Hey guys. Hey Mustakettu85, didn’t realise you were a regular over here.
... I’ve also been doing a lot of crazy Bouce GI stuff with UberEnv.
Hi Jim! About as regular as a rain in the desert, actually LOL And there’s some neat stuff you’re squeezing out of the UE2Bounce, especially that thing with the point-cloud rendertime script… I gotta take a closer look at the script code, what if it transfers the whole GI thing into the point cloud, not just ambient occlusion?! It had never occurred to me to test the script with any IDL/GI mode.
...there’s some funny comments to your renders over there on dA One thing I never understood about CG, is why people keep blaming the software, not their skills. I would say that no traditional artist would blame their paints or pencils for their inability to produce “lifelike” shading =) Good materials make for easier execution, but no expensive brush will paint by itself.
...you have to have some knowledge or at least a material property database somewhere to rely on.
Hi Takeo! Yeah, that sort of a material property database is something that I’d love to see. Even if we are using “cheat” shaders. It’s still great to have a less biased source to work from than just eyes.
And what I’d also love to have is a visual chart for Fresnel effects with various refraction indices. All the omnifreaker shaders have a very weird way of controlling Fresnel and the shader mixer brick I´m not sure I found the right formula to calculate its refraction ratio input. So I’d love to have a pictorial “database” to test and calibrate them against. I have Carrara 6, but I’m not sure it does Fresnel correctly?..
There you’ve done a pretty skin and for not too realistic render it’s OK. But if you look for realism, then I think it’s a no go unless you programmatically break the uniformity
I’ve made some “almost” skin procedural shaders too.
Thanks! And yeah, you’re right, I can’t imagine a 100% procedural solution looking close enough to fool people… unless there is at least a sort of a distribution map for different procedural algorithms…
And I’d say everyone here would want to see your shaders (constantly evolving, right?), so… you should probably whip out those cropping scissors =D
If you don’t want to deal with surface group then render in Point Cloud mode. It’s quicker, and saves from the ID Group messing problem.
Soooo… so that script does enable point cloud for everything we throw at it? Or are you talking about writing pointcloud scripts of our own?
Thanks also for posting your aoa sss settings, I saved them and will try to set them up this weekend and do some test.
Hi! You’re definitely most welcome! This AoA Subsurface is a very complicated shader, since it’s so complex and has all those features… maybe a bit too complex.
I actually have run into this whole nail issue, too - with other shaders as well. Even if all the SSS settings are totally the same, there is still a visible difference if the GroupID is different between nail and skin surfaces. It can be used artistically, though, I believe.
Here’s an example… please note the shadows that appear around the nails when they are “separate” from the skin.
And yeah, Takeo is completely right, if the SSS settings are different but the GroupID is the same, unpredictable things may happen.
EDIT: IIRC, at least for omnifreaker’s shaders, GroupID of 0 meant that the shader would decide automatically if the meshes are the same object or separate ones. I have not used this “automatic” feature for a looon time, so I can’t really say how well it works. I usually err on the side of caution and apply different IDs to different surfaces and different meshes, but this may not be necessary. I would say that the renderer will differentiate between a human figure and its clothing, even if their surfaces have the same ID…
Just in case there are people lurking here who have never come across docs for omnifreaker’s shaders, here they are again… there is some useful info there:
Click thumbnail to see full-size image