Another Object Issue

Started by j meyer, April 10, 2008, 04:40:01 pm

Previous topic - Next topic

j meyer

Hi there,
found another problem with objects and displacement and,most
likely,light.Please see attachments.
The object has normals this time,i've looked into the file.And i tried
a higher subdivision level as well.Object was made with wings3d.
I have no idea,do you?

Harvey Birdman

Is it the same interpolation issue at the terminator described here?

j meyer

Well,i don't think so,because it's still there with a higher subdivision
while the other one gets better when you use a higher polycount.

PG

Are soft shadows enabled? if so what's the sample count?
Figured out how to do clicky signatures


j meyer

Any official explanation/statement?

Matt

April 22, 2008, 07:23:08 pm #6 Last Edit: April 22, 2008, 07:24:56 pm by Matt
I don't know for sure, but it looks like the interpolated surface normals (smooth normals) are to blame. If so, I don't really have any ideas how this could be fixed. But I would expect the higher subdivision levels to reduce the problem to some extent.

The reason why interpolated surface normals can cause this problem is that the normals have to be interpolated so that the facets aren't visible, but this causes the normals to be inconsistent with the surface that is geometrically visible to the renderer. It's actually the same as the terminator problem in the thread that Harvey linked to, but with a camera instead of a shadow-casting light source.

Perhaps it's something else, but the above seems like the most likely explanation. I don't really have a solution unfortunately.

Matt
Just because milk is white doesn't mean that clouds are made of milk.

Matt

OK, thinking about this some more, the first renders show the problem a lot worse than I'd expect with the light coming from the side. Maybe there's something else going on. There are some weird things I have to do at glancing angles to make surfaces render correctly, and maybe it's doing the wrong thing on some objects.

Matt
Just because milk is white doesn't mean that clouds are made of milk.

j meyer

Thanks for taking the time to reply Matt,i do appreciate.

Cyber-Angel

I think the problem could be solved as I understand it by talking to some one who knows how to solve the problem of reversed normals which form my vary basic understanding of the subject is a J Dimensional problem within the field of Euclidean Geometry.

It could be that at render time (Just my theory) is that edge normals (in some cases) are been reversed causing the problem been shown in the provided images when those normals are handed off to the drawing engine for rendering resulting in the black edge: how to solve it I  am not qualified to offer a guess let alone a solution since I am not a programmer, this could be completely wrong if it so, my bad.

Regards to you.

Cyber-Angel  8)   

JimB

I agree with C-A. Looks like inverted normals/polys on the far side. Did you merge/combine the object from two others at any point? A mirrored object?
Some bits and bobs
The Galileo Fallacy, 'Argumentum ad Galileus':
"They laughed at Galileo. They're laughing at me. Therefore I am the next Galileo."

Nope. Galileo was right for the simpler reason that he was right.

j meyer

No i didn't merge/combine anything in this case and it's also not a
mirrored object.But i think that doesn't matter anyway,because the
effect appears even when the object is rotated.