CGTalk - zero rotation bones

st45

06-11-2004, 06:25 PM

On some of my characters I have been using the "record pivot rotation so that the bone at a joint has a rest rotation of 0,0,0. This greatly reduces popping. However I've found that there are some drawbacks with this. For example, it seems that the IK Match goal orientation ignores the recorded pivot rotation, therefore, if the IK goal has zero rotation, the bone's rotation is not also zero.

To workaround this, I'm considering making clones of the joint bones, and then parenting ther original join bone and it's heirarch to the clone. This produces the same effect as the record pivot rotation operation, but hopefully without the "bug."

My question is, are there any drawbacks to having so many bones in the rig. Will it slow down my rig significantly, etc. If there is anyone who has tried this technique, please let me know if you have any thoughts on this.

Mark

Ramon

06-14-2004, 01:29 AM

As to your question, I believe that Kretin had posted something about cloned bones for that purpose on another thread. He could answer that question.

As to RPR:
It can be a life saver or a problem maker.
I have to say that I have also had issues with RPR with regards to morph targets (joint morphs). The morphs only work correctly when the bone's rot is unchanged in Layout. So if you convert skelegons to bones then use RPR on a joint bone and it rot axises change, then forget about your joint morphs working correctly, unless you use bone twist to realign the pitch. Good luck on that though because I haven't been able to get bone twist to work.

Castius

06-14-2004, 02:43 AM

You shouldn't have any problems at all. I created a script to create deformer bone on top of my Z+ game rig for TV work. It basicly does just what you are doing. When i get my site updated i'll have this one on there. No idea when that will happen though.

Rig_deformer video (http://www.steelronin.com/media/mov/rig_deformer_medium.html)

Other aspects of LW also don't read pivot changes. So I have stoped using RPR as much as possible. The best way to set up bones for rigging is with little to no rotation. So I hope we can see this problem go away in time.

Eugeny

06-14-2004, 06:26 AM

To eliminate this problem follow these steps:
1. Make sure your goal object have 0.0.0. rotation.
2. Make this goal child of the bone which mast be affected with " Mach Goal Orientation " , do this with "Parent in Place" OFF.
3. Make Parent in Place ON and make the goal object child of any other object (depend on your rig) or simply unparent.
4. Make Record Pivot Rotation for the goal object and activate him as a Goal object for the affected bone and turn ON " Mach Goal Orientation ".
Some note here - parenting with "Parent in Place" OFF will move your goal object - but if u use " Mach Goal Orientation " only its doesn't matter.
Nevertheless if u use goal object with "Full time IK" and " Mach Goal Orientation " u mast enter 0.0.0. on the move numeric field before step 2.
Good luck,
Eugeny.

SimianLogic

06-14-2004, 12:07 PM

Something else you can do is to make a reset bone. At the top of my IK chains, I usually add in an extra bone that doesn't do anything, oriented how I want the child to be oriented. It's usually really short, carries no weight, and it's starting point overlaps that of the bone. This seems to help with rotational wonkiness.

CGTalk Moderation

01-18-2006, 11:00 AM

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.