Topology - Check

CreativeSheep · Nov 2015

If someone can take a look at the following image and, although it's not complete verify if the topology is adequate for rigging deformations ?

I may have to do a major adjustment to the design, I'd still like to have some comments whether things are in the right direction and proceed as I am; or not.

Image: https://us.v-cdn.net/5021068/uploads/editor/bw/pp11xkwqzuml.png

JustMeSR · Nov 2015

It looks like quite a mess to me. Is there a reason for those insets and redirected edge loops?
I recommend looking at the wiki for some limb topology (even if your mesh isn't limb, it should use same topology if it deforms the same way).

CreativeSheep · Nov 2015

There is a reason for the insets, not to make it a cylindrical object; that and I did a slight change to the topology. That is some what debatable, the link you supplied.

skankerzero · Nov 2015

it would help if we had some kind of context. I can only assume this is a limb, but which one exactly?

CreativeSheep · Nov 2015

Limb of a frog; it's for high polygon; although I may have to model it in a rest position, rather then similar to a frog-T-pose.

Noth · Nov 2015

when in doubt; bendy-straw!

http://wiki.polycount.com/wiki/ShoulderTopology

CreativeSheep · Nov 2015

I don't want straw type modeling and no formation, that is what the bendy straw method is portraying, edge loops, which I know are required, some termination of some loops which cause some triangles, yet no formation. I'd like to model with no triangles, there is no way to terminate a edge loop unless you make a triangle, if you delete the edge that makes the triangle you get an n-gon, it's like a roundabout.

CreativeSheep · Nov 2015

Here is the same leg, vanilla skinning; IMO it bends well, except there is the hose effect, whereas the mesh deforms as if it's a garden hose. Anyway to model this whereas some simple weight painting would fix at the bends ?

I know there are sophisticated rigging methods to which I could use, I want to keep it simpler; sometimes those rigging methods are not reliable or don't directly fix the problem.

Image: https://us.v-cdn.net/5021068/uploads/editor/bl/uzo6lt3hmhoy.png

Eric Chadwick · Nov 2015

You just need to adjust your weights. I don't see any inherent problems with your topology. The indentations in your rest pose might not match a real frog leg, but maybe your target style is different from normal?

CreativeSheep · Nov 2015

That leg is just a proxy leg, I want to get an understanding, it would need more formation then what is has, and I would be shooting for a almost, if not realistic frog leg so in that arena much work would need to be done. Since this is a proxy not a concern

Could the bent joint areas; not be modeled in a way whereas a adjusting the weights couldn't, maybe not directly solve, but close ?

Compared to my first screen shot and my second, is there an improvement as in, cleaning up the topology helped ?

I ask because I wonder how far I can go with formations to the topology before I run into joint issues; as the above image may be good with regards ready to animate or expand on the rig, but should I keep it simpler where there will be joint deformations, for example, without going into the correct terminology of a frog anatomy, if the top of the frog leg were a thigh, one could do some bevels on the thigh for whatever, but keep where the frog leg bends and deforms simple, with the exception of displacement maps, which is a different topic based on as mentioned, the expansion of the rig ?

Eric Chadwick · Nov 2015

Omg, what a convoluted post. It's very hard to understand what you are asking.

If this is only a proxy, then why do you want critiques?

Do you want to weight a lower-resolution control mesh? Max allows this, with Skin Wrap, not sure about Maya.

Look again at the limb topology page. The simplest topo there, but you can add more detail as needed.

CreativeSheep · Nov 2015

Alright, two questions at a time from my previous post;

Could the bent joint areas not be modeled in a way whereas; adjusting the weights could maybe not directly solve, more or less, understand ?

Compared to my first screen shot and my second, is there an improvement ?

Eric Chadwick · Nov 2015

The first question makes no sense. Remove the double negative, and simplify the question, make the smallest sentence that is possible. Maybe you mean: "Does it have topology which cannot be solved by weighting?" No, it does not.

For the second question, it is difficult to see. They are in different poses. It would help to show an animated gif of the bending motion. Try GifCam. Make something like this, but using your model:

Image: http://wiki.polycount.com/w/images/d/d1/BenMathis_limb_deformations.gif

skankerzero · Nov 2015

Also, don't be afraid of triangles in hinge joints.

CreativeSheep · Nov 2015

I'm not modeling for low poly, rather high. I'm only showing the low poly mesh because, it's easier to skin and show formations then increasing topology and skinning.

Maybe you mean: "Does it have topology which cannot be solved by weighting?" No, it does not

Then, how can I model so I solve most, problems though weighting ?

Should one avoid beveling regions if it hinders bending joint areas ? As I mentioned, add bevels to the thighs, not to the knees, keep the knees simple, as in edge loops etc ?

Eric Chadwick · Nov 2015

Then, how can I model so I solve most, problems though weighting ?

The wiki Topology pages show exactly how to model for good weighting. I don't understand how this is not clear.

Bevels and edge loops and high-res modeling all build on top of these basic topology rules.

If the core lowpoly is correct, all the rest will weight properly. You do need to learn how to rig and weight though, that's another set of skills.

Topology - Check

Replies