From: Lauri K. <ca...@gm...> - 2013-09-15 05:47:58
|
On Sun, 15 Sep 2013 00:25:57 +0200 Lionel Fuentes <fu...@gm...> wrote: > To me, the only advantage of doing fog in a post-process is to reduce the > number of pixels treated (as would happen with a Z-prepass). No, it's mainly for shader complexity. Practically _every_ shader had duplicated fog logic, and it was different in some, due to obvious reasons of someone only changing some. > Being able to tune the formula for how height influences fog can be done > during lighting as well, but without the weakness you mention, so I don't > think having it as a post-process is the best way to go. It would have the exact same weakness, as transparents aren't lit, only solids are. Lights cannot separate sky from transparents either. > So I suggest we keep going with post-process for now but would benefit from > having it in the lighting pass later on. Except if the limitation for > transparent objects ends up not being too much of a burden for the > artists... As it would have the same limitation in lighting pass too, I don't see how it would improve anything? - Lauri |