ZBrushCentral

Texuring skins .. multi layered maps.. whats best..

Hey,
I’ve been a modeler / animator for nearly 10 years… but people / human-like models are a new area of approach for me…

I’m nearing completion of my first human figure, and about to start the texturing phase…

I was wondering how to get the best skin textures…

photo references?.. painting detail by detail?

multi layered maps projected onto it?..

Not so much the how to apply it in zbrush…

but I’ve seen several examples where multiple textures were painted on then composited onto eachother via multiple render pass’s

Is this the best approach?

I’d love to see some broken down examples…

Painting a singular skin texture seems very limited… but im not sure how to acheive a realistic look to skin textures… even my projected photo references never seem quite right…

Love any insight into this you talented folks might provide!

Thanks

-Blair

Ahh the art of CG skin, this is somthink im working on myself, here is some of my personal insite so far.

Projecting photo as textures can give a realistic look but in a limited eviroment. Essencialy using a photo as has light, specular, diffuse colour, subdermal epidermal layers are baked in. This all looks well untill you put this texture map into a scene that has totaly different lighting and angles. Of course that dont mean you cant do some magic in photoshop to remove somes spec highlights shadows e.c.t.

The maps im using at the moment comprices of a basic Diffuse colour, this is just simple skin tones and if needed eye brows.

Id then make a bump map which I turn into a specular highlight map for sharper details.

With the diffuse map I make a 2nd spec highlight map, this would be the 2nd pass which is the larger surface area, for this map I inverse the diffuse map. I would alter some values to get this right.

Then once this is done, id paint, or over saturate a diffuse colour map to make a Subdermal map, being I dont have Max, I use this Map in Modo as my SSS colour map which also includes veins that look good for a back scattering effect, like the ears.

I then create a SSS weight map that controls the SSS amount, id make this from a diffuse map,this of course is grey scale.

Then the very important step is scene lighting, this will make or break the quality of the skin. Getting the right ballance between the maps and lighting taks quite a alot of time, and learning this whole process is what im doing right now.

Id be ineterested if any pros out there can shed more light on this subject.

This might be as good a resource as any to get you started (depends quite a bit on your choice of rendering engine of course also…):

http://www.lamrug.org/resources/skintips.html

good luck :slight_smile:

C.

What brings frustration to many people is not the creation of the hand full of maps needed to create the effects to simulate skin but the fact that this is one of the rare times where scale becomes a critical factor. Most artists do not care if their sofa model is half a mile big in terms of the render engine, or only a few micron. Normally that is not an issue at all as the proportions matter more than real size. But certain phenomena require the static adjustment of things like adsorption depth and other values. If results are not taking place as expected, most often than not it is the wrong setup to blame, not the quality of the maps. So one has to keep in mind that the ‘dimension less’ world of modeling is not dimension less any longer once physics engines, certain rendering phenomena, and objects like IES lights introduce value ranges which have to be accommodated.
Cheers
Lemo

That is so true about the scale, Biggest mistake I made was forgetting to scale it correctly. In Modo it alows me to still take them values up, but when your adding meters instead of milimeters is a good sign that somthng is not right.

Another thing to add is setup your lights first. Have your scene setup with all lights in place before playing with SSS and maps. Nothink is more anoying than perfecting the skin values to then think, ohh thats add a few lights and turn on GI, hours of tweaking them maps and SSS ballance is gone i the wind.

I was quite successful in setting up rig’s driving parameters while changing lights. The distance to an object for example can be used to drive shader characteristics and so on… Tricky to set up, but then allows animations and not only the static image.
Lemo

Iv found in Modo it varies too much. I had spent some time setting up a sss last night and was going to add a few lights after but forgot to scale down the mesh. When I looked in my scene the physical size of the lights where so small that I had trouble finding them.

I scaled down the model but there where just too many paramenters to adjust to get what I had setup before. Mind you when I say setting up the SSS I mean with all the maps that produce the skin, just moving a light seems to upset the outcome between looking good and looking awfull.