PDA

View Full Version : Using a Null in a Bone Hierarchy



WilliamVaughan
10-14-2008, 04:53 PM
Using a Null in a Bone Hierarchy:
Right-Click Here to save (ftp://ftp.newtek.com/multimedia/movies/LW_9/Layout_NullBone.mov)

jin choung
10-14-2008, 07:59 PM
interesting... but WHY?

primary thing i can think of is because of aesthetics in the viewport - frequently in maya, when we do bone driven face deformations, the head becomes a mess because all the joints on the face are visually connected to the head, creating a crazy radiating spike ball appearance that can confuse people having to work with it.

but is there any other purpose besides a cleaner viewport presentation?

cuz a connected hierarchy would have the shoulder move just as you have in the video and while you say that there are reasons to have a disconnected hierarchy, you don't say what they are (except the thing about terminating ik but again, since a method to terminating the chain exists, i'm left wondering why).... and i can't actually think of any.

i understand that with lw and blender, being able to have bones parented but not visually connected allows you to work in a certain way that is kinda jarring when you go into something like maya that requires a visual connection.... but i learned pretty quick that there's not an actual functional difference (actually, there is with blender - it's kind of a halfway point between lw and maya).

so i don't get it.

jin

Castius
10-14-2008, 10:21 PM
Because LW joints are different. That connection can actually deform the mesh. It's not just a visual representation of the hierarchy.

If you want a really extreme example would be if you need to rig two characters with one mesh. You could put each joint hierarchy under there own nulls.

So on one hand it's an extra step because joints can do more. On the other hand nulls can be used with zbones as well to to zero out transforms. This way you don't have to worry about disabling extra bones. LW null bones can now just be nulls.

jin choung
10-14-2008, 10:42 PM
If you want a really extreme example would be if you need to rig two characters with one mesh.

?

when would you do that?

also, even if you did rig two separate skeletons inside a single mesh, whatever you could do with a disconnected mesh can also be done with a connected mesh. as long as shared joint is set to exert no influence.

again, if the answer is just viewport aesthetics, that's reason enough but it sounded like there would be more to it than that....

also, every joint in maya can be set to deform as well and is set this way by default. it IS a visual representation but every joint exerts a pull in terms of deformation... so i still don't see the point.

jin

Castius
10-15-2008, 01:57 AM
You know better than anyone that you're comparing apples to oranges. Maya you can just leave out joints you don't want in the skinCluster. The joints you do use require 100% weight mapping. They are just a transform for a weight map.

For LW Joint connections are more then aesthetics. If you're not using weight maps only. That joint connection represents where the joint is effecting the mesh. So if you are using joints, nulls can be used to limit that deformation.

I'll give you more more example. Take Protons video and lets say that the arms of his character are disconnected and are floating out side the body. You can see in the image that on the left. Moving the joint only side takes into a account the rotation. Where as on the side that uses the null does not.

I could disable the joint in between or set it's strength to 0. But now we have another way to handle this.

jin choung
10-15-2008, 11:31 AM
You know better than anyone that you're comparing apples to oranges. Maya you can just leave out joints you don't want in the skinCluster. The joints you do use require 100% weight mapping.

no they don't. every vertex needs to have a total of 100% weight but a bone itself can be affecting the mesh at 0%.



For LW Joint connections are more then aesthetics. If you're not using weight maps only. That joint connection represents where the joint is effecting the mesh. So if you are using joints, nulls can be used to limit that deformation.

I'll give you more more example. Take Protons video and lets say that the arms of his character are disconnected and are floating out side the body. You can see in the image that on the left. Moving the joint only side takes into a account the rotation. Where as on the side that uses the null does not.

I could disable the joint in between or set it's strength to 0. But now we have another way to handle this.

that's what i mean - your last sentence says it all... we already have a way of addressing this. same with the ik stop that proton mentioned.

and it's easier and faster using the old ways!

so it doesn't seem to buy us anything in terms of *functionality* that we didn't have before.

it's another way of doing something... still, the only new thing that registers is aesthetics - if you want to use joints so they visually look like our zaxisbones or prevent a spike ball mess in complicated skeleton setups for something like a face - which is valuable but then that's the functionality that should be noted.

jin