Glass problem?

[RevitGary]

In the image there is a low glass rail in front of glass sliders.

 

Why is the glass railing showing up soo dark?

 

It is thin wall Arch / design Glass.

[Crazy Homeless Guy]

Can you upload part of the file? That way we can have a direct look?

[RevitGary]

I think I am finding a solution, stilll working on it. It seems to be something with back face culling.

 

If I check back face culling things look much better, including the slider glass.

[Peter M. Gruhn]

I've duplicated your problem with a simple test scene.

 

Box with matte plastic on it.

A single poly in front of it (not a box, not double walled) with thin glass.

A ground plane.

mr

mr daylight system

draft fg

1:4

 

 

The dark area remains if the ground is deleted.

The dark area is not at horizon height.

The height of the dark areas is dependant on the distance to the box behind. The closer to the box, the higher the dark.

Moving the camera doesn't change it.

There is an area near an edge where it does not appear behind the glass.

It goes away if I turn the glass's "cast shadows" off.

 

Lower IOR decreases it.

 

Therefore... yup, making the sun shine more perpedicularly on it will also decrease it. And move it higher. And change all the light in your scene ;-).

[RevitGary]

I tried cast shadows off. That didnt help much.

Clicking back face culling did the trick.

 

My windows are boxes

Edited February 6, 2010 by RevitGary

[Crazy Homeless Guy]

Lower IOR decreases it.

 

Maybe I am not following, but isn't the point of thin walled glass to eliminate the need for IOR? ...I mean. you can use a single sided face, and thin walled allows it to act as though their is no refraction. Or at least that is what I assumed it was doing.

[Peter M. Gruhn]

Hmm...

 

Backface culling didn't do much for me. You had something shinier behind yours though and it was looking at the backside of the glass... so...

 

If there is no IOR then you've just got a boring material with transparency. The point behind glass is the reflection and refraction. The point behind thin glass is to get something that looks like glass but w/o the expense of glass. So, and I'm guessing here, thin glass would say not "this is the IOR that you have to painstakingly calculate as you split off all those expensive rays" but rather "this is the IOR you should jerry rig hack".

 

Maybe.

 

It interesting that the situation I had didn't solve your problem. It looked so close and happened so easily. One thing, yours did look darker than mine, a fact that I put down to materials and setting.

[mattclinch]

is it not a case of simply upping your max trace depths to allow enough light and refraction bounces to pass through the glass?

[Peter M. Gruhn]

That crossed my mind Matt and I didn't test it as in my mind the default values are high enough for some fancy bouncing and I wasn't asking all that much of it at all. Ah.. just thought - if the bouncing fails then we won't see "what you're looking at but dark" we'll see "fail color". Full on black or environment or whatever.

[Justin Hunt]

Hove you got AO on the Walls? If so, how far away is the balustrade glass from the wall?

 

jhv

[Jennifer]

Backface culling just eliminates the polygons facing the other direction, effectively making the boxes appear as single poly planes. It sounds like a trace depth issue, since this helps. You need to make sure that both the material and mental ray are set to enough levels; the clear box window is two traces for each ray w/o backface culling.

 

The A&D material's Glass (Thin Geometry) template has 8 for the Max Trace Depth (in Advanced Settings), however mr may be only set to 6 by default. Make sure the material is at least as much as your mr settings, and vice versa.

 

The glass template is set to "By IOR" for BRDF; changing the IOR just changes the reflectivity based on angle, and the Thin Walled option prevents actual refraction. Maximum reflectivity is set with the Reflectivity setting, 1.0.

 

Jenni