3D Chips and Cards

Anisotropic Filtering

This feature is another catchword in the chase after customers and readers. I try to explain it as simply as possible. Bilinear and trilinear filtering, as used today, are ways of interpolating a pixel as a part of a texture so that it blends in into its environment reasonably. Those two filtering techniques have one big disadvantage, they filter evenly in each direction. This means that the filtering technique is the same at the edge of a wall as in the middle of a wall or the filtering is also the same regardless if you look at a surface under different angles. Unfortunately this isn't very realistic, at the edge of a shape the filtering has to be different and the viewing angle should also have an impact on the filtering. Anisotropic filtering doesn't mean anything than that the filtering is done 'unevenly'. An algorithm has to find which part of a texture needs which kind of filtering. Thus anisotropic filtering is not a clearly defined procedure as bi- or trilinear filtering is. There are many different ways of implementing anisotropic filtering into a 3D-chip. We will have to see which one's implementation looks best and most realistic. So far there isn't any hardware implementation of anisotropic filtering in any of the new or announced 3D-chips.


So far about the requirements of a well performing 3D-chip or 3D-card. Of course there are a lot of other effects that 3D-chips can do today as well, like fog, transparency or different kind of shadings, but those 'common' features don't really need to be mentioned again. However, there are still cases where features are implemented at a low quality level, when 3D-chip makers rate frame rates over image quality.

Realism is what we really want, the users and game developers have to tell the 3D-chip makers what they expect. The hardware developers have made another leap forward in terms of realism with the latest generation of 3D-chips and 3D-games should follow soon to take advantage of it. A look at the CPU-roadmaps shows that we can expect a lot more performance in very short time.