Page 1 of 1

Another pre 2.3 light test

Posted: Tue Aug 17, 2010 11:06 am
by Bulwerk
300,000k poly, one light. Rendered on my Gtx 465 and getting .62 Megasample/sec. at 2560x1600
Used Defly on it and that's it.

I tried rendering the scene again with more light sources at different colors to see how the light would blend. But really could only get color based on light temperature(blue and yellow hues). Any way to emit light say as violet and or green?

Re: Another pre 2.3 light test

Posted: Tue Aug 17, 2010 5:52 pm
by DayVids
Use texture emision

Also, this is a way to fake a shadeless material if you're using Direct lighting and not path tracing.

Dave

Re: Another pre 2.3 light test

Posted: Tue Aug 17, 2010 7:52 pm
by tomas_p
Not bad.

How long?

Re: Another pre 2.3 light test

Posted: Tue Aug 17, 2010 8:11 pm
by justix
Bulwerk wrote:300,000k poly, one light. Rendered on my Gtx 465 and getting .62 Megasample/sec. at 2560x1600
Used Defly on it and that's it.

I tried rendering the scene again with more light sources at different colors to see how the light would blend. But really could only get color based on light temperature(blue and yellow hues). Any way to emit light say as violet and or green?
NIce, directlighting..pathtracing? render time? ;)

Re: Another pre 2.3 light test

Posted: Tue Aug 17, 2010 8:47 pm
by Bulwerk
Pathtracing my firend,
Render time, was somewhere at 5 hours. I just let it bake while i left the house and it looked pretty good when i got back. So it reached roughly 1600 samples in 5 hours

Re: Another pre 2.3 light test

Posted: Tue Aug 17, 2010 9:33 pm
by justix
Bulwerk wrote:Pathtracing my firend,
Render time, was somewhere at 5 hours. I just let it bake while i left the house and it looked pretty good when i got back. So it reached roughly 1600 samples in 5 hours
thanks...5 hours :!: should we expect such rendertime with one single emitter? OK it is 16000 samples AND 300,000k poly but It sound quite ...long? maybe i'm wrong
and your gtx460 is not that far from a GTX470 :ugeek:

Re: Another pre 2.3 light test

Posted: Wed Aug 18, 2010 3:06 am
by Bulwerk
Keep in mind that the render could have been acceptable after one hour i was just not there to see. And this render only got to 1500 samples by the way :) There is a point where baking longer will have minuscule effect after a certain point.

Re: Another pre 2.3 light test

Posted: Wed Aug 18, 2010 8:15 am
by justix
Bulwerk wrote:Keep in mind that the render could have been acceptable after one hour i was just not there to see. And this render only got to 1500 samples by the way :) There is a point where baking longer will have minuscule effect after a certain point.
Ahhh. Obviously :mrgreen: pheeww...I though the days of shortening things up were to an end...LOL!

Re: Another pre 2.3 light test

Posted: Sat Aug 28, 2010 8:20 am
by radiance
Hey,

we currently have not enabled the optimization that let's emitters with a high polycount render faster.
That's why this is so slow.

(all triangles need to be checked at each ray, so having an emitter with more than a few triangles is currently *SLOW*)
we'll add the new optimized code in the next pre-2.3

Radiance

Re: Another pre 2.3 light test

Posted: Sat Aug 28, 2010 10:32 am
by justix
radiance wrote:Hey,

we currently have not enabled the optimization that let's emitters with a high polycount render faster.
That's why this is so slow.

(all triangles need to be checked at each ray, so having an emitter with more than a few triangles is currently *SLOW*)
we'll add the new optimized code in the next pre-2.3

Radiance
very good news