04.21.2020 | Project 2 - Shader Experimentation
Updated: Apr 25, 2020
Now that my scene has been set up, I imported my Stanford Model, Lucy, and began adjusting the attributes of the aiStandardSurface shader. I researched the IOR of honey calcite, which is 1.486 and plugged that value into the shader. I experimented with the transmission color and added a little bit of transmisson depth because my stone has a cloudy interior. In order to get the varieties in the color, I used an aiNoise map to drive the transmission color attribute.
I also noticed that the transparency of my stone shifts depending which angle you look at it, so I needed to find a way to make the edges that faced the camera more transparent compared to the rest which appeared more cloudy or saturated. In order to do this, I created 3 new render layers that I plan on compositing later in Nuke. I created a layer for my fresnel mask, a layer for the clear glass and a layer for the more cloudy/saturated glass. To create my fresnel mask, I assigned a new aishader and attached a ramp node to the emmission color. Then I deleted the "place2Dtexture" node, because I want don't want the ramp to be driven by UVs. I instead, substituted it with an aiFacingRatio which will map the ramp based on the facing angle. I adjust the bias as a way of controlling how much transparency I want to have on the edges of my model.
Then in my clear glass layer, I create a new aishader and transmission attributes to get clear glass and I add a value of 10 dispersion abbe to give the effect of chromatic aberrations which my stone has.
Once that is done, I render out the three layers and bring them into Nuke. I shuffle the fresnel mask layer and tell Nuke to use the red channel as the alpha. Then I merge(in) my clear glass into my mask and merge(out) my cloudy class with my mask. Then I merge the two together to give me the final composite.