View Full Version : Lattice/Bone Blend
10 October 2002, 10:07 AM
So about a week ago I was all excited about lattices and there ability to animate an organic form on the fly (no premade morphs). Today I successfully combined the lattice (which is composed of bones for every point in the defined object) parented the lattice bones to my skelegon drawn bones, controlled the capsule deformation with weight maps (using the weight map only option) and made a happy capsule flexing its muscles! Without use a morph, no expressions fuss, and can be manipulated intuitively. Aah Yeah!
Chech out the simple avi. http://www.stoehrsartstudio.com/Animation/lattice_boneBlend.avi
Now I'm trying to use IK with it, but I'm having trouble. I think there are too many bones in the lattice conflicting the influence of the IK on the mesh. I even tried a goal stregth of 5000, but that didn't override the jittering and shaking.
10 October 2002, 11:05 AM
Sounds to me like what you did was more work and fuss than expressions or morphs...
In any case, with not really knowing about what you are doing, here's an uninformed idea I'm throwing out there to see if it sticks. A little brainstorming.
What if you have the IK bones NOT affecting the mesh at all, just the lattice bones. And the lattice bones parented to their respective IK bones, as you said. So you IK the lattice, not the mesh. The lattce bends the mesh, which you can then tweak with the lattice. Does that make sense?
Is that what you're already doing?
Hmmm...now that I write that out... it does sounds kinda cool. But maybe really slow for animating with all dem bones all over the place.
Keep up the good work and let us know how your experiment turns out! :thumbsup:
10 October 2002, 11:26 AM
Actually, it is a little more work when animating all of those bones in the lattice to get accurate deformation. The advantage is the ability to breath life into each movement, intuitively, and each being slightly different. Whereas when using a morph linked through expressions, the morph is "cloned" everytime time, say, an arm bends, which can look unrealistic over a period of time.
Ya know, now that I'm writing this out, I realize I can use morphs to compliment this setup. As a way of "roughing out" the deformation. Aah Yeah! I'm on it now!
To answer your question, Yes, the lattice is parented to the bones, so I animate the IK/FK bones first, then go through and tweak the lattice for tiny movement or muscle like deformation. The only real challenge in this setup is accurate weight maps, As always, and spending the time to do it. Oh yeah, and proper parenting.
10 October 2002, 11:29 AM
I'm not using the weight map only option. It tends to conflict with all of the other bones in the lattice. (it multiplies the effect, effectively blowing up my model like a ballon when I bend it). But I am using a weight map, just without the weight map only option.
10 October 2002, 09:02 PM
So I have this working with IK now. Yea! The only catch I my setup, so far, is I need weight maps for every skeleton bone (not the lattice, their automatically created), and the weight maps all need to be Zero percent influence. this allows the lattice bones to influence the mesh, otherwise the skeleton bones are too big and exert too much influence. So there has to be a weight map for the mesh to deform properly and "communicate" from the skeleton bones, yet zero influence so the lattice bones can effect the mesh as well.
Because a mesh object can only except bones from itself or one other object all bones need to be in the same layer, therefore effecting the mesh proportionately.
This brings me back to my original question a week ago. Is it possible to develop a lattice sytem that is not dependent on bones? Can Lightwave be changed enough to except another way of deforming a mesh other than using bones or morphs? The methodology I'm developing ACTS like another way of animating, which is extremely intuitive, yet ultimately, is still based on bones, which creates conflicts between weights.
Though I'm still working on it, I'm hopeful for some powerful results.
01 January 2006, 09:01 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.