Sign in to follow this  
SaTANO

Correct normal map

Recommended Posts

I am just curious... I implemented normal mapping without any significant problem long time ago but now, when use Blender to bake normal map it was not correctly rendered. When I compared blender generated map with normal maps I found on net (I think 3ds max generate them this way) it looks like blender map is horizontally flipped I also found some educational document where map is absolutely mirrored (horizontal + vertical) which seems to be right because I don't need to modify TBN formula (otherwise I used negative binormal or tangent) So I am just asking which one is correct or if they are special for every soft normal problem

Share this post


Link to post
Share on other sites
I don't know if there is any specific "standard" normal map orientation, just adapt your program for whatever kind of map you have.

Also I should point out more importantly that those look like object space normal maps, not tangent space. You wouldn't use these with any kind of TBN formula.

Share this post


Link to post
Share on other sites
Entirely dependent on coordinate system. To the best of my knowledge there is no real 'correct' standard per se, and I've seen all three. Conversion is simply the process of inverting one or more color channels. For simplicity's sake I prefer the bottom -> top, left -> right, down -> up convention, (all from 0 to 1) but as mentioned it's all taste/your coordinate system.

EDIT:
Quote:
Original post by karwosts
I don't know if there is any specific "standard" normal map orientation, just adapt your program for whatever kind of map you have.

Also I should point out more importantly that those look like object space normal maps, not tangent space. You wouldn't use these with any kind of TBN formula.


Nah, those are tangent space. Object-space normal maps tend to have a lot of areas that are not bluish whatsoever due to not all the normals pointing along the positive Z direction. These don't have any bright yellow regions, (positive X/Y and negative Z) meaning it's highly unlikely.

Share this post


Link to post
Share on other sites
You may be right about the tangent/object thing, just all of the tangent space maps I've seen look like 99% blue (like so). I've never seen one that was bright colored like that. But I haven't spent much time with them so I'll defer to someone who knows more about it.

Share this post


Link to post
Share on other sites
Quote:
Original post by karwosts
You may be right about the tangent/object thing, just all of the tangent space maps I've seen look like 99% blue (like so). I've never seen one that was bright colored like that. But I haven't spent much time with them so I'll defer to someone who knows more about it.


It's just something that comes with practice. For reference, there's a comparison between an object-space and tangent-space map here, which coincidentally looks to be the same as the one you posted above. Note the *extremely* colorful appearance of the object-space normal map as opposed to what's in the OP.

For what it's worth, it seems you've only been exposed to mediocre tangent-space maps, too :) It's possible to have lots of variation in orientation/curvature on a surface, but sadly people just tend to toss the diffuse map into a normal map generator and call it a day.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this