Chaos Corona Forum

Chaos Corona for 3ds Max => [Max] Feature Requests => [Max] Resolved Feature Requests => Topic started by: lacilaci on 2013-03-28, 08:29:20

Title: Point Cloud Approximate GI occlusion color bleeding
Post by: lacilaci on 2013-03-28, 08:29:20

would it make sense, or is it even possible to implement such stuff to corona as an alternative to pathtracing??
It's been in PRman and 3Delight for awhile now and the results are fast and clean altough not unbiased..

http://www.cgsociety.org/index.php/CGSFeatures/CGSFeatureSpecial/pixar_points

Title: Re: Point Cloud Approximate GI occlusion color bleeding
Post by: Polymax on 2013-03-28, 09:08:32

Not a bad idea, well, I think it's after to finish the development.

Title: Re: Point Cloud Approximate GI occlusion color bleeding
Post by: Ludvik Koutny on 2013-03-28, 09:24:26

I don't think it would merge well with raytracing based renderer. I might be wrong, though ;)

Title: Re: Point Cloud Approximate GI occlusion color bleeding
Post by: lacilaci on 2013-03-28, 10:14:43

I don't really know that too, and I'm not sure how does that really differ from IRmapping or Keymasters implementation of Irradiance caching that was in corona before... Or maybe some HDcache mutants :D lol one for primary GI maybe to do some more detailed high frequency details/lighting other to do a base rough computation if that's possible..

Now you guys might not like this :D but something like mental ray is doing when used GI+FG or IP+importance driven FG yields really nice convincing lighting extremely fast. The only problem that I think could appear when something like this was present in corona, would be how to minimize the amount of controls for this to be able to get good results..

Title: Re: Point Cloud Approximate GI occlusion color bleeding
Post by: Ludvik Koutny on 2013-03-28, 10:39:10

I have never ever seen IP+Importance driven FG to show any good results. Could you post any example?

Title: Re: Point Cloud Approximate GI occlusion color bleeding
Post by: lacilaci on 2013-03-28, 11:28:27

No I can't  :) And I really don't want to get on a mr discussion with you, especially with you :D lol... So please let me apologize and I'm taking back that statement about convincing mental ray results of IP+FG...

The point however remains... Despite the mental ray's approach on the topic, the idea itself might not be wrong.. Or is it?

Title: Re: Point Cloud Approximate GI occlusion color bleeding
Post by: Ondra on 2013-05-30, 21:43:15

will not be implemented in forseeable future