Selecting Level of Detail (LOD)

The next step in the process is to determine the level of detail for any nonculled objects. Level of detail takes advantage of our lack of ability to discern object details at a distance. Because of this, we can model an object with fewer polygons when it is far away and more polygons and detail when it is close to the eye point.

In the past, we would need to store a version of the objects mesh at each level of detail. We would then select the best mesh to use based on the current distance from the camera. While you can still use this method if you like, Microsoft does provide a better way. This improved method is called a progressive mesh. The details of creating and using a progressive mesh will be covered in the next chapter. The point to take away from this chapter is this: by specifying how many polygons to draw for the mesh based on how far away the mesh is, we will improve performance. Small, simple objects will not benefit from progressive mesh handling. This is reserved for those models that will have a lot of detail when close to the camera.