Page 1 of 1

Correct method for colored glass?

Posted: Tue Jun 16, 2015 5:50 pm
by itsallgoode9
When creating a colored glass, is the correct way to create it by putting a color in the Transmission channel or leaving the transmission channel white putting inverse of the desired color in an absorption node?

Re: Correct method for colored glass?

Posted: Wed Jul 01, 2015 3:54 pm
by itsallgoode9
Anybody from octane chime in one this? I've read the manual for both entries but it doesn't make it clear if one is more "correct" than the other. in my limited testing, I didn't see a major difference between the two but maybe my tests just weren't setup in a way to show the differences.

Re: Correct method for colored glass?

Posted: Wed Jul 01, 2015 5:20 pm
by ristoraven
I don't know is there a "correct way" to do anything in CGI. I think there are only ways to do something and then better ways to do the same.

Re: Correct method for colored glass?

Posted: Wed Jul 01, 2015 5:28 pm
by oguzbir
Correct method should be giving color to the absorbtion. Which outputs negative color as you would know.

But in vray and such we tend to use giving color to refraction.
You might do some tests either way. And investigate which way does works best in terms of lighting (colored shadows) and such.
I have no definite answer

Re: Correct method for colored glass?

Posted: Wed Jul 01, 2015 6:56 pm
by itsallgoode9
I did a new test that is more extreme than what I was doing before, so it shows the obvious differences. Top row of image is using absorption, with transmission set to white and bottom row is using transmission only. these are stair stepped cubes with heights noted.

Image

So now my question is: what exactly is the purpose of transmission, if it doesn't [seem to] work physically? If I am making colored glass, should I ever have the transmission at any other value than full white? Is this more of a legacy setting before Octane had the medium nodes? Or am I just using it in the wrong manner so i'm not understanding the setting correctly?

this is just me trying to understand the setting better is all. thanks! 8-)

Re: Correct method for colored glass?

Posted: Wed Jul 01, 2015 7:38 pm
by oguzbir
Perfect test.
I guess you're right bout the fact that the medium feature is added later.

Think of a colored stained glass, a thin glass.
I am thinking of on top of my head. I would go for using color @ transmission . With out messing around with absorption, and setting its scale. I have a solid and quickest solution.
It could be read as more of a legacy setting resulting a freedom to choose not physically correctness.

And of course there times where physically correctness =! (not equals) Better.

Re: Correct method for colored glass?

Posted: Wed Jul 01, 2015 8:06 pm
by itsallgoode9
for my work, the absorption model is much more desirable actually. that's just me though, doing realistic liquor/beer bottles for a job. I kinda started thinking about this issue recently on a whiskey bottle project where i noticed the whiskey in the bottle neck was the exact same color as the whiskey in the body of the bottle (neck should be lighter color/more yellow), but haven't really had time to fully test things out till now.

thinking about it more, i suppose i would need to turn the transmission down if I needed a dark glass, like a wine bottle or something, maybe. I'll have to play with this test more later but based on this result it seems absorption only isn't making the glass darker, just more saturated.

Re: Correct method for colored glass?

Posted: Fri Jul 03, 2015 2:18 pm
by bepeg4d
hi,
the correct way for colored glass and liquid is to use the absorption. The absorption is related with the volume and can accept only gray, gaussian or RGB color. The transmission is not related with the volume but can accept also image texture input, so is used in different materials than liquid or glass. the specular material is the most powerful node that we have, you can create from diffuse to glossy going through SSS, emitters and transparent materials or a combination of them :D
ciao beppe

Re: Correct method for colored glass?

Posted: Sun Jul 05, 2015 8:57 pm
by itsallgoode9
Thanks for the explanation! It might be good to make the difference more clear in the manual. I've been using Octane for about a year now and didn't realize the difference between the two methods even after reading the manual. I had always been using transmission, having to adjust color issues in post (different thicknesses of glass being the same color) not realizing that I was just using the wrong method all along! The way materials are organized (absorption being in the medium node) it seems as if it's something you'd use with SSS, not as a basic glass material. That could just be me though! 8-)