Interior rendering tests Tweet

[glimpse]

If you take into the comparison the cost of one i7 CPU and fairly fast GPU then I think Corona would be a winner here. Still I prefer Octane for it's interactivity and quality. Corona's interactive viewport is going to be implemented in the future so maybe then it can be usefull as Octane

well it would be actually a nice test using top of the shelf 8core i7 (for 1000$) vs TitanBlack (for the same 1000$) - personally, I would not bet for CPU..

Here we don't try to make overall Corona Vs Octane case, just comparing some speeds..-both engines have their own strengths & usage scenarios =) The best thing is that You can use them side by side rendering two projects on one machine simultaneously!

[bbeepp]

...
well it would be actually a nice test using top of the shelf 8core i7 (for 1000$) vs TitanBlack (for the same 1000$) - personally, I would not bet for CPU..

Well, we'll have to ask someone who owns them both

[smicha]

12k DL - same as previous settings, 20 minutes.

[bbeepp]

12k DL - same as previous settings, 20 minutes.

Suuuuuper fast

[smicha]

4k PT, 30 minutes, 6k samples

[aoktar]

Hi guys,
Corona is really good renderer on many aspects. But you should learn the our tricks and use better the Octane
.. my tests.

Oaktar, Thanks for Your input!

Actually I don't even think that Corona could be faster, so I was curious to make a test with Octane with the same scene too.
It's unbiased engine & on GPUs these should render faster then on CPUs, considering available compute power.

P.S. what kernel You've used for nigh shot? =) it came really pleasant.

Hey,
Common pitfall is to compare the different concept. BRUTE FORCE against to GI CACHE estimations. Corona has two modes as i could see. I think it's impossible to hit GPU rendering by BRUTE FORCE methods. And i'm very curious about results of GI cache on animated meshes. Vray had this problems(GI filickering) on GI estimation methods.

NOTE: All renders are on PT mode with coherent. Night shot was harder than daylight shot. Took 12 min. on this resolution.

[aoktar]

Hi!
I would also like to ask for a link.
I'm using Octane for quite a while (3yrs to be precise) but still haven't got the courage to post anything on the forum yet. I'm still not satisfied with the results. Mainly due to lack of time to develop my skills. But this scene would be a great start for faster development

Octane has fastest feedback to see what are you doing. Also has many option to can achieve good results. So you should check more forum posts and do more tests. I can't distribute this scene because it's from Corona web page. Only i can show and reply questions.
I hope you have good gpus

[bbeepp]

...Difference is some camera imager parameters and little octane post effect. Second image shows the most parameters.

I have noticed "Highlight compression" setting. Is it new in OctaneRender 2.1?

[smicha]

yes

[glimpse]

Hey,
Common pitfall is to compare the different concept. BRUTE FORCE against to GI CACHE estimations. Corona has two modes as i could see. I think it's impossible to hit GPU rendering by BRUTE FORCE methods. And i'm very curious about results of GI cache on animated meshes. Vray had this problems(GI filickering) on GI estimation methods.

I though Corona is BRUTE FORCE..so, this given scene, how it was made to render?
need to do more homework, as it's not fair to compare apples to oranges then..