[WIP Clip] Rog's Sock Tease - Cloth Sim and Shader Test (Patreon)

Another WIP clip - this time of the newly done cloth simulation for Rog's socks, along with their new shader in action. This one is a test render clip, so it’s been rendered with lower samples and no denoising to speed up rendering time.

As before, you can find the WIP clip attached to this post.

I'm in the process of working out some vocal stuff ahead of time, so Rog's facial animations aren't set in stone. Also, animating the chair buckling has led to some clipping with the legs when he stands, so I'll be fixing that too.

For the socks, I'll be animating the thickness of the hair curves so that they appear thicker when they’re closer up to the camera. I'll also trim them down a bit shorter too. As for getting some of that translucency back into the shader, that requires getting a bit more in depth with things.

A side tangent on transparency/translucency:

So, I don't want to get too into the weeds with shader stuff with this post, but the important thing for this topic is that it relies on one main procedurally generated texture for most of its functionality. And it's very simple to produce an alpha mask from that texture:

When plugged into the shader, the white parts will remain opaque while the black parts will be transparent. You can also easily set it up to produce translucency in the shader instead. Then by modifying the values of the alpha mask I can control the size of the gaps and how much light is able to pass through them:

Currently, my setup filters the alpha mask into two variants (One with closed threads and one with opened ones), which is then blended between by a mask texture created using a Fresnel Node.

In Blender, a Fresnel node basically outputs a black and white map describing how reflective a surface is depending on the angle you're viewing it from. In this case I'm using it as a very hacky form of edge detection, since I want the outline of the socks to remain opaque (to preserve the illusion of thickness in the mesh).

Currently there’s very sharp fall-offs between the values in this Fresnel mask, which gives the fabric more of a nylon feel in places. I’m aiming for cotton, so I’ll be working on fixing that (or replacing it with a new masking method). I’d also prefer for the size of the gaps to actually be determined by how much the mesh is physically stretching. That’s something I'll need to do more research on.