Maxwell Render | Official public forum

Just by pure coincident I came about the following problem and did some tests.

I have a sphere with around a 10.000 LEDs on it, loaded with an HDRI Image. The sphere is in a closed box.

seize of the sphere is 3 meters (diameter), all units are meters so the scene is right.

The thing is, if I turn physical sky and the sun on, the scene renders much faster as when rendering with no environment at all, what you usually would set, when you are inside a closed box;)

I uploaded the results (screenshots of the console), reulting images, mxs.file and the rhinofile.

http://rapidshare.com/files/192130023/t ... em.7z.html

I did some more test, rendering straight out of studio, just to check that it is not an rhino plugin issue.

in short:

A. exporting from rhino - sun on - SL8 after 5m11s BM 75.9
B. exporting from rhino - sun off - SL8 after 6m39s BM 59.2
C. rendering from studio - sun on - SL8 after 5m10s BM 76.2
D. rendering from studio - sun off - SL8 after 7m34s BM 52.0

the ruslting images when the sun is off seem to be a little less gainier, but no big difference!!!

CONCLUSION:

turning on the physical sky system, although beeing inside a "black box" seem to speed up things by 25% !!!

feel free to test;)

when you're saying that the images with sun are a little grainier ... you should run the test again and let all images render for the same time, independently of its sample level. this should give you a real hint whether turning the sun on really speeds things up (which should result in a better image quality) or if it just boosts the sample level.

thx kami

that is a good guess, I let you know.

maybe I give it 30 mins of render time;)

so I ran another test for 30 minutes.

the rendering with sunsystem on seemed to boost the sampling level, but the overall result was better for the rendering without any sunsystem.

so it is just as you supposed kami;)