Level of detail

Level of detail

This is a smart resource. Use a QR reader or the Aurasma app to access additional content.

https://goo.gl/ofoTXC

Level of detail is a method of reducing the number of polygons or triangles of 3D geometry based on its distance from the player or camera. This technique increases the efficiency of rendering by reducing the workload on the graphics card or GPU.

This is usually achieved through the use of level of detail groups. The way these work is that for each asset or piece of scenery in a game, a number of models are actually created. Each of these models has a different polygon count ranging from the most detailed version which may be tens of thousands of triangles down to a version featuring very little detail and a vastly reduced polygon count. This is demonstrated in this example featuring a bunny that we’ll call Nigel;

As you can see each model represents exactly the same bunny. All of these are clearly Nigel. These four models would represent the different levels of detail in a LOD group. The clever thing about the way LOD works is that the closer the object is to the camera, the higher the level of detail that is shown to player. So if in a game we’re standing right next to Nigel, we’ll be shown the version with 69,451 polygons but if Nigel's right at the other side of the level we’ll only the the 76 poly version. The next example demonstrates the way that this works;

As you can see, when we’re close to Nigel we need to see the highest quality version in order to see all of the detail but as he moves further away the polygon count can be reduced without us really being able to tell. This avoids wasting processing power rendering detail that we can’t see because the object is too far away and it ensures that the game can maintain an acceptable frame rate.

Have a look at the video below to see a LOD algorithm in action working on reducing the polygon count of the terrain which makes up a planet as the camera gradually moves away from it.

http://goo.gl/Up2UZq