second relfection bug?

Post your tutorials and help questions in this forum
Locked
NiGMa
Posts: 129
Joined: Thu Sep 14, 2006 2:38 am
Location: Australia

second relfection bug?

Post by NiGMa »

Hi, is this a bug? I can't seem to get rid of it. I have checked and there are no overlapping polys.

Am i correct in saying that this is rendered wrong?
Attachments
secondrefl.png
secondrefl.png (237.99 KiB) Viewed 4703 times
Captain Obvious
Posts: 165
Joined: Mon Oct 10, 2005 6:15 pm
Location: London, UK

Post by Captain Obvious »

Looks reasonable to me. Glass usually produces two reflections.
jure
Posts: 2142
Joined: Thu Jun 02, 2005 6:53 pm

Post by jure »

You're probably using double sided surface flag. This would cause reflection ray see the polygon even if it's facing away.
- Jure
NiGMa
Posts: 129
Joined: Thu Sep 14, 2006 2:38 am
Location: Australia

Post by NiGMa »

nope, im only using "recurse=10", it looks worse on glass bals and glass sliding doors, its definiately not physically accurate.
User avatar
acidarrow
Posts: 140
Joined: Wed Jan 31, 2007 7:42 pm
Contact:

Post by acidarrow »

Are you using air polys? Since kray doesn't need air polys having them could have weird results I guess.

Can you make a texture-less scene that replicates this behavior and post it here?
NiGMa
Posts: 129
Joined: Thu Sep 14, 2006 2:38 am
Location: Australia

Post by NiGMa »

no air polys, I will make a scene. gimme say... 360secs. hehe
NiGMa
Posts: 129
Joined: Thu Sep 14, 2006 2:38 am
Location: Australia

Post by NiGMa »

Wow, look at these. without the "lwo2airpoly" switch turned on, it renders wrong. I only have outer polys no inner, so the top piece of glass is only 29polys total, with beveled edges.

Am I understanding kray incorrectly or is this a bug, I thought the lwo2airpolys was only if you have outer polys plus air polys?

hmm.. what do you guys think? also look at the render times! no wonder my renders have been so slow lately!
Attachments
tailer commands=recurse 10
tailer commands=recurse 10
Room01_00001.png (138.07 KiB) Viewed 4685 times
tailer commands=recurse 10;lwo2airpolys 1
tailer commands=recurse 10;lwo2airpolys 1
Room02_00001.png (128.51 KiB) Viewed 4694 times
User avatar
acidarrow
Posts: 140
Joined: Wed Jan 31, 2007 7:42 pm
Contact:

Post by acidarrow »

(moved my answer from the other thread here)

I got the scene from the other thread and played around with it.

I confirm the double reflection. I made the scene even more minimal and it still happens. So, it seems that the air polys that (I am assuming) kray adds have some reflective properties, maybe to simulate/fake the internal reflection happening, I am not sure.

The thing is, if the refraction index is the same for both materials (the glass in scene has nd 1 same as air) there is no internal reflection happening, none at all. (go to http://www.phy.ntnu.edu.tw/oldjava/optics/prism_e.html and put nd 1 to both materials).

So the only way to not have the second reflection happen is to use lwo2airpolys which I am assuming switches to the lightwave model which needs air polys, so kray doesn't add its own.

I believe this is a bug. If it is not, some explanation on how kray handles glass is needed.
NiGMa
Posts: 129
Joined: Thu Sep 14, 2006 2:38 am
Location: Australia

Post by NiGMa »

Captain Obvious wrote:Looks reasonable to me. Glass usually produces two reflections.
YOU ARE RIGHT!

I was on the toilet yesterday, and was looking at a reflection in the shower glass and it there were two! I was surprised!

So, this is not a bug. :D
Locked