If you do specular separately, then you’ll either have to double your passes, or get into trouble with shadows, and have to generate shadow volumes twice, or have to hang on to all your shadow maps for the entire frame.
I’m also not quite certain how your solution would actually get diffuse color modulate into the initial lighting pass. At a minimum, I think it would involve textures with inverse esposure compensation, which might lead to precision problems.
Hi,
I have to say, I don’t really understand all you’re saying. I’m not very familiar with the details of per-pixel lighting techniques. Yes, it would double the number of passes, but in what case do you have to generate shadow volumes twice?
Also this
Originally posted by jwatte:
I’m also not quite certain how your solution would actually get diffuse color modulate into the initial lighting pass.
[/b]
means nothing to me. Do you mean the way diffuse color is used to modulate total diffuse lighting in the current engines. Because like I said, I don’t thing that is correct anyway, it’s more like an optimization/limitation. Inverse exposured shouldn’t cause precision problems if they exploit ther full range themselves.
But of course using this technique means you have to drop out some other features on older hardware, not least because it requires one texture unit for each pass.
Edit: Removed some unrelated comments. Sorry…
-Ilkka
[This message has been edited by JustHanging (edited 01-13-2003).]