What are normal maps?

AnotherUserName

What do they do? Do they increase render times? In what situations would I use them?

mjc1016

Basically, they are another type of bump map...they add surface details that don't need to actually cast shadows.  One that includes more information about the details than a simple greyscale bump map.  You would, depending on how they are made, use them as a replacement (or they can be made to provide different details) for bump maps.  And if they do increase render times, it's in the fractions of a second.

AnotherUserName

mjc1016 said:

Basically, they are another type of bump map...they add surface details that don't need to actually cast shadows.  One that includes more information about the details than a simple greyscale bump map.  You would, depending on how they are made, use them as a replacement (or they can be made to provide different details) for bump maps.  And if they do increase render times, it's in the fractions of a second.

They increase detail? So I would use normal maps for closer shots and standard bump maps for longer shots?

mjc1016

No...use them anywhere you would use a regular bump map.  They have not just height information, but lighting information, too.  It's a 'more accurate' bump map.

AnotherUserName

mjc1016 said:

No...use them anywhere you would use a regular bump map.  They have not just height information, but lighting information, too.  It's a 'more accurate' bump map.

So if they dont really affect render speed, is it more of just a use them if you have them sort of situation?

mjc1016

Yes...they are a more accurate/detailed type of bump mapping.

AnotherUserName

mjc1016 said:

Yes...they are a more accurate/detailed type of bump mapping.

Thanks mjc1016, I appreciate the help.

mikael-aronsson

To be picky they dont add any detail, they add the apperance of more detail, the surface is still flat so just like a bump map the edges will be flat if you get to close or look from the side, it only works with small "bumps", the higher the bumps the more fake it will look,

And no, they do not contain any lighting information, they just contain a normal mapped to RGB values, the advantage is they do have a unique surface normal for each pixel in the texturemap instead of just a simple "bump" height value like you have with a bump map, and that gives more accurate lighting.

 

 

nicstt

AnotherUserName said:

mjc1016 said:

Basically, they are another type of bump map...they add surface details that don't need to actually cast shadows.  One that includes more information about the details than a simple greyscale bump map.  You would, depending on how they are made, use them as a replacement (or they can be made to provide different details) for bump maps.  And if they do increase render times, it's in the fractions of a second.

They increase detail? So I would use normal maps for closer shots and standard bump maps for longer shots?

Normal maps should take less processing power to deal with; bumps/displacement take more (think both do - one certainly does) so they extend rendering times, which normals don't.

StratDragon

you can also scale an object with a normal and you don't need to adjust any values. the map dictates the strength of the effect on the model, there are no numbers to tweak. While this a potential problem is some instances it take a lot of guesswork out of rendering in other apps.

mjc1016

nicstt said:

 

Normal maps should take less processing power to deal with; bumps/displacement take more (think both do - one certainly does) so they extend rendering times, which normals don't.

They take the same as bump maps. 

 

mikael-aronsson said:

 

And no, they do not contain any lighting information, they just contain a normal mapped to RGB values, the advantage is they do have a unique surface normal for each pixel in the texturemap instead of just a simple "bump" height value like you have with a bump map, and that gives more accurate lighting.

 

If you want, ok...but

the unit vector from the shading point to the light source is dotted with the unit vector normal to that surface, and the result is the intensity of the light on that surface.  (https://en.wikipedia.org/wiki/Normal_mapping)

 

jestmart

That is not lighting information it is directional information derived from the vectors that influences the cast shadows appearance.