"Bryce cannot have it's mirrors reflect light off itself." I was really disapointed when I learned that Bryce would not do that and I keep thinking it's a trick that would be nice to figure out, as I'm not really satisfied faking it; but once found it would lengthen render times even more. Bryce rendering is already way slow and if it weren't for being able to set rendering to the background (just minimize Bryce while rendering) while I'm doing other things like checking out the net,sleeping or away I'd be at wits end with it. So the compromise between render time and image quality is always in play. Every once and a while though I get the idea in that in Bryce the light angles aren't always calculated right; I noticed this once with sunlight actually getting the visable lightray effect as described by Peter Sharpe but with the sun - about 30 or more degrees from where one might expect (a picture showing this is archived somewhere on this forum on some thread I started some while ago - something about faking trees using spikes) . . . so there might be some bug or feature . . . anyrate I keep trying new things with light . . . but then in all likelyhood someone would have found mirror lights by now if such existed. As far as refraction goes, Bryce appears to try to emulate the real world closely. Although a completely opaque surface won't show or use refraction(at least not in the real world and in my observations not in Bryce either) , making things only slightly transparent might allow Bryce to begin to work with it. It might be a worthy experment to see at what point (percent transparency) an object begins to show differences in the way it looks due to refraction in Bryce. - TJ BTW: Refraction is only noticed at the interface(surface) between two objects with noticably different indices of refraction such as (approximately): air/glass (1:1.6 ) air/water (1:1.3) glass/water (1.3:1.6) but it has to work by light actually going through the object. It is a confusing subject but once learned not easily forgotten.