POV-Ray : Newsgroups : povray.binaries.images : mesh2 coloring problems Server Time
5 Nov 2024 22:22:03 EST (-0500)
  mesh2 coloring problems (Message 1 to 3 of 3)  
From: Tom Austin
Subject: mesh2 coloring problems
Date: 8 May 2006 10:09:23
Message: <445f5113@news.povray.org>
Maybe someone has some info that might help me out.  The only thing I 
can think this could be is the scaling problem that comes up when 
someone tries to render our solar system and a pencil on earth in real 
scale in relation to each other - something to do with working with 
really large and really small numbers at the same time.

I'm trying to render a mesh2 and am running into some problems with the 
pigment on individual triangles.

The following scene is a single mesh2 face that was painstakingly 
isolated so that I could better understand the problem.
The first image is the render of the scene - a white triangle, not the 
expected results.

But if I modify one of the vectors I get something similar to image #2.

image 2 - modify vector 1
<57312.569,784.27165,53547.105>,  original
<57312.568,784.27165,53547.105>,  modified
          ^



But results vary with what I modify - image 3

image 3 - modify vector 2
<57310.236,784.35367,53548.234>,  original
<57310.237,784.35367,53548.234>,  modified
          ^


Image 4 is what the big mesh looks like with the little white faces 
sticking out like a sore thumb.  The white face on the far right is my 
example face that I was able to isolate.


This is not a mesh that I want to modify by hand.  It is about 400,000 
faces with about 200,000 vertexes.  The final scene will likely have 
many more faces and vertexes.


Does anyone know what might be going on, and/or know what can be done to 
fix it?  It would be really nice to get rid of the white as it detracts 
from the presentation that we are trying to do.



***********************************************************
***********************************************************
***********************************************************

light_source {<57312,1784,53547> rgb 1}
camera {location <57312,804,53547> look_at <57312,784,53547>}

mesh2 {
vertex_vectors {
3,
<57312.569,784.27165,53547.105>,
<57310.236,784.35367,53548.234>,
<57312.756,784.26509,53547.605>,
}
texture_list {
3,
texture{pigment{rgb<0.1,0,0>}}
texture{pigment{rgb<0,0,1>}}
texture{pigment{rgb<0.3,0.3,0.3>}}
}
face_indices {
1,
<1,0,2>,1,0,2,
}
}

***********************************************************
***********************************************************
***********************************************************


Post a reply to this message


Attachments:
Download 'white_triangle_isolated_1.jpg' (3 KB) Download 'white_triangle_isolated_2.jpg' (3 KB) Download 'white_triangle_isolated_3.jpg' (3 KB) Download 'white_triangle_many_4.jpg' (88 KB)

Preview of image 'white_triangle_isolated_1.jpg'
white_triangle_isolated_1.jpg

Preview of image 'white_triangle_isolated_2.jpg'
white_triangle_isolated_2.jpg

Preview of image 'white_triangle_isolated_3.jpg'
white_triangle_isolated_3.jpg

Preview of image 'white_triangle_many_4.jpg'
white_triangle_many_4.jpg


 

From: Shay
Subject: Re: mesh2 coloring problems
Date: 10 May 2006 12:45:05
Message: <44621891$1@news.povray.org>
Tom Austin wrote:
> Maybe someone has some info that might help me out.  The only thing
> I can think this could be is the scaling problem that comes up when 
> someone tries to render our solar system and a pencil on earth in
> real scale in relation to each other - something to do with working
> with really large and really small numbers at the same time.


Have you tried search/replace the camera and vertex_vectors sections 
'>/10' for '>'? Fixes your examples.

  -Shay


Post a reply to this message

From: Tom Austin
Subject: Re: mesh2 coloring problems
Date: 11 May 2006 08:40:51
Message: <446330d3$1@news.povray.org>
Shay wrote:
> 
> Have you tried search/replace the camera and vertex_vectors sections 
> '>/10' for '>'? Fixes your examples.
> 

You are saying scale down the scene by a factor of 10- right...

I don't believe that we've tried a pure scale like that on the whole 
data set.  I'll have to try it to see if it kills the problem.

For now we translate the whole thing closer to the origin.  It's worked 
for now, but if we have to do more of this work, it might not.

I am curious as to what causes this behavior in POV-Ray. When one 
understands the cause, then one has the ability to avoid it.

Thanks for your help



Tom


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.