Normal vs Displacement maps

Sastrei

Hi all. Quick question, looking at artistic, not technical answers:

Normal maps and displacement maps would seem to affect a lot of the same spots on a texture - would I be correct in understanding that while a displacement map is used to alter the height of the poly or micropoly, the normal map then provides the lighting information?

In other words, I should make sure my normal maps pay attention to the edges of a surface, while the displacement map does the work of raising or lowering the actual surface?

-Stefan-

jocose

You are correct. Normal map ONLY adjusts the lighting on a surface to make it look as though it were raised, when something its not. This visual trick only holds up reasonably well and at very oblique angles visually fails to impress.

Displacement maps offset the rendering of geometry so the resulting render is raised just as it would have been if you had modeled in that detail with polygons.

Michael Knubben

Displacement maps (in realtime rendering) need actual geometry, and are used with Directx11's Tesselation for instance. A normalmap's detail is limited mostly by the texture resolution, as well as the above mentioned limitations.

ZacD

@MP

Parallax also uses height/displacement maps, and is a more common realtime effect.

Sastrei

Thanks for the advice all! And yes, I'll be using the displacement map for a parallax shader.

-Stefan-

Michael Knubben

ZacD: Right you are! I was strictly talking about displacement, I'd never heard of them called Displacement maps (only height maps) in regards to parallax.

ZacD

displacement maps and height maps are basically the same thing, but displacement maps sometimes account for subdivision.