Wrong calculaton of Lightbreaking in Octane?

[cfrank78]

Hello everybody!

I experimented with a logo from a company called Video2Brain, which is made of 3 glass meshes. A red, green and blue glass plate. RGB. I layed one glass plate above the other and put a 6500Kelvin light source behind the logo.

In reality, when you do that, the light gets broken and for example the red above the green glass will become yellow. Ass usual in physics. I tried this in reality. I took a 6500 K Hensel softbox, placed the glassplates like in the logo and it looked like in the picture called LOGO 2.

When you have a look at the picture "Logo 1", the png file, you can see that octane does not calculate the colours as it should! I took like in reality 3 glass plates, specualr material, ior 1,44, transmission "pick colour" from the logo (pure rgb values = r 255, g 255 and blue 255), reflection white, dispersion 0, roughness 0! Also experimented with fake shadows on/off! No success!

I tried it with direct light, pmc and pathtracing, but i dont get accurate results. Am i doing something wrong?

As a little hint. I placed the same mesh in Arion and arion calculated the results correctly! I really hope i do something wrong and you guys can help me. I added the OCS as a rar file for your convenience!

Kind regards Chris!

logo2.jpg (28.45 KiB) Viewed 8153 times

v2blogo.rar

(61.2 KiB) Downloaded 180 times

[Proupin]

Include .obj please

[cfrank78]

Thanks Proupin!

Completely forgotten. Added now in v2blogo.rar in the first post!

Kind regards Chris!

[MaTtY631990]

Interesting, I'm not sure how to configure properly exactly but just try overexposing the scene and you will find the colors work, Kind of. Was Arion render light and camera settings matching the settings as octane.

[cfrank78]

Interesting, I'm not sure how to configure properly exactly but just try overexposing the scene and you will find the colors work, Kind of. Was Arion render light and camera settings matching the settings as octane.

Arion uses real life camera settings and they matched my DSLR settings. The result looked pretty the same. I tried to match the settings in octane as good as possible. I also tried to overexpose - yes yellow was right, but black not and also the blue became magenta, which is not correct too! I tried to overexpose my light (hensel) and the blue never became magenta, just an overexposed blue! Somehow i can not get accurate complementary colours!

Something very weird is going on here!

Chris!

[BorisGoreta]

I don't know what is the problem with your scene but I have constructed mine and I think it works fine. Materials are default specular with different transmission colors. Kernel is PMC. I added imager node and set camera response to one of the cameras.

[cfrank78]

boris yours is wrong too. physically correct would look like "logo2.jpg", like in the first post!

[Proupin]

I was puzzled when you said you tried this in real life and it was like in the logo... as far as I know, colored panels block light, the red one only lets red go through etc... and each layer acts in a substractive manner. I don't see why superposing several colored layers would add instead of substract. I believe you though! just don't see how you achieved this (maybe upload the real picture, with the setup?)

[JimStar]

boris yours is wrong too. physically correct would look like "logo2.jpg", like in the first post!

Physics is crying looking at "logo2.jpg".

The colored glass passes through only the light of this color. So, if you look at the light through green colored glass - your eye gets only the green portion of light from the source. If after it you place the red glass between your eye and green glass - this glass passes through only RED light. But the light coming from the green glass is already filtered and contains only the green portion of light. And this green light does not pass trough red glass, so - your eye does not get almost any light from these two glasses, so you see black.

But. If you will project the lights from green, blue and red glasses on some white surface, in this case they will add to each other. As the full not filtered range of light frequencies enters each glass, so they will produce their own colors each, and the sum of these will give you the white...

[Karba]

boris yours is wrong too. physically correct would look like "logo2.jpg", like in the first post!

Really?!

What about that?






Areon is shit.