Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!clyde!cbosgd!osu-cis!ogg!mark
From: mark@ogg.cgrg.ohio-state.edu (Mark Jansen)
Newsgroups: comp.graphics
Subject: quaternions and homogeneous matrix transforms
Message-ID: <955@ogg.cgrg.ohio-state.edu>
Date: Fri, 13-Nov-87 12:45:34 EST
Article-I.D.: ogg.955
Posted: Fri Nov 13 12:45:34 1987
Date-Received: Sun, 15-Nov-87 17:11:07 EST
Organization: Advanced Computing Center for the Arts and Design
Lines: 25
Keywords: orientation quaternions transforms 3dgraphics

Hi,

we have been having some discussion about whether we should be doing our
3d orientation transforms in the old homogeneous 4 by 4 matrix method or
moving over and using quaternions.

I was wondering what experiences positive or negative that people have
had with quaternions.  We have read the couple of siggraph articles on this.

Duff's paper in the 1986 siggraph course notes for "State of the art: image
	synthesis"
Schumaker's papers in the 1987 siggraph course notes for "3d motion control,
	computer animation".

Obviously there are a number of things that quaternions do better, but what
	if anythings do they do worse?  There must be some limits on their
	usefulness, what have people found?

Can you do keyframe interpolation easily?  Will it walk the dog?


-- 
{cbosgd,ucbvax}!osu-cis!ogg!mark

osu-cis!ogg.cgrg.ohio-state.edu!mark