Model Structure

[ArchangelTalon]

I'm pretty certain most of that is done via intersecting mesh. Except possibly the long ridged areas, the might well be part of the jacket mesh and not separate elements (although there is quite a harsh texture change at the joins... it's hard to be sure). But note that all the parts that are separate mesh are on areas where the model doesn't deform through animation.

Having separate elements is fine, really, but there are instances where it should be avoided and heavily deforming areas of meshes are one of them.

[Buzzy]

Yeah we use intersecting geometry all the time...its nearly impossible to avoid if you want to keep tricounts low. Cutting that eyeball into that guy's hat would probably nearly double its tricount, and for no visible difference.

But, as was mentioned before, that's fine for something like a hat.....but if you are working on a shirt or something else that deforms alot, its best to try and avoid it.

[kyuketsuki]

Thanks for the information. I figured as much and I knew about the deformation issues, but I needed to make sure from some industry people, so thank you.

[Swift3D]

I would say the ridges and studs in the bottom left are attatched and everything else you've highlited is penetrating.

The way I try and tell is if the loops that pass each other have the same number of edges if that makes any sense, for example on the studs you can see both the base of the stud and the section it's plugged into have about 12 edges each and are lined up perfectly.

to me this looks looks like a [Extrude>0, Scale>.75, Extrude>1.5, Scale>0] job.

Also to me the inside ot the ridges look like they suck in a little bit leading me to believe that they're attatched.

just my opinion, probably wrong :p

[SpiralFace]

a quick note though.

Depending on your game engine, there are instances where you will want to avoid intersecting meshes. The biggest probably being if your engine bakes its lighting information on a dynamic per-vertex basis instead of a per-pixel basis. If you have buried verts in a per-vertex lighting engine, then your going to run into issues with the models as the lights won't ber able to hit your verts that are buried, and you'll run into lighting issues.

This is more of a last gen concern, as ANYTHING with a normal map on it must be rendered on a per-pixel basis, but it does crop up.

Also, some engines don't like open meshes. So sometimes, even if it is wasted geo, your going to need to keep faces buried on all the little bits so it doesn't give you issues in engine.

Just some food for thought. If your just doing this for some student studies then you don't need to worry about these things, but just keep in mind that they are problems that can come up, so its best to at least recognize and be able to deal with the possibilitys if they ever crop up in your actual work.

[CGTalk Moderation]

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.