Week 11: Finalize and polish

Week 10: Texture Baking and Testing in Unity

This week, I baked textures and exported the arterial network as an FBX. I have always known how to bake normal and other texture maps from Houdini, but I was never able to bake color maps until now. I have tried and tried for weeks to no avail until now. I have exported two different artery network models, and baked two different versions of textures for each. The second variation has more realistic detail in the normal map, but no color map. The first variation was baked from a Houdini procedural Material Builder, the second was baked from geometry displacement from a VOPs network. The VOPs network is geometry dependent, and therefore requires more processing power to work with. So, I will combine the procedural noises from both in a Material Builder in order to have the basecolor texture map and the other texture maps. The HBV model has also been exported with baked textures. Materials have been applied to all objects:

Here are some videos to further illustrate:

My goals for the upcoming week are:
⦁ Scale down the resolution of the procedural Material Builder's texture to add more detail.
⦁ Combine the VOPs and Material Builder in some meaningful way.
⦁ I like the bumpiness of VOPs, and the procedural Material Builder's color.
⦁ Bring the noises, with the exact same settings from VOPs into the procedural Material                         Builder.
⦁ Bake height, metalness, and roughness maps, in addition to the color and normal maps, 4K!
⦁ Remove texture seams in Blender.
⦁ As a stretch goal, finish solving the rectified truncated icosahedron problem, for a more realistic        HBV particle.
⦁ Basically, combine existing approaches, and add more detail and realism to everything.

Week 09: Artery Material

Started bringing assets into Unity, created a procedural skybox in Unity.

Exported the HBV particle and baked texture maps from Houdini, created a transparent material in Unity.

Close-up of capsid with material, texture, and flat color applied. The arrangement of surface proteins isn't totally technically correct, but will do for now. An instructor and fellow student are helping me develop a technically correct T=4 icosahedral capsid, which is pretty much a rectified truncated icosahedron. The vertexes in this shape will serve as points to duplicate the surface proteins.

I began developing this material in Houdini for the inside of the artery walls. So far, I haven't been able to figure out how to convert point color to vertex color in a way that would allow me to bake the detail here from the material node as a color texture map. If I could figure out how to do that, then I would have the color map, then I would generate geometry displacement from the procedural texture in the material node and bake that in normal and ambient occlusion maps. 

So, I decided to continue development of the artery wall in an Attribute VOP node instead of a material node. Below are a couple of variations of the results I have achieved so far in VOPs applied to a half pipe. This geometry detail can be baked to texture maps and a lower resolution model.

These are the sample branching arteries developed last week by using a polywire node on a curve.

Here is the Attribute VOP applied to the sample branching arteries. The polygon normals have been reversed, and the view is from the inside. It's a good start, but needs some serious tweaking.

Development will be continued on the arteries and HBV particle next week.

Week 08: Arteries

This week, all I really did in Houdini is create some curves, fuse them, then polywire them. Here is the resulting node network:

As one may see, there is a high-poly version, and a low-poly version to bake to.

Week 07: HBV DNA

The first couple of times I attempted to create a double helical strand, I would have been successful if I continued, the only problem is that I would not be able to control them via a curve because of the methods I was using.

Here, I can control this strand via a curve, and it has ladder rungs alright, but it does not have that characteristic double-helical twist that a strand of DNA has. Creating a poseable double-helical shape has proven to be far more complicated than I originally thought. This will have to do, for now.

Week 06: HBV Particle Capsid

Here is the icosahedron, the base shape of the capsid that all other components of the viral core are built upon:

From this, the protein coat was created:
High-poly:

Low-poly:

Low-poly with textures baked:

The model has been kept fairly simple at this point, because it has not yet been determined what the exact needs for the thesis project are. It can be easily adjusted because it was created using procedural workflows in Houdini.

Week 05: HBV Particle Envelope

The Houdini asset scheduled to be completed this week was the hepatitis b viral particle envelope. It has been completed. The overall shape is driven by a sphere with smaller spheres duplicated across it's vertices. The overall shape and amount of surface proteins can be easily adjusted or totally changed.

 This is the high-poly model.

The low-poly model and it's topology can be seen here below.

Here is the low-poly model with the texture baked down.