Bump maps work only with greyscale to provide either up or down information.
I try to simplify a bump map
See a bump map as water.
black is 0 percent water and white is 100 percent water.
Now take a glass and pour black in it. (0% water, the glass is empty)
Now pour dark grey in it. (25% water in the glass)
Now fill the glass to grey (50% water in the glass)
Now fill the glass to light grey (75% water in the glass)
And now fill it to white (100% water in the glass, it is full)
- water example.jpg (29.09 KiB) Viewed 9595 times
This is the basic how a bump map works.
black is no bump and white is full height all the shades of grey are calculated as height.
This is the picture i used on a simple plane
Here the result, only bump map added
Normal maps are bump maps on steroids.
It holds more information in it on all axis.
It gives you high detail on low poly objects.
Next-Gen game consoles rely heavily on normal maps.
You have two kinds of normal maps.
1: Tangent Space normal map. the most used in Poser and Studio. It is a mixture of purples and blues.
This is very good for animations and characters.
2: Object space normal map. This is a mixture of rainbow colors and is a little bit more advanced.
Here the two different normal maps
How does normal maps work.
In a normal map all the heights and depths are stored in the RGB and is translated into X, Y, Z axes.
So basically it tells how much water is there in the glass and is it a glass, bucket, or swimming pool. This is not completely true, but it is very near.
This is the normal map used
Here the result, only Normal map added
The displacement map works in a similar way as a bump map, but this time it gives real physical depth.
A displacement map can be baked from an object or like a bump map made from an bitmap. Normally the bitmap will be 8 bit, but this is actually not so good for a displacement map.
The higher the bits, let’s say 16 bit or even 32 bit. The render will benefit from the detail of your displacement map.
Displacement Mapping is very much dependent on the amount of polys in the mesh, a high poly mesh will give better results & more detail than a low poly model. Unless use of such things as "Sub-Poly Displacement" (called Micro-Facets in Reality) which subdivides the amount of polygons during render time giving the appearance of more polys in the model than there actually is. (Thanks Fuzzy70, for clearing this out)
But is displacement mapping the holy grail? No it is not. Displacement is hard on your system when rendering. But if you use it, it will definitely pay of as the surface is transformed. That will bump maps and normal maps never do, they are just shaders telling the light how to bounce of the surface. that's why the borders never have depth. (look at the examples from bit map and normal map)
If you use only, lets say, normal maps for a wooden floor with only a top light, your normal map will have a real problem. it shows flat all the way.
It needs angle light to work.
Here the result, only displacement map added
Using These Maps Together
Regardless of which map you choose to use, understanding how each map works and both it’s strengths as well as it’s weaknesses will only make your decision easier. Ultimately the map you go with should be the one that best fits the needs of the scenario you find yourself dealing with.
All the maps together
Here some handy sites to help you!
Online normal map generator and displacement map generator
I cannot use it, Chrome has no Web GL
It should work with Firefox, but i cannot confirm this.
I have used "Know the difference" as reference.