POV-Ray : Newsgroups : povray.advanced-users : clipped_by, complex unions and bad csg results Server Time
24 Nov 2024 10:04:10 EST (-0500)
  clipped_by, complex unions and bad csg results (Message 1 to 6 of 6)  
From: Samuel Benge
Subject: clipped_by, complex unions and bad csg results
Date: 11 Oct 2007 18:04:26
Message: <470e9dea$1@news.povray.org>
Greetings,

I've been making an object library which allows me to make beveled, 
filleted and chamfered objects, to be used for stone pillars and the 
like. Not wanting to deal with slow csg operations, I opted to make 
everything from basic shapes with the clipped_by modifier applied.

It all works very well, since the (hollow) surfaces meet up with each 
other if used properly. To see how well, I tested the closed objects 
with interior_texture, transparency, media, refraction, and interior 
attenuation. All these features work as they should, and the render time 
is relatively fast, as expected.

The problems come when I try to apply traditional csg techniques to my 
objects (difference, merge, intersection). Strange artifacts appear, 
such as transparent planes, partial csg results, and other strange things.

I'd rather not post my code here, as the definition for each object is 
rather large.

Has anyone else encountered strange results when trying to difference, 
intersect or merge unions of clipped_by objects? Was there a workaround? 
I'm beginning to think there isn't a solution....

Sam


Post a reply to this message

From: Bruno Cabasson
Subject: Re: clipped_by, complex unions and bad csg results
Date: 11 Oct 2007 18:15:01
Message: <web.470e9fc35c01085976e65db0@news.povray.org>
Samuel Benge <stb### [at] THIShotmailcom> wrote:
> Greetings,
>
> I've been making an object library which allows me to make beveled,
> filleted and chamfered objects, to be used for stone pillars and the
> like. Not wanting to deal with slow csg operations, I opted to make
> everything from basic shapes with the clipped_by modifier applied.
>
> It all works very well, since the (hollow) surfaces meet up with each
> other if used properly. To see how well, I tested the closed objects
> with interior_texture, transparency, media, refraction, and interior
> attenuation. All these features work as they should, and the render time
> is relatively fast, as expected.
>
> The problems come when I try to apply traditional csg techniques to my
> objects (difference, merge, intersection). Strange artifacts appear,
> such as transparent planes, partial csg results, and other strange things.
>
> I'd rather not post my code here, as the definition for each object is
> rather large.
>
> Has anyone else encountered strange results when trying to difference,
> intersect or merge unions of clipped_by objects? Was there a workaround?
> I'm beginning to think there isn't a solution....
>
> Sam

Good initiative!

Unfortunately, I nearly never happen to use objects with clipped_by, and I
can't be of much help for your problem. Hope someone else can.

Bruno.


Post a reply to this message

From: Alain
Subject: Re: clipped_by, complex unions and bad csg results
Date: 11 Oct 2007 23:03:58
Message: <470ee41e@news.povray.org>
Samuel Benge nous apporta ses lumieres en ce 2007/10/11 18:05:
> Greetings,
> 
> I've been making an object library which allows me to make beveled, 
> filleted and chamfered objects, to be used for stone pillars and the 
> like. Not wanting to deal with slow csg operations, I opted to make 
> everything from basic shapes with the clipped_by modifier applied.
> 
> It all works very well, since the (hollow) surfaces meet up with each 
> other if used properly. To see how well, I tested the closed objects 
> with interior_texture, transparency, media, refraction, and interior 
> attenuation. All these features work as they should, and the render time 
> is relatively fast, as expected.
> 
> The problems come when I try to apply traditional csg techniques to my 
> objects (difference, merge, intersection). Strange artifacts appear, 
> such as transparent planes, partial csg results, and other strange things.
> 
> I'd rather not post my code here, as the definition for each object is 
> rather large.
> 
> Has anyone else encountered strange results when trying to difference, 
> intersect or merge unions of clipped_by objects? Was there a workaround? 
> I'm beginning to think there isn't a solution....
> 
> Sam
It could be a bounding problem.
Have you tried adding: bounded_by { clipped_by }? This will add a manual 
bounding conforming to the clipping object.

-- 
Alain
-------------------------------------------------
Unitarianism: Come let us reason together about this shit.


Post a reply to this message

From: Samuel Benge
Subject: Re: clipped_by, complex unions and bad csg results
Date: 12 Oct 2007 02:02:53
Message: <470f0e0d$1@news.povray.org>
Alain wrote:
> It could be a bounding problem.
> Have you tried adding: bounded_by { clipped_by }? This will add a manual 
> bounding conforming to the clipping object.

I tried several variations of your suggestion, and either nothing 
worked, or it made things worse. I will have to try a simpler test to 
see if I can find success.

Sam


Post a reply to this message

From: Warp
Subject: Re: clipped_by, complex unions and bad csg results
Date: 12 Oct 2007 07:09:24
Message: <470f55e4@news.povray.org>
Samuel Benge <stb### [at] thishotmailcom> wrote:
> The problems come when I try to apply traditional csg techniques to my 
> objects (difference, merge, intersection). Strange artifacts appear, 
> such as transparent planes, partial csg results, and other strange things.

  A short example scene demonstrating the problem would be helpful.

  My guess is that it's the infamous coincident surfaces problem.
(I actually don't know if this problem happens if two *clipping* surfaces
coincide, but it could well be.)

-- 
                                                          - Warp


Post a reply to this message

From: Kenneth
Subject: Re: clipped_by, complex unions and bad csg results
Date: 17 Dec 2007 00:10:01
Message: <web.4766047b5c01085978dcad930@news.povray.org>
Alain <ele### [at] netscapenet> wrote:
> Samuel Benge nous apporta ses lumieres en ce 2007/10/11 18:05:

> > Has anyone else encountered strange results when trying to difference,
> > intersect or merge unions of clipped_by objects? Was there a workaround?
> > I'm beginning to think there isn't a solution....
> >
> > Sam
>
> It could be a bounding problem.
> Have you tried adding: bounded_by { clipped_by }? This will add a manual
> bounding conforming to the clipping object.
>

I have a suggestion--which may not have any relevance--but you might try
changing Remove_Bounds=on/off and/or Bounding_Threshold=*some integer* in your
..INI file. That has helped me in some *odd* situations. Sorry I can't be more
specific as to *why* this trick (sometimes) works; I don't quite understand it
myself. The default Bounding_Threshold in POV-Ray is 3--or so it seems--not 25
as mentioned in the docs. Try alternately changing it to 1, or to 50 or more
(i.e., two extremes.)

I'd be curious to know if this helps.

Ken


Post a reply to this message

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