Skin Shading Workflow

[shushens]

Hello Softimage Lovers!

I have been reading a lot of articles on skin shading in Softimage in various blogs/forums. A lot of people have said a lot of things and a lot of meaning has been lost in technoblabber. Being new in skin shading, a lot of it did not seem to make sense at all. A lot of the posts are also pretty old, so it is difficult to figure out which problems have been fixed since and which haven't been.

I am trying to make skin for the head I modelled here:

viewtopic.php?f=24&t=517

Instead of endlessly experimenting with the shader with the render region, it would make much more sense to get all the passes rendered out so that they can be comped back together in post. If all light, specular and scattering contributions can be rendered separately, then life can get a lot simpler. After looking around for quite a while, I found the Split SSS fast skin v2:

Code: Select all

http://sjannuz.com/sj_sss/sj_sss_fast_skin_lightmap_v2.spdl
http://sjannuz.com/sj_sss/sj_sss_fast_skin_material_v2.spdl

along with some screenshots here and there. But no proper explanation of how this really works.

Watching the Gnomon DVD, where Ben Neall uses the Nexus shader to render out all sorts of possible passes almost effortlessly, I am very much curious to find out if any smart method of isolating light contrbutions and scattering contributions exists so that one can tinker with them in post and find the skin composition that looks best.

Any suggestions, guidance, thoughts would be very much appreciated.

Thanks in advance.

[Hirazi Blue]

As far as I recall (but my memory isn't all that good ) the split_sss shaders stopped working with XSI 7. At least "mine" did. (Not really helpful, sorry! ) I have stopped my own struggle with skin shading a long time ago, but IMHO you're doing just fine with the eman's "Alternative" in your "Human Head Experiment"...

BTW - There once was a "skin shader workshop" on the German Softimage Forum, which sported some quite impressive results. Might be an idea to shamelessly copy the idea and do a similar workshop (in english, obviously) here...

[shushens]

Very helpful post, as always!
Since you speak German, (evidently) you should be the one to host the workshop
I will try to observe the screenshots and see if I can gather something from it.
But not being able to read the text is a huge disadvantage, needless to mention.
If you attempt such a workshop, I am very much in I can run tests round the clock
But, oh well, so can probably every other Softimage lover

One of my big disadvantages are, I tend to get too vertical with things and forget to look around. So there is often a lot going on in the scene that I miss.

-------- EDIT ---------

I looked at all the caps.
I have known Oglu long time from CGSociety. He is quite a Mudbox addict. Helped me with Mudbox maps in Maya/XSI many times before. Doesn't use Softimage any more though. He told me he quit. I am kinda sad to say this, but I think compared to average Max or Maya shading artists, the guys who worked on this seem... unmotivated. Naturally they could do better than this. But seems like they just felt like it's enough and dropped it. But of course, I can't read the language.
What do you think about this?

[Hirazi Blue]

Well, you're probably right, but this is in the true spirit of how this works in most forums I'm afraid. Most people, if they even do participate, hardly go "all the way". The stuff "bt3d" AKA Tim Borgmann is the best contribution in that thread...
If you look at it as "final production work" it obviously doesn't work, if you view it as "let's have a community workshop, see where it goes", well that's a different story...

BTW - That workshop was held very, very long ago, when working with the SSS shaders was even more daunting than it is now...

Moderator edit: the following posts about color curves have been split into their own little thread - HB