PDA

View Full Version : problem with FK control curves orientation on atypical model


cynicalbug
11-21-2010, 11:20 PM
I am rigging a model (not mine) that is not in the typical pose I have seen in rigging tutorials. (ie: arms stretched out across the x axis.)

The character has his arms down at his sides. My problem is that I can't make the orientation of the FK control curves match the joints AND freeze the curves' transformations without changing their rotation changing back to 'World'.

I have considered using a joint chain that is only constrained to the skeleton to act as an FK control curve as I can match the orientation of the joints while keeping their rotations zeroed out.

Does anyone know if there is an easier way to do this? (without going back and editing the character model to straighten his arms out ;) )

mccollom73
11-22-2010, 01:23 AM
Instead of doing an orient constrain on them, try this method.

1) make a circle at world space.
2) select it and then press down (to grab the shapenode)
3) shift select the joint aka shoulder, elbow, wrist
4) then type "parent -r -shape

this will make it like u r select the joints, but with controllers.

Let me know if this works

Josh

cynicalbug
11-22-2010, 02:48 AM
It did properly orient the curve to the joint. But it added values to the control curve's rotations.

mccollom73
11-22-2010, 03:29 AM
It did properly orient the curve to the joint. But it added values to the control curve's rotations.

It should have only did that it you have rotation in your joints.

Josh

cynicalbug
11-22-2010, 04:26 AM
It should have only did that it you have rotation in your joints.

Josh

hmmm... you're right. And, by the way, thanks for the help :)

The pole vector on the arm is doing it. I can switch to FK mode and zero out the joint rotation but when I go to IK mode the arm shifts. The control curve I used was snapped to the knee joint and moved forward the Z axis and then had its transformations frozen before I created the pole vector. Any idea why it would add the rotation?

Edit:
I've concluded that the pole vector is adding rotation to the joint because the arm joint chain was not laid out perfectly on one axis. I added a slight bend where the elbow was to make sure the chain bent in the proper direction when I added the RPik. I did some experiments and found that I don't get rotation from adding the pole vector when I first lay out the chain in a perfectly straight line... so...now what?

Edit Again:
Scratch that I don't know what I am talking about : (
Created several other chains and added pole vectors none of the joints received rotation values.
I've attached the rig if anyone wants to look at it as opposed to trying to make sense of my babbling ;)

mccollom73
11-22-2010, 06:22 AM
I'll take a look at your rig, I sort of understand what you are dealing with, but I'm more visual and need to look at the problem. I'll give you an answer tomorrow sometime.

EDIT: I looks like you have an orient constrain on your joints. Where you trying to get some twisting action going? My conclusion, it may be wrong, you have something affecting the joints when you move it. What I would do, is redo the arm, for one it's good practice, and two it'll make sure that you know that it's not something you did.

Josh

cynicalbug
11-22-2010, 05:29 PM
yea eventually the rig is supposed to have an arm and forearm twist. I'm taking your advice and redoing the arm. thankfully it looks like that was the only joint that had rotations on it. thanks for your help :D

mccollom73
11-22-2010, 05:55 PM
Ya, wouldn't you twist the arm in FK, and just have an orient constraint from your arm controller to the forearm for the twisting of the forearm?

Josh

cynicalbug
11-22-2010, 10:43 PM
Ya, wouldn't you twist the arm in FK, and just have an orient constraint from your arm controller to the forearm for the twisting of the forearm?

Josh

Yea that's kinda what I thought. The tutorial I am using has the instructor creating an extra joint chain that's sole purpose is to allow the animator to tweak the twisting of the arms so that the mesh/skin deforms properly. He claimed the extra joints would also help the mesh maintain its volume and prevent collapse that can be caused by twisting... I thought that with as simplistic of a model as I am working with that I wouldn't need it.

I really like the foot controls and love the spine of this rig setup but I thought his way of rigging the arms was kinda convoluted. Then again, this is only the second human rig setup I've worked with.

CGTalk Moderation
11-22-2010, 10:43 PM
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.