[Solved] Why not displacement maps instead of normal maps?

[Roland]

Is there any reason why normal-maps seems to be preferred (or are they?)

compared to displacement-maps. If I got it right, displacement-maps gives

better results. Are the slower, consumes more GPU memory or whats the reason

why normal-maps are used instead?

[Canardia]

Displacement maps create additional faces, so they are much heavier to render:

http://www.ru.is/kennarar/hannes/useful/blendermanual/htmli/ch11s03.html

I think instead of displacement mapping you could just create the details with real faces and then use a normal map on top of the details.

[Roland]

Displacement maps create additional faces, so they are much heavier to render:

http://www.ru.is/kennarar/hannes/useful/blendermanual/htmli/ch11s03.html

I think instead of displacement mapping you could just create the details with real faces and then use a normal map on top of the details.

 

Thanks Lumooja.

 

Yes. They write the answer in your link

 

For distant/non-critical items, NOR mapping should still be used. NOR maps, compared to Displacement maps, impose very little additional CPU cost per renderface, and you can NOR map independently of renderface count. Use of Displacement maps quickly leads to millions of faces in a scene.

[Rick]

Does LE even allow displacement maps? I'm still fighting for maps to be able to make 3D geometry on the fly. I think that would be very cool and the displacement map seems to always do that to some extent.

[franck22000]

There is the parrallax occlusion mapping but i never managed to make it working correctly i get some quite ugly results :s

[fuego]

I couldn't make it work too

[BrokenPillar]

Does LE even allow displacement maps? I'm still fighting for maps to be able to make 3D geometry on the fly. I think that would be very cool and the displacement map seems to always do that to some extent.

I think displacement maps don't so much create geometry as tell geometry where to go once it is created. From how I understand it, the typical method is to create a mesh that can be subdivided nicely (like you would for sculpting). Then the displacement shader would tell the geometry to tessellate, and then use the displacement map to tell the new geometry how far to displace. I don't think Leadwerks currently supports it, but then again using it in any kind of meaningful way in real time 3D apps is a fairly new thing as far as I know.

 

There is the parrallax occlusion mapping but i never managed to make it working correctly i get some quite ugly results :s

Parrallax occlusion mapping is more similar to normal mapping than displacement mapping in that it "fakes" geometry rather than adding more. It basically occludes pixels from being drawn based on viewing angle, which gives a false sense of depth. Think of brick pixels hiding the mortar pixels as your viewing angle gets closer to parallel to the surface. It isn't creating bricks out of geometry to cover up the mortar, just telling the engine not to draw the mortar pixels anymore.

 

You have to really understand how POM works and what it is and isn't good at to get it to look decent. You also have to create a POM map by baking it from a high res mesh. Subtle differences in value dramatically effect the end result, which makes creating one from hand near impossible. Crazybump can create POM maps, but that is just because it is basically creating a high res mesh, displacing it off a gray scale image of the image you feed it, then baking the maps from that (behind the scenes). If you are going to use Crazybump, you really need to clean up the image first since the POM effect needs to be so precise to look right.

 

-edit- Ha. I just saw that this thread was marked as solved. Sorry for the TMI.