I am now looking to do a fast version of atmospheric scattering. This scattering only has to be viewable from space, not with in the atmosphere or anything like that. I have read up on Sean O’neils articles, newest being in GPU gems 2. Unfortunately when I try and run the demo, I am getting CG Compile errors. It’s running under the ARB profiles (I have a 9700 pro), but doesn’t seem to be working. If any one is interested in getting the demo it’s at Nvidias link. I’m really hesitant on trying to implement his version when his doesn’t work and I am unsure how the results aare going to look. If they are anything like his past gdnet article (which his article says it builds upon) I found it was prone to getting visual artifacts with an “unrealistically wide” atmosphere (I can post ss’s if needed). So without being able to test his new code, I was wondering if there were any pointers here on where to get started. I’ve found a plethora of different equations and constants and well at this point I figure it would be good to get some advice before getting into this. So if any one has some information or places to look, I would lvoe to hear what you have to say. Sean uses a 2 sphere approach, 1 for the planet, and one with only back faces drawn (larger for the atmosphere). Is this typically the method for drawing it from space?

Yeah, that site for Atmosphere from Space just has the Sean paper I talked about and the 1993 paper it was based off of.

Any one able to get his code to work? Or have any suggestions?