Frosted glass

[re:FORM]

I had a look at previous threads on frosted glass, but they all seem out of date since the new version of vray.

What is the best approach to frosted glass:
Refraction glossiness or translucency? The later seems faster but less accurate(with the same no, of samples) than the former.

I guess it depends on the way the glass has been frosted in real life, ie transfer or sand-blasted(yes I am that pedantic).

I'm interested in your suggestions for both accurate and fast approaches to this shader.

[jow]

I when down the Refract glossy road on a recent project, about .85 (subs about 15-20) i think. And kept the reflect glossy at 1. Times where good but it was only a small scene.

[re:FORM]

An unscientific test: Transparencies, and glossy refractions. Each vertical strip of glass has different parameters.

The scene is lit with vray sun/sky. Vray lights at the first floor slab edge, and in the spandrel box at the top. Interesting to compare the dispertion of light with the different materials.

[jow]

Interesting results. What where the settings you used for the translucent mat?

[andrewjohn81]

I'm also curious about that. We have a need right now for some translucent materials similar to frosted glass. People have also been asking for a milky white, similar to a plain white corian.
The problem is they want the materials with a light behind them. Of course, the light then needs to transmit through the object pretty well.

I can get the ojbect looking good, but it has a protrusion comming out of it that should also be lit. What I did for tests was just make a plan with thickness, out of a cube, and extruded a tube out of it after making a hole. That's basically the shape of the final object. I can get the majority of the plane to look nice, and even a bit of the tube, but for the life of me I can't get light to get transmitted throughout.

Is there a way to get light transmitted a bit more evenly? I've got it looking better, but it's still not great.

Right now I'm using Refract 100% transparent with IOR 2.2
Glissiness is a low 0.1 subdivs 20 for now use interpolation checked ON
Fog color at 123, multiplier at .02
had wax model, scatter @ .43
Fwd?bck @ 1.0
mulitplier @ 1.0
Thickness at 1000 to encompass the entire object just in case.
This is using the Blinn BRDF if that makes a difference.
Also double sided is turned off.

[jreynolds]

Just an observation...

The left image looks like it is frosted through and through.

The right image looks like the surface is frosted.


Jon

[andrewjohn81]

I'm pretty sure that's what should happen since the translucency should simulate a volumetric effect, but the refraction glossy should only be a surface glossy I think.