a simple case and send it on over and I'll explain what's going on for you) – G(lad to help)