CompuServe Messages

Jonas, Thanks for your note. Having a bend in each of the joints definitely makes a difference. I had some bend in the intial positions but the follow object stretched the chain to a nearly straight alignment, which caused problems. As long as the follow object is within easy reach and the rotation isn't too great, IK works easily. But if the rotation is fairly large, there can be problems. If the motion is broken down into steps it seems to help. The "snap" definitely appears when objects reach their rotational limits so limits (and which axes are active) need to be changed for each section to get smooth lateral motion. I guess my mistake was thinking that IK could solve for this type of motion in one pass. IKs solutions are technically right but look weird (so much for computers being intelligent machines <g>). A wishlist item: it would be very useful to have IK with force control (like gravity) to allow chains to have more natural motion tendencies with follow objects. I look forward to the day when I can just tell the models what to do and they do it <bg>. Cheers, John Mateer PolyMedia