misss tests

#981

Super: i’m not sure, but i think may be increasing the sampling from 64 may help you out.

I’ve been trying to tweak my own character for a while now using this thread for as much help as i can. I am hoping you guys can all help me get this a little more realisitc looking. There are a few spots in particular that i’d like some help with if anyone can.

 1.  ok, his eyes are crazy looking.  I tried to use the SSS shader for the color, with a transparent blinn for the clear reflective part.  Following Krishnamurti's tutorial on 3dtotal, but i'm getting some crazy reflections and tried turning off all the lights linking to it and still they look glazed over.  I tried turning the reflectivity of the colored sss down as well, but no luck.

  1. His skin doesn’t seem real. It’s too smooth or something, i don’t feel like i can place my finger on it. I currently am only using a displacement map from zbrush, and don’t have a bump map on at all. I’m a little confused about that part, since in zbrush there was tons more detail (crows feet, wrinkles cracks, pores) but when imported into the sss shader i seem to have lost a lot of it. if i increase the displacement map he just seems to become bloated looking and doesn’t gain any of the detail. I realize this may not be the place to post this, but thought it might also be something with the SSS shader.

  2. I don’t feel like i’m getting the variation in skin tone that i painted on my color map. all the parts seem to be the same color almost, even though they don’t appear that way on my map.

Here is an image of my maya mental ray render.

this is my zbrush render done using Krishnamurti’s skin shader.

Thanks for any help!

#982

Hey dude, thx for the tip, I’m at College atm, I’ll try that when I get home.

What you might like to do is generate a normal map from Zbrush and plug it into the bump node of the Mi SSS shader. I did that and it brought out quite a lot of detail that is lost in the displacement map.

#983

After reading this:

I realised the actual size of my mesh was too small, so I just scaled it up and viola. I also tweaked the scale conversion a bit. But I seem to be on the right track now. :smiley:

#984

I am using misss_fast_simple_maya for my model.The problem is i want some parts of the model to have transparency and bump map, does anyone know any workaround ( i bump mapped a file but its not showing bumps and the model looks different and for texture i selected few faces and applied another texture(blinn with transparency) but the model is becoming darker).
I really appreciate if anyone suggests a good tutorial for this bcoz maya documentation doesnt have info on this material.

Thank you

#985

I know almost nothing about texturing so please feel free to rip these apart.
I would also really appreciate any tips on making the texture more photo realistic,
as well as any tips on improving the eyes. I really have no clue what i’m doing with those.



#986

does anybody know why scattering is influenced by the scale of the camera? I compared renderings where camera size is 1 and where camera size is 10 and the renderings look different? it’s almost like altering the scale of misss_fast_skin node…
I know that (quasi) raytracing is somehow done with position of light camera position and point to render internally in the misss_fast_skin. Is that right?

does somebody have an idea?

#987

i’ll look into that, thats odd… what about camera distance? is it possible to have the shader look great from close and then when you move the camera back it looks bad?

#988

delarge…I’ve encountered a problem when rendering my character near and far…did you run into the same problem?

…when rendering near… http://www.geocities.com/lorely3000/sss_near.jpg

…when rendering far… http://www.geocities.com/lorely3000/sss_far.jpg

/lorely3000

#989

new misss test

#990

loooking good SilentRage. maybe soften it out a little bit, the bump feels a bit harsh in some areas. what is your light setup like? are you just lighting it with an HDRI?

#991

It’s HDI and 1 spotlight.

Yeah the bumpmap itself is not very consistent in terms of sharpness.
I’ll probably redo the texture later.

#992

Hey guys, thought maybe I would toss what I am working on into this thread. I am an animator, so all this shading and rendering stuff is black magic to me, but here is a characters head from a personal project. It’s much more stylized then most of the work in this thread, but since I am using the misss shader I figured why not.

 I am trying for a more cartoony look. He is supposed to be a kind of Nosferatuesque guy.  A little bit of detail here and there, but more simplified over all.  Let me know what you all think.

  

 Grim
#993

I think instead of just the rims of the ears having a little translucence, the inner portion which is very thin as well, should have some redness too.

Also while I can’t base this on any real reference because I havn’t seen a nose that thin, I think it should have some redness too due to the tip and wings being very thin, and that there’s no solid bone underneath to produce whiteness like the chin has.

Hope that helps.

#994

Hi

this is my contribution:

#995

Thanks for the input SilentRage. So here is my neophyte render monkey question. Is there a way to some how paint the various scatter radius and have the lower and upper numbers bound to meaningful numbers in my scene. If I start tweaking stuff to get the inner ears to have more translucence I got lots of scatter around the mouth, eyes, top of the head etc. So can I paint like a grey scale map and have black represent a base scatter radius and have white be the higher one. In all the maps I know of black = 0 white = 1. So how do I change it so that black = the minimal scatter radius I want and white = the max I want? Is that even possible or a good course of action?

Thanks,
Grim

#996

that is exactly what you do, put in a grayscale map where brighter values represent areas you want more translucency and vice versa.
Now before I say anything I have to declare that i’m in no way even remotely expert at this shader so I’ll just go off of whatever info I know by trial and error.
I believe you just plug the grayscale map into say Back Scatter Depth, then tweak Back Scatter Radius, which can be used more or less as a global control to scale how translucent the white in your map is going to be and how solid the black will be.

also it looks like with the thickness that the tip of the nose has it should already be more translucent than it already is, maybe you can try putting a higher value in the global ratio, then a higher value for scatter depth and radius.

#997

@ ZippZopp:

Hi Peter,
I am very curious of your model scale in the viewport. What is the relation to viewport grid? I though you need to crank the algorithm - scale UP for huge model, but instead you taking it down to 0.3. I think this value is crutial before trying anything else.
One more thing, I saw facing ratio used in the network. Where does it plugged into? I must’ve missed it if it is explained before :smiley:

My 0.02 cents tips on SSS:
When you had Bump2D node attached to misss bump slot and you decided you want to use bump_combiner or something else down the chain then you disconnect it from misss shader, pay attention to the value left after bump2D disconnected. By default it was 0 0 0 means no bump at all, but after disconnecting bump2D it became 0 0 1 which gives you jagged shadow on renders. I spent half day try to solve it and end up rebuilding my network.

Ft

#998

that orc, where the scale conversion is really low, is huge in the scene. using the distance measure tool, he is 200 units tall. much bigger than i usually work, so thats why my value was very low.

facing ratio had been used just on the cornea shader of the eye. it is really just to get a fresnel effect on the reflectivity of the eye. it is slight, but a little more reflective at glancing angles.

#999

I have a question (I think a dumb one). Did you hand painted all those maps? Or are they the result of the individual layers of your shader rendered? It would really help to know!

#1000

@ LiquidSnake: i know you asked ZipZopp, but i’m willing to bet he only painted the bump and color maps. the other maps are probably just color corrected and combined versions of those. that’s what i do in mine.

@ Jester: i can’t believe nobody has commented on your image yet. it’s really incredible. for me, there are two things that are a dead giveaway that images are CG though: the hard edge around the eyes, and the lack of the dust/fuzz layer that catches light on skin and fabric. being a perfectionist, i would look into that if you have time. if you haven’t already, submit that image to the Character Modeling 2 submissions at Ballistic. it’s really great.