PDA

View Full Version : A little mystery with glass material


nycL45
11-05-2005, 04:02 PM
While doing some lighting experiments, I came across a mysterious "wall" that bisected the box room and is linked to the glasswall material being used. Delete the plane & glass and it goes away. I'm not a pro with the finer settings and can't see anything in the glass settings that is causing this. Can anyone else see the cause?

Edit: "Old Glass" and colored glass behave similarly.

vid2k2
11-05-2005, 11:42 PM
Hi Leonard, I'm not seeing the wall going away ???

Edit: I moved the glass and added some thickness
for the render. Here's what I got. Email me if you
have further concerns.

nycL45
11-06-2005, 12:07 AM
Hi David, A slight change. With the glass turned on, I have removed the back wall and inserted a cone just beyond the back opening. See the attached. It's Escher like. The glass takes the cone and box end and pulls them forward in the opening while the bottom back edge does not appear to align visually with the lower left rear exterior edge of the box.

You can do this with the file I uploaded.

So it's not a wall but distortion. Does that make sense?

vid2k2
11-06-2005, 12:19 AM
Yes, I see. I covered this anomoly in one of my past posts
dealing with curved glass.
Same solution applies here: add thickness to the glass and
the anomoly disappears.Thickness is needed for the refraction
to work.

Edit: you can also take your refraction back to "1" if you
don't want to add thickness to the glass.

HTH

nycL45
11-06-2005, 12:35 AM
You were spot on, David. Yikes, I should have known it was the plane and not the glass after the same thing happened with six different glass materials. Well, it has been busy and it must be fatique getting to my normally agile 200 teraflop brain.

Are you the only one to encounter this? Has this been covered in the manual or anywhere else?

Thanks for the help.

Per-Anders
11-06-2005, 12:49 AM
This is core to non solid volume raytracers, what you are seeing is refraction, one of the fundamental properties that makes glass look like... well... glass. It's more a thing of knowing how raytracers work and a little physics 101.

Raytracing works by hitting surfaces and basically shading that surface, a refractive surface works by casting another ray onwards at a refracted angle, if it then intersects another surface with the same ior facing the other way it'll bend the ray back again, just as if the ray has gone through a solid volume. if you don't have another surface then it's going to treat the area like a volume of glass (you can use this to your advantage in for instance swimming pool type situations where you want a volume of water, you dont' have to make a real solid volume, just make a flat plane on top). If it didn't work this way then it would have to calculate if it was inside of a volume (which would make glass and refractive textures much much slower and stop them from working in many situations).

nycL45
11-06-2005, 02:12 AM
No doubt I'm rusty with physics 10x, but raytracing is something I have known in name only. Thanks for the Raytracing Intro 101. So, the effect I had was essentially looking through "a volume of glass" that extends from the front plane on to some point in the distance? If true, that would explain the foreshortening.

This is good start and info for the files. Thanks.

BTW, I went looking for a definition of "ior" (index of refraction) and found these tutorials: The World of Reflections Tutorials by Philipp Zaufel at 3DCreative Magazine
http://www.the123d.com/tutorial/general3/reflections01.shtml
http://www.the123d.com/tutorial/general4/reflections2-1.shtml
They appear to take light behavior in 3D cg into some depth.

Primitiv
11-06-2005, 04:15 AM
I don't understand. Why is the glass shader applied to the plane and not the cube?

Primitiv
11-06-2005, 04:22 AM
Your scene makes little sense to me. But it is obvious that the refraction is bigger than the cube can take. put the refraction down. I tried 1.1 for example and if the refractive effect is not apparent, the wall effect also disappeared. So you need to set it right for the effect you want to make.

nycL45
11-06-2005, 01:11 PM
Your scene makes little sense to me. But it is obvious that the refraction is bigger than the cube can take. put the refraction down. I tried 1.1 for example and if the refractive effect is not apparent, the wall effect also disappeared. So you need to set it right for the effect you want to make.

Thanks Primitiv.

It's not you. I was playing with exterior lighting for a rig approach using interior and different exterior light sources and studying reflections in the glass. This was the first time I had tried a Plane to simulate glass. In the past, I built the 3D glass in CAD and didn't have this issue.

Re your earlier ques.: the cube served as an interior space with an omni and the plane, the exterior glass wall.

CGTalk Moderation
11-06-2005, 01:11 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.