fast skin and final gather problem BIG problem.

[Butachan]

Hi I am having a big issue here that i don't understand what is the problem.
When I use Fast_skin_misss material without final gather is ok but in the moment that I turn final gather on. this error appears:
// ERROR : PHEN 0.5 273 MB error: Internal error: misss_call_shader: Unexpected state->type case.

the image renders fine! there are no visible errors, but the render is painfully slow. when the error appears.
there seem to be a similar case in 3DMax http://area.autodesk.com/forum/autodesk ... or/page-2/
I am trying to accomplish a sss skin similar to pixar's (with MR) but a bit more flat I believe the fast skin was a good approach and the result seem very promising... and it flourishes with final gather. But the error would render the shader useless if it takes so long to render.
Does any one knows about this issue?

[Kerro Perro]

If i remember correctly it's a bug in the "packaged" skin shader compound and if you use the some of the others it won't happen - but i have forgotten wich ones you should use instead and can't quite remember where i heard it. I know it's one of those big messy ones plus something else....

Sorry i couldn't be more precise...

[Butachan]

Hi Kerro Perro,
Any answer is helpful, so thanks for the explanation. However I doubt that is an internal bug unless all the other material nodes, including mi phenomena nodes also have the same bug. For the moment I can confirm that the problem happens when using transparency in the scene. Once I hide the corneas (basically transparent balls) the error dissapears and it renders wonderfull. Nobody else had this problem?

[Mathaeus]

well if you already localized the problem, perhaps you could supply some another cornea shader for final gathering. For example using RayType shader switcher, 'enable FG port for final gathering', put something opaque into FG port. Or just disbale Fg for cornea, if it's fully transparent. It might help, theoretically - just personally, didn't played with mi_sss stuff for a long time.
Definitively looks like transparency invoked the additional computation of mi_sss stuff, together with FG. As long as subsurface shader rely on some pre-computation phase, it's matter of convention, what is enabled in this phase. You don't want costly feature which contributes small or nothing. If you'll be able to fix it in this way, I'd believe it's more like inconsistency in shader system instead of bug.

[Butachan]

Matheus: you are an angel! indeed the raytype switcher made the trick. I basically connect my previous solution to all the ports except final gather and then give a solid color 50% grey to Final Gather. I know that is fruitless to complain about mental ray on the subject, particularly since it has soo many enemies already. But I am quite surprised that a phenomena they themselves create is incompatible with a basic state of oppacity less than 100%. It never cross my mind because is rather ridiculous to create a shader like that.

[Kerro Perro]

Ah yeah sorry i might have been confused with what i got from an Alex Alvarez tutorial where he used a seperate blinn node for the specular, it used to be on the Gnomon site for free but it seems to have been taken off...

I actually opened up the last scene i used sss with yesterday and i did notice a really long pre-compute time before the actual rendering as well.