problem with orthographic camera

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.
mick
Posts: 80
Joined: Mon May 21, 2018 7:57 pm

Re: problem with orthographic camera

Post by mick »

If I mark the glass rough with a roughness of 0.1 the glass shadow becomes dark for to orthographic camera.

But with a roughness of 0.05 again there is more light passing through the glass than through air. This is also true if I remove the world volume, or set its scattering to zero.

Why is there more light energy passing through glass than passing by?

BTW: The extra light is coming from passing through the glass. I tested by coloring. It's colored by volume absorption and by material transmission, not by material reflexion (just a little on the sides).

???
mick
Posts: 80
Joined: Mon May 21, 2018 7:57 pm

Re: problem with orthographic camera

Post by mick »

Idea for workaround: place a perspective camera extremely far away

All other objects are camera invisible. Air volume is clear, absorption = 0 0 0.

Will this work for reliable measurement, or will I run in other issues?
epilectrolytics
Donor
Donor
Posts: 790
Joined: Thu Oct 04, 2018 6:06 am

Re: problem with orthographic camera

Post by epilectrolytics »

Sorry to be of no help.
I tested your blend file and found the geometry with no fault.
I could not get any reasonable behaviour from the orthographic camera either.

Volumes with absorption or scattering introduce distance as a parameter.
While the perspective camera has a defined place in the coordinate system, the orthographic one is in a way a distance-less camera.
My guess is, it cannot deal with scattering or absorption because its relative distance is not defined or in a way not compatible to volumetric algorithms.

I hope you get your last approach to work, good luck!
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: problem with orthographic camera

Post by Dade »

mick wrote: Sat Feb 09, 2019 9:25 pm I created the scene in blender, more or less. At least it shows the same strange behavior for orthographic camera.
File attached.

To compare just move to camera to e.g. z=-1 or switch to Perspective.
Do you mean changing the camera position from:

Code: Select all

scene.camera.lookat.orig = 0 0 -15
to:

Code: Select all

scene.camera.lookat.orig = 0 0 -1
:?:

I'm testing your scene but changing the camera Z position produces the same rendering (as expected).
Support LuxCoreRender project with salts and bounties
mick
Posts: 80
Joined: Mon May 21, 2018 7:57 pm

Re: problem with orthographic camera

Post by mick »

I'm testing your scene but changing the camera Z position produces the same rendering (as expected).
Setting the origin of an orthographic camera close to the floor (e.g. -15) brings a circular pattern below the glass. More distant (e.g. -1) this patter is not recognizable. Where does this pattern come from? And why does distance matter for orthographic?

With perspective camera I also get a pattern:
image_persp_close.png
I guess this is the shine of the sun on the matte floor. And it's mainly visible outside of the shadow. Why is it only seen when the camera is close to the floor?

The main question still is: Why is the shadow below the glass lighter than the irradiated area when rendered with orthographic camera?
mick
Posts: 80
Joined: Mon May 21, 2018 7:57 pm

Re: problem with orthographic camera

Post by mick »

Just another test. Close orthographic camera with another glass in the opposite quarter. That glass is much thinner, hence more light and gloss from the sun. And again almost no light outside of the shadow.
image_ortho_close_2.png
It seems like the renderer is almost neglecting direct rays. (direct between sun the plane that is observed by the orthographic camera) ???
mick
Posts: 80
Joined: Mon May 21, 2018 7:57 pm

Re: problem with orthographic camera

Post by mick »

Just tested to switch the renderer to cycle or blender. Both corrects the situation: shadow becomes darker than the rest also with orthographic camera.

So I really think that there is a bug in the handling of orthographic camera in LuxCoreRender.
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: problem with orthographic camera

Post by Dade »

I don't understand. From the scene you posted I get a totally different output:

test.jpg

What do you get ?
Support LuxCoreRender project with salts and bounties
mick
Posts: 80
Joined: Mon May 21, 2018 7:57 pm

Re: problem with orthographic camera

Post by mick »

I'm using LuxCoreRender Blender plugin 2.1 just download two hours ago.

The attached file gives me the odd result shown in my previous post.

Would you please check. Are there relevant changes in your alpha version?
Attachments
plate.blend
(628.82 KiB) Downloaded 176 times
User avatar
Dade
Developer
Developer
Posts: 5672
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: problem with orthographic camera

Post by Dade »

mick wrote: Sun Feb 10, 2019 5:54 pm I'm using LuxCoreRender Blender plugin 2.1 just download two hours ago.

The attached file gives me the odd result shown in my previous post.

Would you please check. Are there relevant changes in your alpha version?
The reason I was getting different results is because I was using LuxCoreUI. The only way it affects the rendering is by overwriting the "scene.camera.screenwindow" property to match the rendering window.

The source of the problem is in that property. If you check how BlendLuxCore export the property:

Code: Select all

scene.camera.screenwindow = -5 5 -5 5
It looks wrong, I'm not sure and tomorrow I will check but it should be:

Code: Select all

scene.camera.screenwindow = -1 1 -1 1
Is your scene using some strange border render setting ?

I have yet to check where exactly is the problem but the root is in "scene.camera.screenwindow" for sure.
Support LuxCoreRender project with salts and bounties
Post Reply