Hi, I’m new to this thread. I’m glad it’s still active!
flouie – I’m having the same kind of dilemma with the peach fuzz issue, although I did find a work-around. I’m using Electric Image but the idea is universal: The idea is to create a series of fall-off maps like the one you mentioned and composite them in Photoshop (or Shake, After Effects, or Motion) using different transfer modes.
I used the reflection fall-off with a 100% white relfection map around a spherical projection which was all reduced by about 60% and subtracted with a noise procedural. I also created another shader with a more severe diffuse fall-off around a pure white color map to use in “Multiply” mode at about 20% opacity to modulate the underlying render.
I know I’m jumping forums with this link but here’s what I’m talking about if you’d like to take a look:
Thanks Harryb.
I like the results around your model’s jaw. Its good to see someone with a workaround.
However, going into a compositing application is beyond the scope of my project. It’s a little over my head too. I’m looking for a way to get a better peach fuzz to work within XSI 5.0. Something of a shader based work-around, if that is possable.
flouie, for a cartoon project this technique is quite good, but not for realistic projetcs. I’ve spent more than two years with workaround shaders when I’ve realized, that if I’m successfully created a huge shader network, it was killed immediatley by the slightest change in the light rig. The fast skin shaders in XSI are good option IMO, and worth the time factor.
Peach fuzz. I used to make this with a mixer node, where I mix a camera incidence node driven gradient with an ambient occlusion or a light incidence node. This will mask out the areas where the light won’t get in. In addition, I’d connect a texture of tiny hairs into the gradient. Give it it a try.
Thank YOU flouie. I’m sorry you’re not using compositing this time around. I used to think that the 3D program was the main tool for any 3D illustration or animation, but over the years I’ve come to realize that the compositing program is THE weapon of choice in the CG artist’s arsenal. Most VFX houses and CG studios render out in passes and rely heavily on compositing. And those that don’t are coming around to this way of working. Here’s why:
These days, most 3D programs used in production can render multiple passes onto multiple layers directly to your compositing program of choice from single file. So, there’s no time penatly for rendering out separate passes (most renderers, including mentalray, composite these channels internally anyway).
It is far easier and much faster to tweak the look of a scene by directly manipulating a render pass layer in a compositor than to render test after test in a 3D program (even if you have an interactive photorealistic renderer like Maya does).
You can make significant changes without having to rerender your scene. For instance, your supervisor looks at your dailies and wants you to change the hue of your main character to be slightly more pink and wants the highlights to be slightly more blue and have a little more contrast. In a compositor, this change is only a couple sliders away and instantaneous.
Finally, you regain some of the artistry and free-form expression to a 3D project that tends to get lost while you’re waiting for your test renders.
This workflow has saved me so many times it’s insane. Projects that before took me weeks to finish now take me days, with far more flexibility to react to the client inevitably saying, “That’s good. . .but what if we change this to that?”
Our skin is made in layers, right? Maybe our renders should be too. . .
So, I’ve been looking around for skin texture samples (free, of course). You know, something with even lighting and close-up shots around the eyes and nose, and came across this database from Rutgers Univerity:
It says this is a database for visual artists but they’re really just lots and lots of photographs. The cool thing is that every sequence was shot with lighting at different angles (measured too!). So, it’s easy to combine these images, convert them to Lab color in Photoshop and even out the texture. From there you can pull the various diffuse, bump, specular and relection maps.
Looks good Riperton. I think you have a good handle on the lighting properties of the skin. What type of technique did you use?
My suggestion to improve on it would be to add a, VERY subtle, dark red celular or cloud procedural to the colour channel. This would help get rid of the one tone CG look and make the subtle varations within the skin tone. In my opinion this addition is very low investment, high reward.
Thanks for compliment and advice.
What you talking about reminds me of something I played around with some time ago - perhaps I should give it another chance.
No SSS. Blinn-shader with anisotropy; (of course) blue color in the Specular-Channel. In the finalRender-Material I can mess around with more than one Glossiness-Layer in one material, so I have two different settings - but I think you can achieve the same effect with a curve.
I used Falloff-Maps in the Diffuse-Channel(no maps in the self-illu). Some time ago I experimented with weird curves, but every attempt worked just with a few light settings. Now I mainly use linear curves with different maximums and different kinds of red for the light areas.
P.S. The frames above took about 2-3 minutes rendering time
Put more detail on your post on how your skin was created and what the project will be used for. Description on tools and tecniques are requred for anyone to give a proper critique.
Anyways, off the bat, the model requires a better light rig. Start with the lighting before you texture. A simple 3 point lighting rig would do. Better lighting would alo bring out the bump detail.
Try not to render the face straight on. It is easier to see detail at a quarter view.
The reddish area at the sides of the neck seem to be in the wrong place. The reddish hue should appear in the transition between the light area and the shadow area. This is how you can get a sense of what is under the skin.
I guess this is all I can say until the lighting gets better.
is the fast skin shader,i want make a jappanesse man , i have 3 point ligths ,one is back of the modell for create a red zone.i dont knoow how to illuminate well.
there is nothing more.
thanks
That explains the red on the neck and the light colour ears.
I cant tell if the skin is good or not because the lighting is very bad.
Learn to light properly before getting into advanced texturing. If lighting is bad, everything looks bad.
Look for tutorials on basic 3 point lighting. Once it is lit properly, you will be surprised at the difference.
this is my old man i was wondering if its posible to get pixar like skin with the mental ray sss skin shader in max 8 if not i still need help with the skin please crit http://forums.cgsociety.org/showthread.php?p=3308023
