View Full Version : Cleaner mesh = better Claude Bonet map
Cactus Dan 02-07-2004, 10:43 PM Howdy,
I've been racking my brain trying to get the Claude Bonet weights painted properly so that the shoulder's deformation behaves. Nothing I tried seemed to work, so I got to thinking that maybe the problem is in the topology of the mesh.
Well I did a little experimenting on one arm of the mesh and this is the result:
http://home.earthlink.net/~cactusdanl/Cleaner.jpg
The topology of the upper arm seems to work better when the top of the edge loops lean toward the shoulder.
Now, if I can just remember everything that I did to the one arm, I can fix the other one, too. It would be helpful if I could place the curser over a point and get a digital readout of what the CB weight was, or be able to sample a point and have that value plugged into the CB strength setting.
Adios,
Cactus Dan
|
|
JamesMK
02-07-2004, 11:11 PM
Interesting stuff!
I guess I came to just about the same conclusion back in my Blender days... after several less than successful attempts with various approaches to weightpainting, it finally became clear that if each part of the mesh isn't properly modeled for the particular deformation that part is supposed to be exposed to, one can paint weights until hell freezes over - it won't work very well anyway.
On the other hand, with a perfect mesh, there's rarely any need to weight it at all (except for those obvious problem areas around the crotch of course, and a few others).
dsavage
02-08-2004, 01:58 PM
Nice example Dan. I seem to remember somthing about this on a Maya tutorial I found some time ago. Something along the lines of the edge loops being closer together on the outside of a bend (elbow, shoulder, knee) and further apart on the inide of the deformation, though I've no idea where the tut. is, so I could be wrong.
Cactus Dan
02-08-2004, 03:59 PM
Howdy,
Originally posted by dsavage
Nice example Dan. I seem to remember somthing about this on a Maya tutorial I found some time ago. Something along the lines of the edge loops being closer together on the outside of a bend (elbow, shoulder, knee) and further apart on the inide of the deformation, though I've no idea where the tut. is, so I could be wrong.
Hmmm..... interesting because that seems to be exactly what I did (although I just stumbled onto it). I'd be curious to see the rest of that tutorial. Too bad you didn't bookmark it back then.
Adios,
Cactus Dan
dsavage
02-08-2004, 04:56 PM
Dan, I'll have a look for it - it was a link from one of the threads here a couple of months back. I'll let you know if I find it.
dsavage
02-08-2004, 06:57 PM
Dan - good and bad news. Good news is I found the tutorial - www.hippydrome.com.
Bad news is it's the opposite of what I said, and also what seemed to work for you i.e. the tutorial suggests edge loops closer on the inside and further apart on the outside of the bend.
Maybe it's a case of 'whatever works for you'.
By the way, I've been following the development of your cowboy - coming along very nicely. Your face morph work was particularly interesting.
Cactus Dan
02-08-2004, 07:17 PM
Howdy,
Thanks dsavage. Maybe the shoulder is just a special case, or it may have something to do with the fact that I'm trying to simulate the clothing (there's no modeled arm under the shirt sleeve). I don't know, but anyway, I'm going to check out that tutorial and see what I can learn. I've noticed that it doesn't matter what software it's for, as long as the tutorial is good, you learn something (as it was with Jason's book).
Thanks again.
Adios,
Cactus Dan
flyingP
02-08-2004, 07:50 PM
Originally posted by dsavage
Maybe it's a case of 'whatever works for you'.
Or just a case of whatever works .
Cactus have you tried moving this in other directions as well ?
Cactus Dan
02-09-2004, 12:11 AM
Howdy,
Originally posted by flyingP
Cactus have you tried moving this in other directions as well ?
Yes, and I need to make a slight adjustment in the contours so that not only will their tops lean toward the shoulder, but the backside should also. I figure that the arm would have more forward motion than backward so that would work out better.
I went through that tutorial that dsavage posted a link to, and found that I was at least partially right in what I did with the contours. Not only should the contours be tight on the inside of a bend and spread more on the outside of the bend, but they should also be angled toward the root bone on the outside of the bend and angled away from the root bone on the inside of the bend. The tutorial pointed out that this kind of structure helps to define the muscle movements also. Very interesting and informative tutorial. I've added that one to my bookmarks.
Adios,
Cactus Dan
Hey Cactus,
I know you already know this, but in addition to the topology of the mesh, there is some importance to the placement of the bones themselves. This article at Gamasutra really helped me out:
http://www.gamasutra.com/features/visual_arts/19981113/charmod_01.htm
HTH
Gary
wuensch
02-09-2004, 07:57 AM
Hi, cactus-- actually the maya tutorial shows quite a similar shoulder construction.
I totally agree. one of the beginner errors I had to learn was that the single most important thing to good deformation is a good layout of the mesh-topology.
Olli
Cactus Dan
02-10-2004, 01:13 AM
Howdy,
3dg: thanks for the link to that article. Yep, I have noticed (through trial and error) that bone placement is important.
Olli: as long as I learn from my errors, I reckon I'm doing OK. Although, I keep wishing I had known some of these things when I started this project. But even so, sometimes the investigation into solving the problems I encounter, is a rewarding experience.
Adios,
Cactus Dan
CGTalk Moderation
01-17-2006, 09:00 AM
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.
vBulletin v3.0.5, Copyright ©2000-2012, Jelsoft Enterprises Ltd.