Film processes in a games
greentooth
Hi all,
I'm new to the polycount forum - the quantity and quality of information/tutorials available here is great! Loving it!
I've been working in 3D in the film industry for a few years so I know my way around 3D packages but have a few questions regarding shading and texturing and how these areas translate into games, especially next gen gaming.
In film, I normally group objects (or parts of that object) of the same likeness into the same shader. Wood objects point at a wood shader; metal objects point at a metal shader. These shader properties are very different so they are independent to each other. In games would this process still be possible or is it better to have the entire object point to one shader and have texture maps define the different materials of the object?
I keep reading about people now using multiple UV sets to texture large areas of terrain. When the texturing is finished is it baked into a massive texture that defines the entire ground or is there now a better process used to capture all of this information and render it on the fly?
When modelling I leave my assets as quads but is it better to triangulate? In examples online I've seen meshes left as quads and converted into tris so I'm not sure which route to go down!
I have some experience with asset creation for games but as technology has improved have processes evolved as well?
Any input would be much appreciated!
I'm new to the polycount forum - the quantity and quality of information/tutorials available here is great! Loving it!
I've been working in 3D in the film industry for a few years so I know my way around 3D packages but have a few questions regarding shading and texturing and how these areas translate into games, especially next gen gaming.
In film, I normally group objects (or parts of that object) of the same likeness into the same shader. Wood objects point at a wood shader; metal objects point at a metal shader. These shader properties are very different so they are independent to each other. In games would this process still be possible or is it better to have the entire object point to one shader and have texture maps define the different materials of the object?
I keep reading about people now using multiple UV sets to texture large areas of terrain. When the texturing is finished is it baked into a massive texture that defines the entire ground or is there now a better process used to capture all of this information and render it on the fly?
When modelling I leave my assets as quads but is it better to triangulate? In examples online I've seen meshes left as quads and converted into tris so I'm not sure which route to go down!
I have some experience with asset creation for games but as technology has improved have processes evolved as well?
Any input would be much appreciated!
Replies
Generally, you will use 1 material with textures defining the albedo, reflectance, gloss/roughness and normals of each surface type within that material. For certain surfaces, you'll use special purpose shaders, for instance skin, hair, glass, etc, and these may have separate materials and textures just for those surfaces as they may require different/more texture content.
Depending on the engine, you may set up multiple basic materials, and then layer those together for the end result. Again depending on engine, those shaders may be baked down into single maps, especially for smaller items.
For larger objects like terrain and buildings, no, otherwise it would be way, way too much texture memory. Tiling textures are required to get high pixel density over huge spaces.
There is no need to work in triangles. You may want to triangulate before you bake normal maps, but working in triangles while modeling is unnecessary.
Of course, everything in the games industry is constantly evolving.