Page 1 of 1

Interior Tests/Benchmarks/Technical Stuff

Posted: Tue May 27, 2014 4:02 pm
by RobSteady
Maybe we all can post ideas or insight on interior setups, lighting, technical stuff etc.
A good thing could be a default scene (maybe a day and night setup), so we can find the best speed/quality ratio.

For example I prepared this one to find out if portals could be useful, even for rooms with bigger openings:

This is the scene. A simple room with three windows and only lit by white environment.
Scene.PNG
PT9 With Portals 1000s/px 6,8s/s 5min
PT9_portals_1000spx_6.8sps_5min.png
PT9 No Portals 2100s/px 14,3s/s 5min
PT11_noportals_2200spx_14.3sps_5min.png
PT9 No Portals 1000s/px 14,3s/s 2min20
PT11_noportals_1000spx_14.3sps_2min20s.png

This is a 200% crop
200_zoom_detail.png

Re: Interior Tests/Benchmarks/Technical Stuff

Posted: Tue May 27, 2014 4:22 pm
by acc24ex
nice one, this looks very usefull, this should be on some sort of wiki page, or a sticky thread
- and apparently the samples/s doesn't mean much, speed, but not divergence

- so the portals actually work well now? - too bad I'm still on 1.2 because of old scenes..

Re: Interior Tests/Benchmarks/Technical Stuff

Posted: Tue May 27, 2014 4:55 pm
by kavorka
Interesting. I never used portals because my tests showed the Ms/s wasnt advantageous. So maybe I will have to retest it.