I’m wondering if I’m mapping right. I’ve got a load of textures fromm zbrush and want to apply them in blender. I presume the blue/red normal map should be mapped to the normal channel? Why do grayscale images get mapped to the normal channel then? I’ve also a displacement map which is also greycale. This shuld be mapped to the displacement channel? Why do I need to use a displacement map when the nrmal map seems to be doing the job? Whats the difference between normal and displacement maps?
Displacement map (typically grey-scale) actually modifies the mesh (moves the vertices about).
Bump map (grey-scale or the red/blue one) actually a synthesis of the surface along the normal (along the Z only). This does not modify the mesh but influences the light as it passes along the surface.
Normal map (the “baby” blue one) also a synthesis of surface but is tri-directional. The RGB channels each correlate to an XYZ from the surface. This too does not modify the mesh but influences the light as it passes along the surface.
The Bump map “costs” the least as the grey-scale image holds very little data. This is good for game engines and when size matters (GPU rendering for example).
The Normal map is the next cost, but will look better. This is sorta middle of the road.
The Displacement map is typically the most costly as the density of the underlying mesh needs to be high (but usually less so then the multires mesh from the sculpt).
I typically decide which to use by the detail I need vs. proximity to the camera. A wood fence in the background that is not the focus can use a bump map easily. A wood surface the subject is on at a close shoot I would use a normal map. A discplacement map then would be good for various wave effects on a pond of you wanted distinct control (water that kind of burps up around the surface of say a floating life saver).
Lastly, a renderer if both a bump & normal map are specified, the higher quality normal map will be used and the bump will be ignored (though I don’t know about the BI for certain), but it tends to be a general rule of thumb.
Three different channels you have do different things. It is true that Displacement map can do it all. But Displacement map will actually creates fictitious polygon surface to render like in real object, like nose on a face. You are talking about gazillion vertexes and faces computer need to create no the fly. It is like having actual zbrush model, as far as render engine is concerned.
Both Bump and Normal map are rendering tricks to make thing look 3D. They render much faster because they are just tricks. Both are not generally used for large 3D features. Meaning, when you look at the object from different angle or from near on edge, you can tell it’s a fake surface feature. Bump map works best for small shallow surface imperfection, like skin texture on a face. Normal map is good for 3D feature little more larger and deeper, like wrinkle on a face.
Using those channels together has this mixing effect on the eye. When one effect starts to fade other effect still holds. It is hard to spot when feature turn fake. It looks more real than just using one channel.
By the way there is one more important channel, the color and shading map. You need that mix as well.
Thanks guys. Sheds some light on the subject. I was wondering why I needed to use displacement but it makes sense now.
I’ve placed the colour map in the first channel and added normal and displacement after that. Is there a right order? I’ve also aded another colour map after that with some alpha - like scratches. Again i order important for this kinda thing?