There is information about the Materials System in the SL Wiki and Forum. The part I found interesting is the link to: Brief Considerations About Materials. This is an article by a teacher for a Post-Graduate Course on Game Art.
The emphasis in this article is on Specular Maps. This is something many of us have little knowledge about. But, the specular map is going to be far more complex than the normal map. This is the area where knowledge will be key to making nice object textures or great object textures.
The article is a basic start to finish piece. There are a couple of things left out but, it is all most creative types in Second Life™ will need. But, remember. This article is more about what things are (information), than how to use a program to make them.
A point on understanding why normal maps are important is skipped in the article. In fact most articles on normal maps are skipping it, other than to vaguely say they save CPU cycles. I’ll explain what they are leaving out.
If you consume Blender tutorials on making normal maps, you will probably notice professional modelers make a high-polygon model. Next they bake a normal map from it for use on a low-poly version of the model. This is done so the game can use the low-poly model and look like the hi-poly model.
The reason this works in the game is that all the normal directions of the polygons must be calculated to figure out how to reflect light. All those normal calc’s are pre-calc’ed and baked into the normal map. In a way they are caching those calc’s so the game does not have to do them.
He does explain how to get a preview of the effect of your specular map while still in Photoshop, very easy.
This article has given me new insights to what a specular map is and how to work with them. Well worth the read.
The more I read these articles that supposedly provide insight about why normal maps are important, the more I feel compelled to produce an article that touches on every reason why they’re important.
They’re important for various reasons. “Pre-calculating normals” isn’t one of them, as meshes store normals in their vertices as-is, and only ever change when a mesh’s vertices change position (such as on a rigged mesh). It’s all about saving fill rate for the GPU. The more triangles you have, the more work the GPU has to do to rasterize those triangles. With normal maps, you don’t need to have anywhere near as many triangles to simulate wrinkles, folds, or what have you.
Ok… you out teched me… 🙂
Pingback: Dies und Das aus dem Web (KW 15 / 2013) - [O.L.D.] Old London Docks
Brief Considerations About Materials was great. Thank you for (linked) those definitions and practical terms.