Transitioning from VRay to LuxCoreRender

Discussion related to the LuxCore functionality, implementations and API.
marcatore
Donor
Donor
Posts: 463
Joined: Wed Jan 10, 2018 8:04 am

Re: Transitioning from VRay to LuxCoreRender

Post by marcatore »

well..never known this... really interesting.

What kind of problem should generate that kind of light setup?

EDIT: tested just now. It's perfect for my faking&easier lighting setup. In a super simple scene I can't see any artifacts...but if anyone else have tried before and can show what's the problem, I like to see it.
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: Transitioning from VRay to LuxCoreRender

Post by Dade »

marcatore wrote: Wed Jun 26, 2019 7:07 am What kind of problem should generate that kind of light setup?
It is prone to generate more fire flies because the light source can be sampled only with direct light sampling (while you can use both direct light sampling and direct light hits of the eye path).
Support LuxCoreRender project with salts and bounties
marcatore
Donor
Donor
Posts: 463
Joined: Wed Jan 10, 2018 8:04 am

Re: Transitioning from VRay to LuxCoreRender

Post by marcatore »

If I have a bit of time I'd like to test in a project I've done and see the result.

thanks for the info.
happyboy
Posts: 63
Joined: Sat Jun 22, 2019 5:39 am

Re: Transitioning from VRay to LuxCoreRender

Post by happyboy »

Dade wrote: Wed Jun 26, 2019 8:06 am It is prone to generate more fire flies because the light source can be sampled only with direct light sampling (while you can use both direct light sampling and direct light hits of the eye path).
Can you elaborate a bit? I don't quite understand. Thanks!
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: Transitioning from VRay to LuxCoreRender

Post by Dade »

happyboy wrote: Wed Jun 26, 2019 11:31 am
Dade wrote: Wed Jun 26, 2019 8:06 am It is prone to generate more fire flies because the light source can be sampled only with direct light sampling (while you can use both direct light sampling and direct light hits of the eye path).
Can you elaborate a bit? I don't quite understand. Thanks!
LuxCore uses MIS (Multiple Importance Sampling => merge direct light sampling and direct light hit of an eye path to minimize the noise) with path tracing to compute incoming light. If the area light can not be hit by an eye path if it is not visible so you loose one of the 2 results and this can generate some problem with some kind of glossy surfaces.
Support LuxCoreRender project with salts and bounties
happyboy
Posts: 63
Joined: Sat Jun 22, 2019 5:39 am

Re: Transitioning from VRay to LuxCoreRender

Post by happyboy »

Dade wrote: Wed Jun 26, 2019 12:07 pm LuxCore uses MIS (Multiple Importance Sampling => merge direct light sampling and direct light hit of an eye path to minimize the noise) with path tracing to compute incoming light. If the area light can not be hit by an eye path if it is not visible so you loose one of the 2 results and this can generate some problem with some kind of glossy surfaces.
Ok, thanks! I understand now. But I thought MIS is only used in bi-directional path tracing... does it also affect regular path tracing like PATHOCL?
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: Transitioning from VRay to LuxCoreRender

Post by Dade »

happyboy wrote: Thu Jun 27, 2019 5:39 am Ok, thanks! I understand now. But I thought MIS is only used in bi-directional path tracing... does it also affect regular path tracing like PATHOCL?
Yes, the classic example:

Image
Support LuxCoreRender project with salts and bounties
happyboy
Posts: 63
Joined: Sat Jun 22, 2019 5:39 am

Re: Transitioning from VRay to LuxCoreRender

Post by happyboy »

Dade wrote: Thu Jun 27, 2019 7:33 am Yes, the classic example:

Image
ok, this picture explains why MIS is used in PATHOCL, but what you mentioned "merge direct light sampling and direct light hit of an eye path" looks like a different kind of MIS (which looks like connecting eye path + light path as in bi-directional path tracing)?
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: Transitioning from VRay to LuxCoreRender

Post by Dade »

happyboy wrote: Thu Jun 27, 2019 12:48 pm ok, this picture explains why MIS is used in PATHOCL, but what you mentioned "merge direct light sampling and direct light hit of an eye path" looks like a different kind of MIS (which looks like connecting eye path + light path as in bi-directional path tracing)?
I mean "direct light sampling" = "Sampling the light source" and "Eye path sampling" = "Sampling the BRDF".
Support LuxCoreRender project with salts and bounties
happyboy
Posts: 63
Joined: Sat Jun 22, 2019 5:39 am

Re: Transitioning from VRay to LuxCoreRender

Post by happyboy »

Dade wrote: Thu Jun 27, 2019 2:45 pm I mean "direct light sampling" = "Sampling the light source" and "Eye path sampling" = "Sampling the BRDF".
Oh! Now I fully understand. Thanks for your help & patience.

Now the next question: some of our customers are quite "allergic" to color bleeding, so in VRay we reduced indirect lighting's hue by setting gi_saturation=0.3. We also use gi_contrast=1.1, gi_contrast_base=0.5. I think I can do this with LuxCore by rendering indirect lighting AOVs then do some custom composition, but how to make use of OIDN? I can currently think of two approaches:

1. write an ImagePipeline plugin, manipulating AOVs before OIDN
2. LuxCore only renders to AOVs, then I write a separate program to process AOV, then call OIDN manually.

Are there any other approaches? Which one is preferred?

BTW: While I know color bleeding is physically correct, I often think the bleeding looks MORE than it should. Are there other ways to reduce color bleeding while still being physically correct?
Post Reply