LuxCoreRender Forums

https://i.imgur.com/lE6Kd2p.png
Probably it is wrong

JulianoLisboa wrote: ↑Mon May 10, 2021 2:41 am Corona alpha
corona_alpha.jpg

They may have the alpha value related to the amount refraction (more refraction = more white). Can you you run the same test with a sphere and post the alpha channel ?

Dade wrote: ↑Mon May 10, 2021 8:41 am

JulianoLisboa wrote: ↑Mon May 10, 2021 2:41 am Corona alpha
corona_alpha.jpg

They may have the alpha value related to the amount refraction (more refraction = more white). Can you you run the same test with a sphere and post the alpha channel ?

Yes sure. Here it is.

JulianoLisboa wrote: ↑Mon May 10, 2021 12:34 pm Yes sure. Here it is.sphere_beauty.jpgsphere_beauty.png

What is the difference between the 2 spheres ? Is it the amount of reflection ?

It seems to set alpha=0 if the ray is transmitted or alpha 1.0 if it is reflected

Dade wrote: ↑Mon May 10, 2021 2:24 pm

JulianoLisboa wrote: ↑Mon May 10, 2021 12:34 pm Yes sure. Here it is.sphere_beauty.jpgsphere_beauty.png

What is the difference between the 2 spheres ? Is it the amount of reflection ?

It seems to set alpha=0 if the ray is transmitted or alpha 1.0 if it is reflected

The one on the left is solid and the one on the right is thick (solidify)

JulianoLisboa wrote: ↑Mon May 10, 2021 12:34 pm Yes sure. Here it is.
sphere_beauty.jpg

sphere_beauty.png

This doesn't make sense, you have a solid background in your rendering and the alpha channel of that rendering is, by definition, all white (with or without the 2 glass spheres, it doesn't matter). You are using something else, not the alpha channel. Please render a glass sphere over a plane like this (RGB and Alpha):


We can argue about the alpha channel of pixels belonging to the sphere but the alpha of plane pixels must be 1. Otherwise you are using some kind of shadow catcher or similar.

Hello Dade, so ... Yes I am using a shadow collector, but that has been talked about since the beginning. The use of this is to be the transparent glass when you have a product in the studio for example. So I believe that it is the shadow collector that has to be changed.

JulianoLisboa wrote: ↑Wed May 12, 2021 5:01 pm Yes I am using a shadow collector, but that has been talked about since the beginning.

The first MCurto's post in this thread talks only about alpha channel, they are 2 different topics.

I reworked how alpha channel is handled for purely transmitted eye paths and the result should match what was asked:


This change is available in v2.6 daily builds.

The last question is if this new behavior should be used far all kind of transmitted paths (like in the new code) or only for specular one (i.e. glass but not rough glass, glossy translucent, etc.).

Hoah!Thaks a lot, will test, and give feed back!Looks great.

I guess because I'm a bit new to the whole scene maybe I'm not calling things by their names but the ability to have transparency (alpha?) on a glass material believe its invaluable for people who are more focused on the kind of studio stills like product rendering and stuff, I work a lot with glass and liquids (photos) and I'm sure this will prove very handy to help me try to integrate 3D more often.

Anyway: thanks!

MCurto wrote: ↑Thu May 13, 2021 12:50 pm I guess because I'm a bit new to the whole scene maybe I'm not calling things by their names but the ability to have transparency (alpha?) on a glass material believe its invaluable for people who are more focused on the kind of studio stills like product rendering and stuff, I work a lot with glass and liquids (photos) and I'm sure this will prove very handy to help me try to integrate 3D more often.

The reason I was not marking the glass as transparent is because you are not seeing directly what is beyond the glass but the ray is refracted and it lands in a different part of the sky.
So, if you use this new alpha channel for some kind of background compositing, you are not really ending with an accurate result. But I guess "something" is till better than "nothing"