Normal Map vs. Displacement Map
 

Hi,

can someone please explain, why game developer use normal maps on low-poly models? What is the advantage of normal maps compared to displacement maps?

I know that is a basic question, but I am a bit confused. :headbang: I always thought disp. maps would be the right way to go from a high-resolution model to a low-poly model.

Thanks!

displacement maps will literally add geometry to the model to achieve a high-res result. Game engines, while able to handle a lot more polys than previously, still can't handle THAT much more geometry. Normal maps achieve a simulated high-res look without adding geometry at all.

I'd like to add a little to this question. Since visually they're alike, why whould you wanna use disp. when you can use normal maps?
Since normal maps are usable in realtime rendering and disp. maps aren't, I'd assume that disp. maps are a lot render intensive than normal maps

Yes, displacement maps are a lot more render intensive, but normal maps don't ACTUALLY alter the geometry. For instance if you look at the silhouette of your normal mapped model, there will still be edges. - normal maps still have the same basic idea of normal bump maps in how they only affect the lighting during the render process - so it still has the problem where the more parallel you view your surface the less of the normal map detailing will be seen. That's the way I see it anyway...

If you have no restrictions regarding poly usage (ie. rendering a movie) there's no reason to use normal maps, as displacement maps give an actual high-res result.

As was mentioned, because normal maps are only simulated high res detail and the detail ISN'T actually there, you still see the profile and geometric limitations that the low-res model originally is. Which isn't very condusive to high-res movie rendering.

I'm working with Zbrush and Maya and therefore I use disp maps to transfer the model from zbrush to maya without bringing >1 million polygons to maya.

I don't have anything with game engines to do. So, therefore I think I do not need realtime rendering.

Actually, what I do mostly is creating small aniamtions, still-pics, and adding some special effects to self-captured videos. (all non-professional)

Am I right, if I say that I actually don't need normal maps for my stuff and disp maps are better for me, since rendertime is not really a problem for me?

In my opinion, yes, you are correct.

Ahh, so what I was saying is actually kinda like saying "why model holes when you can just use transperancy maps?", right?

Erm, imo you said the other way around, but yea.

Can you make normal maps in Photoshop 7 and can you see the affects to in PS 7, or do you need a program that surports normal maps.

You will need some 3D software to visualize the normal map, but I think it would be possible to create one in PS7/8/9/...20

but you will have absolutely no control.