PDA

View Full Version : Omega SSS problem



sammael
02-02-2007, 10:24 PM
Hi all, I have a problem with my current wip, when I use the omega shader on the model you can see the inside of the mouth through the head. I am using a weight map to control the sss amount and penetration, even when I set the value in the weight map to '0' in this area the inner mouth still shows through.

I had had this problem in the past when using image maps to control the ammount and penetration but I got around it somehow (I forget), it was temprimental and slight adjustments of settings would cause the error to occur again.
I want to make this new character with 100% procedurals and gradients no image maps.

I wonder if anyone else has had this problem and knows how to get around it?
I also have a problem where you can faintly see the wireframe of the model.

Here is my surface setup for the skin, I have tried to set it up as it states in the documentation.

Hope someone can help, cheers!

sammael
02-02-2007, 11:20 PM
Heres an image of the head as well so you can see the problem.
Its a shame this stuff happens because the omega shader is great, I realy like the results minus the render errors.

Werner
02-03-2007, 12:58 AM
Just give it a try with air polys and leave the spot info nodes out. You should use the spot info node with a logic node on transparency. Do a search here on the forum...someone posted the correct way of setting this up.

If that does not work, the SSS distance might me too much, thus seeing the inside of inner geometry.

sammael
02-03-2007, 01:33 AM
Thanks Werner but if you mean this thread it does not realy make much sense to me. I get the exact same result following this method to the best of my understanding.

I have not used air polys before, does this mean I would have to create a whole new mesh within my existing mesh and apply appropriate subsurface material settings? How would you recommend avoiding the inner mouth showing through using the air polys method? those polys would still be inside there and I cant see why they wouldnt show up using this method as well. I will try it and see what happens.
I would rather avoid adding substantually to my poly count though and thus render time. At the moment I can render 1 frame at 600x800 in under 3 mins, im pretty happy with that.

sammael
02-03-2007, 02:55 AM
I must say, implimentation of this node and others could be improved by about 10000%
how about a system for artists instead of programmers? all this crap you have to do to apparently make it work and then it does not work anyway. How about an option within the node itself where you specify double sided polys or air polys? None of this confusing BS with multiple other nodes of which the purpose is unclear at best unless your some sort of nerdy genius that understands this rubbish. No offence to people who do understand the purpose of all these individual nodes... I envy you.

Simpler is better NT

sammael
02-03-2007, 09:45 AM
ok I fixed this problem, it turned out to be the scalar node plugged into the oren nayer diffuse shader. My appologies for my outburst earlier LW can be fustrating at times.

serge
02-04-2007, 06:38 AM
My appologies for my outburst earlier LW can be fustrating at times.
No need to apologise I think. The lack of an artist-friendly node-manual is frustrating for most users I bet.


I also have a problem where you can faintly see the wireframe of the model.
Yes, it's a bug. Link (www.newtek.com/forums/showthread.php?t=57808)
It has been reported in the Beta-forums.

sammael
02-04-2007, 07:21 AM
More artist friendly tools, less science geekery, please! By all means include both, but I don't see why things need to be so overly complex when others can make such elegant solutions.

Agreed!

Werner
02-04-2007, 02:07 PM
well, I have to agree, but I'm just happy that you fixed the problem.

:D