Blue glass caustics less intense

Post your tests, experiments and unfinished renderings here.
Post Reply
Philstix
Posts: 13
Joined: Sun Feb 04, 2018 8:48 pm
Location: Brisbane, Australia

Blue glass caustics less intense

Post by Philstix » Wed Feb 21, 2018 2:40 am

While creating a scene in Blender with glass caustics (using BlendLuxCore 2.0alpha5) I noticed that it was difficult to produce intense caustics from blue glass.

I decided to do a test comparing the caustics produced by applying different colours to the same glass object. I've attached the result, as well as the .blend if anyone is interested.

All of the coloured jars have HSV saturation of 100. The top row has HSV value of 100, while the bottom row has HSV value of 50. I've also included a clear glass jar in each row, again with HSV values of 100 and 50.

The test does seem to confirm that blue glass does not produce caustics of the same intensity as other colours. Now I'm wondering why. Is that physically accurate, or is it a limitation of the RGB colour space used for the image file output, or is it perhaps an artifact of the algorithms used to produce the caustics?

(Hey, at least I got an interesting picture out of it :D )
2018-02-21_12-23-55.png
Render result: comparison of caustics vs colours
2018-02-21_12-23-55.blend
The BlendLuxCore 2.0alpha5 scene file
(919.4 KiB) Downloaded 101 times

User avatar
Dade
Developer
Developer
Posts: 2808
Joined: Mon Dec 04, 2017 8:36 pm

Re: Blue glass caustics less intense

Post by Dade » Wed Feb 21, 2018 10:50 am

Philstix wrote:
Wed Feb 21, 2018 2:40 am
The test does seem to confirm that blue glass does not produce caustics of the same intensity as other colours. Now I'm wondering why. Is that physically accurate, or is it a limitation of the RGB colour space used for the image file output, or is it perhaps an artifact of the algorithms used to produce the caustics?
It is likely to be more related to human eye sensitivity: we are far more sensitive to green and red than blue. Our luminance perception is not uniform at all, just look at the code to calculate luminance out of an RGB color:

Code: Select all

float Y() const {
	return 0.212671f * c[0] + 0.715160f * c[1] + 0.072169f * c[2];
}
Green is far more important than red and red is far more important the blue :!:
Support LuxCoreRender project with salts and bounties

User avatar
B.Y.O.B.
Developer
Developer
Posts: 2809
Joined: Mon Dec 04, 2017 10:08 pm
Location: Germany
Contact:

Re: Blue glass caustics less intense

Post by B.Y.O.B. » Wed Feb 21, 2018 4:13 pm

I agree with Dade.

If you pick the colors and look at the HSV values, you'll notice that both green and blue are at V = 100 in your image, so both have the same brightness.
Attachments
sdfsdf.png
Support LuxCoreRender project with salts and bounties

Asticles
Donor
Donor
Posts: 150
Joined: Thu Jan 11, 2018 8:52 am

Re: Blue glass caustics less intense

Post by Asticles » Wed Feb 21, 2018 6:51 pm

Modern screens also have blue light attenuation, for eye care and such.
Salvador Ureña

Philstix
Posts: 13
Joined: Sun Feb 04, 2018 8:48 pm
Location: Brisbane, Australia

Re: Blue glass caustics less intense

Post by Philstix » Wed Feb 21, 2018 10:37 pm

Thanks for the responses.

@Dade, yes I think you're right about this being a perception issue. BYOB's observation demonstrated that well.

@BYOB, good idea about checking the colour values numerically, I should have thought of that. The result is even more dramatic when getting the colour values in the faint fringes of the caustics. Some parts of the blue fringes are barely visible to me, but the HSV value numbers are about the same as the corresponding green fringes that seem bright.

And, as Asticles pointed out, that perceptual difference is probably magnified by the filtering in display screens.

I started looking online for information about caustics, but I quickly found myself wallowing in research papers about catastrophe optics using integral mathematics. A bit too much for me at the moment. :lol:

Post Reply