By slow i mean a cornell box will be 200+ hours if you have enough objects in it.
the thing is it will give you Every light effect spot on accurately (you can’t turn them off infact)
caustics, and shadows are unavoidable parts of the design, you can’t render in it without them. lights need to have a form they don’t exist simply as a point in space, they exist as only an object (sphere or plane or whatever)
also effects like correct spectrum splitting (over the entire range, not just rgb splitting), and the effect where colours go blurry around the edge when looking through a lens co of their different refractive indexes…
The renderer is super damn spot on… if you have weeks to wait for each render.
it renders from the Phototon (not photon mapping or aproximation, it does every single photon), and renders starting from the light source, and not from the camera, hence all its results are correct.
if i were the coder of the software, i would design it to remember the data from multiple camera angles at once, since it would be no extra rendering time to do so. (perhaps it could, i dunno, maybe thats a feature request)
Jaycun (and anybody else curious about it):
thanks to the link provided by Alltaken, I’ve looked around in the galley and found a page with a picture of a scene with a list of links to shows comparison of different rendering by different softwares of the same scenes.
IanC: I’ve never had much luck with Radience either. Never got it running.
Alltaken: Yeah…ummm…200+ hours is a little too long for the type of stuff I am intending. My parents went crazy with a 68hr render time…I rather not risk a 200hr render time (and summer is already hot enough w/o the CPU )
Dittohead: Yes…dispersion and absorption. I want to be able to replicate most of the well-known optical phenomena (caustics, dispersion, absorption, polarization, total internal reflection, etc.) I recently looked at a list of features for Yafray and I discovered that most of these can be simulated with Yafray (hence the question of yafray .9 above). I want it to be as accurate as possible b/c it might be used in a classroom setting (don’t want obviously faked results, the effects need to be accurate enough to point to and explain as something that would happen in real life).