Quote: "so displacement mapping IS normal mapping?"
Yes, and No...
When your talking in terms of DirectX/OpenGL graphics, then really Bump, Normal and Displacement Mapping are all evolutions of the same technique.
Bump Mapping, uses a Dot Production of the Normal Vector Multiplied by the Light and then placed on a strength of light scale to alter the brightness of a texture map and fake depth.
Normal Mapping, evolves this by taking an rgb image and using that as a 3-float Vector (+Softness) using a Dot Production 3 of the same technique as above to provide a 3D depth.
Displacement Mapping, takes this one step further by altering the texture coordinates of the Normal Map based on the eye vector to make sure that it's height is altered to give depth even at extreme angles. (Normal/Bump Mapping only effectively fools the user upto 45-50° off center, Displacement does it upto 89°)
However in order to stop artifacts appearing from the Displacement and to even use it effectively in the first place the object needs to have subdivision of UV points and take more than one pass at an object.
(this add vertex requirements and pixel pipelines)
Effectively Normal Mapping works wonders and is fast for organic models, basically things that are curved and are constantly moving.
Displacement Mapping is great for static levels where surfaces are flatter and can be more heavily inspected.
Bump Mapping is good as a fast solution when your just adding character to a surface and not trying to fake it being made of more polygons than it actually is.
... non of those solutions will subdivide a surface in order to give it more detail. Infact doing so for a game would infact be counter-productive, the idea is to lighten the processing load; not add to it.
[edit-]
Oh yeah about the part of DBP.
It is possible to give users the exact functionality as Blitz3D has in order to create that Normal Map effect, much quicker and without the use of Shaders (meaning we wouldn't have to rely on shader cards) the same goes for Bump Mapping, and infact the original 1.0 release didn't use FX to achieve it.
But TGC are going down the route of 'any special effect for the engine is going to be Shader based' ... which means everything from normal mapping to reflections, refractions, or even simple texture blending. Will eventually all be part of one big Material (HL2 Style) Engine.
Great for simple graphics creation, unfortunately cuts a VERY large market from bedroom developer markets. It also means that people will need some very very awesomely fast cards just to do some effects rather than now where you just need a relatively fast CPU.
CPUs have and always will be cheaper and more available to upgrade than GPUs; so it is an annoying and quite frankly stupid idea for TGC. I'm not opposed to having a Shader Material Engine; but I am opposed to it replacing what we have and making the markets for people who can use this stuff smaller.
