question

Use this forum for general user support and related questions.
Forum rules
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

question

Post by daros » Wed Oct 21, 2020 9:21 am

Hi to all, we are integrating Luxcore in to www.felixrender.com with the purpose to offer a bidirectional rendering tool for architects at reasonable rendering speed.
We have a problem with the attached scene.
Rendering on luxcore 2.4 is much slower as on Luxcore 2.2 and a wired artifact is showing up on the ceiling, in the upper left corner of the attached image. In theory this slightly brighter area on the ceiling should not be there.
Can someone please explain us what is wrong with this scene?
Thanks
Attachments
2.zip
(399.22 KiB) Downloaded 18 times
artefacts.jpg
Last edited by daros on Wed Oct 21, 2020 12:12 pm, edited 1 time in total.

User avatar
B.Y.O.B.
Developer
Posts: 3872
Joined: Mon Dec 04, 2017 10:08 pm
Location: Germany
Contact:

Re: question

Post by B.Y.O.B. » Wed Oct 21, 2020 11:54 am

Curious. That bright rectangle on the ceiling even shows up when rendered in Blender after importing the PLYs, with completely different lighting.
And it even shows on a fresh plane mesh that is slightly rotated, so it's probably not related to a problem in the ceiling mesh.
Attachments
a.jpg
b.jpg
imported.blend
(1.27 MiB) Downloaded 11 times

daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

Re: question

Post by daros » Wed Oct 21, 2020 12:33 pm

Yes, probably we are writing something in the wrong way but we cannot figure out what.

daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

Re: question

Post by daros » Wed Oct 21, 2020 2:00 pm

If i turn off 4 lights sources of 5, the artifact still shows up. At the beginning i was thinking about caustics caused by the chrome material, which is the only shiny one of the scene, but no, even if i turn off the chrome the artifact still shows up.

daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

Re: question

Post by daros » Thu Oct 22, 2020 8:58 am

Apparently if we move the entire geometry 50m away from the origin the artifact vanishes. I publish our results soon.

CodeHD
Posts: 399
Joined: Tue Dec 11, 2018 12:38 pm
Location: Germany

Re: question

Post by CodeHD » Thu Oct 22, 2020 9:35 am

daros wrote:
Thu Oct 22, 2020 8:58 am
Apparently if we move the entire geometry 50m away from the origin the artifact vanishes. I publish our results soon.
I did have a similar artefact the other day, also with a cube object illuminated causing square artefacts that dissapeared with distance. I think it was GPU only in my case but not sure. I don't have time anymore right now but I can maybe provide a simple test scene later.

daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

Re: question

Post by daros » Thu Oct 22, 2020 9:53 am

Interesting, let's see. Thanks.

daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

Re: question

Post by daros » Thu Oct 22, 2020 10:09 am

Hi, sorry for the dumb question: Is it normal in LuxCore 2.4 that if you use completely white materials, for a closed space scene, the engine do not converges anymore correctly and overexposes strongly the final image?
Attachments
View 01 900x450.png
View 01 900x450b.png

User avatar
Dade
Developer
Posts: 4817
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: question

Post by Dade » Thu Oct 22, 2020 10:17 am

B.Y.O.B. wrote:
Wed Oct 21, 2020 11:54 am
Curious. That bright rectangle on the ceiling even shows up when rendered in Blender after importing the PLYs, with completely different lighting.
And it even shows on a fresh plane mesh that is slightly rotated, so it's probably not related to a problem in the ceiling mesh.
It is also BIDIRCPU only problem (inside eye pass) while the scene works fine with PATH(CPU or GPU).
Support LuxCoreRender project with salts and bounties

daros
Posts: 43
Joined: Thu Dec 12, 2019 3:25 pm

Re: question

Post by daros » Thu Oct 22, 2020 10:38 am

We did one test using Bidir+Metropolis+Reinhardt and the artifact is gone. We would prefer to use luxlinear.
Attachments
View 01 600x300.png

Post Reply