photon mapping in halo 2?

BirdMadBoy

Newcomer
Hi all, i´ve read in this article http://www.unlimitedlives.com/special_reports/ul_spec_halo2-3.html that halo 2 will use photon mapping, an extension of ray tracing, instead of radiosity, and in the second page of the article they talk about using stencil shadows ( similar method to doom 3? ). phothon mapping is used for static lightning only and is pre-calculated? as i´ve read for radiosity but being faster, or will it be real-time? and how would it interact with the shadows/dinamic lightning? or how would it work in exteriors, open zones? what differences would be with doom 3 lightning?
 
If it's going to be done in real time I should hope that everyone has a VERY fast videocard/cpu. :)
 
so if its not possible to be done in real-time on xbox, how are they gonna make use of it? how will it interact with the shadow/lightning system?
 
I think Halo used radiosity alorithms to calculate it's light maps (not realtime of course).
The switching to photon mapping in the sequel.

PM has the advantage of properly calculate with light bouncing off shiny/mirroring surfaces. (That is it can do caustics.)

Stencil/per-pixel lighting can calculate direct light only - the good way to combine these schemes it to store bounced light only in lightmaps.

It won't have bounced lights for dynamic lightsources, but I think it would be too much to ask.
 
We use photon maps as well, it's no big deal. Actually it's quite a bit easier to implement than radiosity. It's just another way to compute light maps. You basically shoot out a photon from your light source and let it bounce around and stick to the scene. You shoot out tons of them, then combine the photons which are stuck to surfaces into lightmaps. Cool thing is you get things like caustics for free.

A guy here is working on keeping a rough copy of the photons around to modify the lighting in realtime.
 
As a side note, I was considering attempting to find an analytical way to develop caustics from a fourier representation of a reflective surface. I never got the math to work out properly, but, if it could be done well, it would certainly provide for some impressive effects for water surfaces (above and below water).
 
Back
Top