Frustum Culling Here is a camera placed over a terrain model. In the left view below, the view frustum is visible.	Below is the view from the camera in the scene.
	If the polygonal mesh making up the terrain model is one single object, the camera takes in to account not only the polygons visible in the view of the camera, but every polygon behind the camera as well.
	When modeling the terrain, you can optimize the scene by cutting the model into multiple parts that match together. This image shows the terrain cut into nine sections.
	In this view, the highlighted sections of the terrain model are the only parts detected by the render engine as being within the view frustum of the active camera. The number of polygon faces being calculated and rendered is reduced significantly. Consider this optimization whenever possible, such as in an architectural space. It may be beneficial to build a room out of seperate walls rather than a single box. It depends upon the complexity of the scene. Do a web search for more info on view and frustum culling, or start here: www.flipcode.com/articles/article_frustumculling