light + attenuation + scissor

hello all.

Well, I have been reading about the scissor optimization technique (Mr.Lengyels article in gamasutra and Mr.Everitt’s demo on nvidia). I just cant figure out whats happening. If someone could point me to a more detailed explanation on this topic I would be grateful.

I realize that the lights’ boundaries have to be found to cap the shadow within this volume so as to avoid unwanted fill. but how exactly are you guys doing it, somehow all of it does not make a connection in my head. Could someone please explain?

Thank you.

I think the basic idea is:

Project the intersection of the object shadow volume and the light bounding volume into screen space. Set the scissor rectangle to that. Draw the (possibly infinite) shadow volume. The scissor saves on fill rate for the cases where the shadow volume extends beyond the light bounding volume in screen space. Note that you can cap however you want; you don’t use the intersection to determine capping.