Happy New Year render

[ViennaLinux]

This is my first render in the year 2008
I hope you like it.

I dont know why it takes THAT LONG to render. Its still cooking on my old Pentium4 Server (2.26 GHz and no HT). After 23 hours it made only 315 samples/pixel at 1920x1080.

In my eyes the "orange juice" is the problem. Is SSS always that slow? I think I will copy it back to my core2duo pc and continue rendering here.

[BbB]

Hi. Nice image. SSS is dead slow. All unbiased renderers have the same problem, though not all produce as nice a result as Indigo does. I don't use it if I don't have to.
High bump values on surfaces that occupy a large section of the image will also slow you down, as will lots of glossy surfaces, light through speculars, complicated emitters or a large number of them.

[ViennaLinux]

w00t i have all in one picture .. lol ^^
okay I will remember that.

why do you think to bump maps influence render times? the just change the surface normals - okay mabe its harder then to hit the light source etc. but complicated bump maps should work without performance loss I guess.

[BbB]

That's above my paygrade I just know they do. Maybe one of the techies could answer this one...
(Is it perhaps because they send light rays fly all over the place?)

[Kram1032]

Just a thought...
Many many lightsources shouldn't really affect render speed negatively. Actually, they should speed things up:
Lightsources currently are like black walls (perfect absorbers, which kill every single ray), which then colour the light, so that the pixel (or sumpersampled part of a pixel) gets it's final colour.
So, they actually should be nearly as fast as an absolutely black material.

Ah, and you forgot that, BbB.

White walls also slow things down, as well as any transparent/translucent stuff with low absorption. Every ray, that gets sorted out, when being useless anyway, will make things faster. Every ray, that gets sorted out, although it's next bounce would reach a lightsource, cost some detail, but it's still faster.