 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I have a somewhat complex scene which gives me strange parsing
times on different computers :
Athlon 1333 Mhz - Win 2000 - 512 Mb
POV 3.5 : 23 seconds
POV 3.6 : 44 seconds
Intel P4 2.8 Mhz - Win XP - 512 Mb
POV 3.5 : 18 seconds
POV 3.6 : 55 seconds
The scene contains CSG, many macros, several big meshes, a big
image_map, #reads a big file. It does NOT contain photons,
isosurfaces, HF...
Rendering time seems normal (though I didn't compare accurately).
Does anyone noticed similar behaviour ?
Fabien.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
in news:402a932d$1@news.povray.org Fabien Mosen wrote:
> Does anyone noticed similar behaviour ?
>
Yes, I noticed using my mesh generation include files. The use loops and
functions a lot.
Ingo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
news:402a932d$1@news.povray.org...
> I have a somewhat complex scene which gives me strange parsing
> times on different computers :
I just observed a similar behaviour - worse in fact. With a scene containing
essentially big meshes (mesh2):
POV-Ray 3.5 - parsing 65 s
POV-Ray 3.6b1 - parsing 149 s
With one mesh (#include only, about 400000 tokens): 19 seconds in 3.6b1 vs 8
seconds with 3.5.
Intel P4 1.7Gz, Windows XP HE SP1, 1GB RAM.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
news:402a932d$1@news.povray.org...
> Does anyone noticed similar behaviour ?
In addition to meshes, a simple loop causes it too.
#declare i=0;
#while (i<100000)
sphere{0,1 translate x*i*2 pigment{rgb x}}
#declare i=i+1;
#end
Takes 19s with 3.6b1, 8s with 3.5
Intel P4 1.7Gz, Windows XP HE SP1, 1GB RAM.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
An early pre-beta version of 3.5 had a similar problem and it was caused
by it using C++ streams to read the input. When they were changed to
C streams the speeds went back to the 3.1 speeds.
I wonder if the reason is the same now...
--
plane{-x+y,-1pigment{bozo color_map{[0rgb x][1rgb x+y]}turbulence 1}}
sphere{0,2pigment{rgbt 1}interior{media{emission 1density{spherical
density_map{[0rgb 0][.5rgb<1,.5>][1rgb 1]}turbulence.9}}}scale
<1,1,3>hollow}text{ttf"timrom""Warp".1,0translate<-1,-.1,2>}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
> [...]
>
> #declare i=0;
> #while (i<100000)
> sphere{0,1 translate x*i*2 pigment{rgb x}}
> #declare i=i+1;
> #end
>
> Takes 19s with 3.6b1, 8s with 3.5
> Intel P4 1.7Gz, Windows XP HE SP1, 1GB RAM.
I did a quick test here on Linux and it parses in ~31s on 3.5 and 21s on
3.6 here, that difference is at least partly due to different
optimizations - i assume the same for your 3.5/3.6 difference observation.
Christoph
--
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 11 Jan. 2004 _____./\/^>_*_<^\/\.______
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <402fdefa@news.povray.org> , Warp <war### [at] tag povrayorg> wrote:
> I wonder if the reason is the same now...
No.
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
news:r03### [at] tritonimagicode...
> I did a quick test here on Linux and it parses in ~31s on 3.5 and 21s on
> 3.6 here, that difference is at least partly due to different
> optimizations - i assume the same for your 3.5/3.6 difference observation.
Perhaps someone on a non-Windows platform should test the mesh parsing too.
If you haven't one I can see if I can put one for download.
Actually I'm back to 3.1 as far as parsing time is concerned :D
I should add that this seems to affect only the parsing time, not the render
time. I noticed that it's a different compiler too.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
>
> Perhaps someone on a non-Windows platform should test the mesh parsing too.
Mesh parsing is about the same here as the loop, i only tested a very
small mesh that only took about 5 seconds so it is not very accurate but
3.6 is slightly faster than 3.5.
> Actually I'm back to 3.1 as far as parsing time is concerned :D
3.1 is indeed much faster for the loop here as well (the mesh was a
mesh2 so i could not test it), but the changes from 3.1 to 3.5 were
fairly large so this is quite understandable (more features -> slower
parsing).
> I should add that this seems to affect only the parsing time, not the render
> time. I noticed that it's a different compiler too.
You could try Megapov 0.7 and Megapov 1.0 in comparison. You should
also notice that with parsing times of only a few seconds you should
always run the test several times because speed of file access will
strongly vary.
Christoph
--
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 11 Jan. 2004 _____./\/^>_*_<^\/\.______
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Program ended abnormally on 15/02/2004 15:44, Due to a catastrophic Gilles Tran
error:
> news:402a932d$1@news.povray.org...
>
>
>>Does anyone noticed similar behaviour ?
>
>
> In addition to meshes, a simple loop causes it too.
>
> #declare i=0;
> #while (i<100000)
> sphere{0,1 translate x*i*2 pigment{rgb x}}
> #declare i=i+1;
> #end
>
> Takes 19s with 3.6b1, 8s with 3.5
> Intel P4 1.7Gz, Windows XP HE SP1, 1GB RAM.
>
> G.
>
I figured that on a slower machine, the difference would be more noticeable,
Version: Parse times:
-------- ------------
POV 3.1: 19s.
MegaPOV 0.7: 20s.
POV 3.5 (ICL): 17s.
MegaPOV 1.0 (MSVC): 20s.
POV 3.6 (beta1): 28s.
I noticed that 3.6 updates the status line while parsing this scene while
previous versions only had a "parsing..." message, maybe the added calculations
and time spent updating the status line have something to do with it.
--
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/* flabreque */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/* @ */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/* videotron.ca */}camera{orthographic location<6,1.25,-6>look_at a }
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
