Radiosity normal mapping as implemented in Resistance

[msxyz]

Does anybody have a link to a whitepaper/presentation of the method used by Insomniac to implement radiosity normal mapping in the Resistance: Fall of Man game?

Radiosity normal mapping, as implemented in the Source engine, uses three lightmaps, calculcated along three perpendicular axis. In Resistance, they do away with a single texture which isn't really a lightmap, as it stores in the texels the direction of the light rays. What about light color? Do they use a second texture, which is just a matrix of RGB scalar values? Any info would be appreciated.

PS: Sometimes I feel I'm still living in the stone age of 3D computer graphics (and, on top of that, I haven't really coded anything since the days of the fixed functions cards)

 

[Davros]

You thing your living in the stone age. i havnt done anything since the days of putpixel and getpixel

 

[msxyz]

You thing your living in the stone age. i havnt done anything since the days of putpixel and getpixel

The first 'personal' computer I ever used didn't even have a screen, only a 32 character DOT line display! It was a scavenged HP9825, a machine that I still own and that still works! (There are a couple of flaky 1Kbit Intel PMOS DRAM chips in it however... It's difficult to get working replacements for such old ICs).

I like to keep me updated on technology but sometimes I really wish we hit a hard wall on silicon technology 10-15 years ago. This would have given me time to catch up!

 

[silent_guy]

I can't be the only one who implemented radiosity algorithms in the 3KB of my VIC20?