- Joined
- Mar 28, 2008
- Messages
- 2,500
- Reaction score
- 623
- Points
- 128
I have been working on the implementation. Texture loader is done as well as the pre-shader. Currently working on sunlight shadows. However, something else came in to my mind and I would like your opinion about it. Let's say that there is a gauge in a VC panel with a rotating pointer on it. How would the pointer be lit to match the baked background. So, the client would need to form a texture coordinates form a vertex positions so that it can fetch lighting information from a background light texture. To do that each panel would require mapping between 3D space and texture space. Each "dynamic" meshgroup that requires background lighting being taken in account would need to be binded to a 3Dto2D mapping group. Or is this getting too complicated or un-nessecary ?
C++:
struct 3Dto2DMapping {
VECTOR3 PanelCornerPointsIn3D[4];
RECT TextureSpaceRect;
}