View Normal Mapping
Normal Mapping is like bump mapping on steriods. The result of normal mapping is so good, that once you use it, you may never go back to old-style bump mapping again. Why are normal maps so superior? In a nutshell, normal maps can account for three axes of information, not just one like bump maps. If you look at a normal map, you will see that it appears rather strange, with red, green and blue shapes appearing to be ramped together. When applied to an object and rendered (some Open GL cards can do this in real time), you will see a great representation of your object’s shape as modified by the normal map.
Another strength of normal maps are the fact that you can actually get real world surface textures into a normal map with a few easy steps (and applications such as Adobe Photoshop). Normal maps are far less a cheat than are bump maps. To learn more about creating normal maps from the real world, google “normal maps.” Have fun reading!
So what are the limitations of normal maps (you gotta have something wrong, true?) Well, like bump maps, the silhouette of an object that has a normal map applied will not be altered. And, they do require more memory to process than a simple bump map needs.