Timeline for How to get pixels affected by a light in deferred shading?
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 28, 2013 at 15:39 | vote | accept | danijar | ||
| Mar 27, 2013 at 21:50 | comment | added | danijar | Thanks for your explanations, though I realized that deferred lighting is a hard fake. | |
| Mar 27, 2013 at 21:11 | comment | added | Nathan Reed | Well, once you've already shaded the pixels in front of the light volume you can't un-shade them, so drawing the light a second time doesn't help. As for pixels inside the light geometry that are occluded by walls - now you're talking about shadow casting, which is a whole other can of worms. It can be done using shadow maps for spot lights, and omnidirectional shadow maps for point lights. But that's really a completely separate issue. | |
| Mar 27, 2013 at 20:50 | comment | added | danijar | What comes to my mind for excluding pixels in front of the light volume is to draw the sphere with back face culling again and use the opposite depth test. But how can I prevent shading pixels inside the light geometry which are occluded by walls, for example? | |
| Mar 27, 2013 at 18:46 | history | answered | Nathan Reed | CC BY-SA 3.0 |