Bump map vs normal map ?

[GABRIEL EM]

Im confused, i know normal map is more advanced than bump map, but i dont understand why, i mean, without technicals explanations, which is the difference when we see the render ???

[Freespace-3DT]

A bump map simulates depth through two dimensions - black and white.
A normal map simulates depth through three dimensions. The end result can fool the eye.

[bb0x]

Basically a bump map only contains height information. While a normal map contains informations from '3 directions'.

Look at the normal maps different channels in photoshop. You'll noticed that the red channel looks like its lit from the side. Green channels looks like its lit from the top/bottom. and the blue channel contains depth information.
While you could make a bump map with just 1 channel.

[Tank[NL]]

if you generate a normalmap(which is also known as a DOT3-bumpmap) from a high mesh, you project the, literally, the directions of the face normals of the high poly on to a texturemap that correspond to the lowpoly mesh.

you basically store the information of the highpoly normals(the directions faces are "looking" at) onto a texturemap. a shader then interprets this information, resulting in the illusion of more detail or perceived geometry then which is actually being rendered.

a normalmap can relay detail more accurately compared to a standard b/w bumpmap, especially with curved surfaces, such as organics or modern cars.

bumpmaps and normalmaps both have advantages and disadvantages. depending on the type of project you're working on, you need to decide the type of bumpmap.

gl and i hope this made any kind of sense.

[Nukie]

This little tutorial explains it quite well.
http://www.cgenie.com/tutorials/arti...-textures.html

Link is a bit slow.

[GABRIEL EM]

Thank you all for explaining me !
It is much clear now, i´ll try to use normal maps from now on

[Nukie]

You should use whichever map is useful to your project.

If you are building a stone wall, you probably will do oke with just a bump map. (Bump map actually works in most architectural renders, but are not limited to only that).

If you have a high poly model, and for some reason you want to use a low poly version of it (a not so important object in an animation for instance), you probably would go with a normal map.

You will use displacement mapping of you want the original geometry to deform. Displacement does add extra polygons to your model and so can highly increase render time. But in some occasions you want to have a displacement. They can be used, for example, for carpets.

Maybe someone else has a better explanation (I suck at explaining these things) but bottom line is to not limit yourself with only one possibility as they are all useful.