Comments on: Rendering normals in eye space http://www.gamerendering.com/2008/09/25/rendering-normals-in-screen-space/ Thu, 29 Jul 2010 23:27:20 +0000 http://wordpress.org/?v=2.8.5 hourly 1 By: admin http://www.gamerendering.com/2008/09/25/rendering-normals-in-screen-space/comment-page-1/#comment-45 admin Sun, 22 Feb 2009 17:18:47 +0000 http://www.gamerendering.com/?p=25#comment-45 Yes, you're correct, it's wrong and I will fix it. Thanks for pointing it out. I think it was the wrong code I posted. Was a while since I wrote it so I don't remember exactly what this code should do. Maybe normals in screen space? =/ Yes, you’re correct, it’s wrong and I will fix it. Thanks for pointing it out.

I think it was the wrong code I posted. Was a while since I wrote it so I don’t remember exactly what this code should do. Maybe normals in screen space? =/

]]> By: Max http://www.gamerendering.com/2008/09/25/rendering-normals-in-screen-space/comment-page-1/#comment-43 Max Fri, 20 Feb 2009 04:44:12 +0000 http://www.gamerendering.com/?p=25#comment-43 Did you mean: "gl_NormalMatrix" instead of "gl_ModelViewMatrix"? If I understand you correctly, you want to move normals from object's coordinate system to eye coordinate system. The transformation from one to another, transforms only points and tangent vectors correctly (the later is true as long as it's linear), however normals won't be transformed correctly this way. One easy explanation is (M is 3x3): cross(M*v1,M*v2)==transpose(adj(M))*cross(v1,v2)!=M*cross(v1,v2) Actually there will be equality iff M is special orthogonal. Or maybe you wanted to do something else, and I totally misunderstood you? Did you mean: “gl_NormalMatrix” instead of “gl_ModelViewMatrix”?

If I understand you correctly, you want to move normals from
object’s coordinate system to eye coordinate system.
The transformation from one to another, transforms only points
and tangent vectors correctly (the later is true as long as it’s
linear), however normals won’t be transformed correctly this way.
One easy explanation is (M is 3×3):
cross(M*v1,M*v2)==transpose(adj(M))*cross(v1,v2)!=M*cross(v1,v2)
Actually there will be equality iff M is special orthogonal.

Or maybe you wanted to do something else, and I totally misunderstood you?

]]> By: admin http://www.gamerendering.com/2008/09/25/rendering-normals-in-screen-space/comment-page-1/#comment-39 admin Tue, 20 Jan 2009 00:01:52 +0000 http://www.gamerendering.com/?p=25#comment-39 This solution does not work with normal mapping if that's what your asking for? You can read about varying variables here if that's what your wondering about. <a href="http://www.lighthouse3d.com/opengl/glsl/index.php?varying" rel="nofollow">http://www.lighthouse3d.com/opengl/glsl/index.php?varying</a> This solution does not work with normal mapping if that’s what your asking for?

You can read about varying variables here if that’s what your wondering about.
http://www.lighthouse3d.com/opengl/glsl/index.php?varying

]]> By: Matt http://www.gamerendering.com/2008/09/25/rendering-normals-in-screen-space/comment-page-1/#comment-38 Matt Mon, 19 Jan 2009 23:48:15 +0000 http://www.gamerendering.com/?p=25#comment-38 How are the 'varying' variables dealt with? There are normally a different number of fragments to vertices, so how does the pipeline know what the value of (in the above case) 'normal' is? How are the ‘varying’ variables dealt with? There are normally a different number of fragments to vertices, so how does the pipeline know what the value of (in the above case) ‘normal’ is?

]]>