PDA

View Full Version : Difference between Bump/Normalmap



HolyMonkey-
04-24-2012, 04:29 AM
Long story short

What is the difference between bump map and normal map?

----------------------

Short story long

You have a piece of wood texture.

You have a black and white heightmap
And a normal map of the very same texture.

What is the difference between plugging the bump map into bump, or plugging the NM into Normalmap slot?

I cant see any difference. The only difference is that it takes longer to create a NM than just using the bump.

biliousfrog
04-24-2012, 05:36 AM
A normal map uses RGB information to create the illusion of bumps which have XYZ values.

A bump map uses a greyscale image to create the illusion of bumps which just stick in or out.

You will probably not see a lot of difference in a static render but, in an animated one, you will see an improvement in how the normal map creates the illusion of surface detail. Neither method will show detail beyond the edges of a mesh but normal maps are more convincing at glancing angles because they look like they are growing out from the normals of the object rather than directly at the camera.

Sensei
04-24-2012, 05:49 AM
Bump mapping is slower in rendering because renderer has to convert bump map to normal vector during rendering. 6 pixels of image have to be read (if blurring and mip mapping is turned on even more)
Normal map in object coordinate space is read directly by renderer. 1 pixel of image is read.

lightmann
04-24-2012, 10:28 AM
Bump mapping is slower in rendering because renderer has to convert bump map to normal vector during rendering. 6 pixels of image have to be read (if blurring and mip mapping is turned on even more)
Normal map in object coordinate space is read directly by renderer. 1 pixel of image is read.

thanks, for that technical note Sensei. You deserves that name.

HolyMonkey-
04-24-2012, 11:13 AM
Bump mapping is slower in rendering because renderer has to convert bump map to normal vector during rendering. 6 pixels of image have to be read (if blurring and mip mapping is turned on even more)
Normal map in object coordinate space is read directly by renderer. 1 pixel of image is read.

Sweet, I have noticed that when I apply NM my texture seems to get darkened, any idea why?