I've been hitting the node references and tutorials lately, and it's been rewarding to finally get a glimpse of understanding on how these nodes interact. Like how the gradient patch size is so important in rendering detail and controlling slope masks.
That's where my current challenge lies. I've created a deep, drift-like snow surface layer for the recesses of the distant mountains in the attached pic. The problem happens when I try a maximum slope to allow the larger rocky peaks, etc, to peek through while still maintaining detail in the foreground. A patch size of 100 gets me the smooth variations I want for the snow, but then the foreground loses a lot of detail. A small patch size of 5 preserves the foreground detail, but makes the snow layer too noisy and unrealistic. I tried branching into two different compute terrains/normals, but wow, that HUGELY increased render times even for a low quality render. I could export the snow shader to Photoshop, clean it up, and import it as a simple mask shader, but that seems a clumsy and error-prone way to go.
An aside question: Wouldn't it make sense in the TG compute terrain to scale the gradient patch by distance? Sort of like how tree populations are rendered with decreasing quality the farther they get from the camera?
Cheers, and thanks for any help. If nothing else, I've provided a rough example of how gradient patch size affects a scene.
Mr. Ed
