 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I'm using POVRay 3.6 at WinXP Pro, and found that the following
scene reports 4 *infinite* objects:
sphere {0, 1358}
sphere {5000, 1358}
box {-1358, 1358}
box {0, 2*1358}
But the following scene reports (as expected) 4 finite objects:
sphere {0, 1357}
sphere {5000, 1357}
box {-1357, 1357}
box {0, 2*1357}
It seems that the condition for an object to be infinite is related to its
bounding box, and that there is some constant (around 2*1357) which acts as
a threshold.
Would this be a bug?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <web.40d9ad47b0d1a68d27bbadf50@news.povray.org> , "Leonardo"
<lmf### [at] mega istutlpt> wrote:
> Would this be a bug?
No, expected behavior that I think got actually documented somewhere in the
3.6 manual, but I don't recall where because this question came up before in
these groups. In short, what POV-Ray reports has nothing to do with the
actual object, it reports internal data (which is useful because it tells
you how POV-Ray handles an object) and it simply cannot report what you
think would be correct because that information cannot be determined for all
objects anyway.
Thorsten
____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Leonardo" <lmf### [at] megaistutlpt> wrote in message
news:web.40d9ad47b0d1a68d27bbadf50@news.povray.org...
> I'm using POVRay 3.6 at WinXP Pro, and found that the following
> scene reports 4 *infinite* objects:
>
> sphere {0, 1358}
> sphere {5000, 1358}
> box {-1358, 1358}
> box {0, 2*1358}
>
> But the following scene reports (as expected) 4 finite objects:
>
> sphere {0, 1357}
> sphere {5000, 1357}
> box {-1357, 1357}
> box {0, 2*1357}
>
> It seems that the condition for an object to be infinite is related to its
> bounding box, and that there is some constant (around 2*1357) which acts
as
> a threshold.
> Would this be a bug?
>
>
you still get the infinte object count of 4 for several scale factors such
as:
sphere {0, 1358 scale 1.2}
sphere {5000, 1358 scale 1.2}
box {-1358, 1358 scale 1.2}
box {0, 2*1358 scale 1.2}
you get a count of 4 finite objects if you have:
sphere {0, 1358 scale 0.1}
sphere {5000, 1358 scale 0.1}
box {-1358, 1358 scale 0.1}
box {0, 2*1358 scale 0.1}
however, if you scale by 0.2, you get 4 infinites:
sphere {0, 1358 scale 0.2}
sphere {5000, 1358 scale 0.2}
box {-1358, 1358 scale 0.2}
box {0, 2*1358 scale 0.2}
i didn't find a number greater than 1 that resulted in finite object counts.
these were just my quick observations, for whatever they're worth.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Thorsten Froehlich" <tho### [at] trfde> wrote:
> No, expected behavior that I think got actually documented somewhere in the
> 3.6 manual, but I don't recall where because this question came up before in
> these groups.
If it's documented, then certainly it's not a bug :P
Haven't readed the 3.6 docs yet
> and it simply cannot report what you
> think would be correct because that information cannot be determined for all
> objects anyway.
Why not? The primitive types are well classified into finite/infinite
The finiteness on the mathematical sense is not computable. I wouldn't be
surprised if I rendered a sphere or an ellipsoid with the Pov primitive
poly and it reported 1 infinite object. POV doesn't know it's a sphere - it
only knows it's a polynomial object, and those objects are handled like
infinite objects.
But if I rendered a sphere with the sphere primitive, why POV thinks it's an
infinite object??? Spheres can't be infinite... neither do boxes...
Would POV-Ray internally convert a sphere primitive into a poly primitive? I
don't think so...
I (still) can't understand it
When the source code comes available maybe I'll find out :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 23 Jun 2004 13:48:02 EDT, "Leonardo" <lmf### [at] megaistutlpt> wrote:
> Spheres can't be infinite...
Well....
sphere{ 0 1 inverse }
:-)
ABX
Post a reply to this message
|
|
| |
| |
|
|
From: Thorsten Froehlich
Subject: Re: Bug on infinite objects counting?
Date: 24 Jun 2004 03:22:56
Message: <40da8150@news.povray.org>
|
|
|
| |
| |
|
|
In article <6fhjd0henv315g5rlg4a0h1e4bg4v81qps@4ax.com> , ABX
<abx### [at] abxartpl> wrote:
> On Wed, 23 Jun 2004 13:48:02 EDT, "Leonardo" <lmf### [at] megaistutlpt> wrote:
>> Spheres can't be infinite...
>
> Well....
>
> sphere{ 0 1 inverse }
Exactly!
Thorsten
____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <web.40d9c252970f5f5527bbadf50@news.povray.org> , "Leonardo"
<lmf### [at] megaistutlpt> wrote:
> But if I rendered a sphere with the sphere primitive, why POV thinks it's an
> infinite object???
You do not usually bound spheres with bounding boxes - you usually do not
bound spheres at all. As I explicitly said, the information POV-Ray shows
is an internal statistic just like the intersection stats, the function VM
call stats, photon stats, and so on. It is not an analysis of the scene
contents in a mathematical sense at all.
Thorsten
____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
