 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi!
I hope someone can tell me, why my picture needs maaaaany hours rendering a
wall!? The wall is created with the mur.inc macro from Steven Pigeon. You
can find it here: http://home.pacbell.net/tylereng/index.html
I use the standard textures from the example file (nothing too special I
think, some dents as I see it).
The macro creates the wall by placing brick by brick. Every brick is a
superellipsoid.
The wall takes hours, even at 512*384 pixel without anything like radiosity
or photons or focal blur on a athlon 1800+! Is it the superellipsoids which
slow the system so much down?
I guess, the time can't be reduced, but I would like to understand, WHY it
takes so long?!
Regards,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi!
I think it are the superellipsoids which slow down the process. You may try
to use rounded boxes instead, altough you need to edit the macro for that.
JWV
"frozen" <fro### [at] thefrozen no-ipcom> wrote in message
news:web.414eda6f5c5e51dd102354040@news.povray.org...
> Hi!
>
> I hope someone can tell me, why my picture needs maaaaany hours rendering
a
> wall!? The wall is created with the mur.inc macro from Steven Pigeon. You
> can find it here: http://home.pacbell.net/tylereng/index.html
>
> I use the standard textures from the example file (nothing too special I
> think, some dents as I see it).
>
> The macro creates the wall by placing brick by brick. Every brick is a
> superellipsoid.
>
> The wall takes hours, even at 512*384 pixel without anything like
radiosity
> or photons or focal blur on a athlon 1800+! Is it the superellipsoids
which
> slow the system so much down?
>
> I guess, the time can't be reduced, but I would like to understand, WHY it
> takes so long?!
>
> Regards,
>
> frozen
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>
> The wall takes hours, even at 512*384 pixel without anything like
radiosity
> or photons or focal blur on a athlon 1800+! Is it the superellipsoids
which
> slow the system so much down?
>
It's been a while since I played with that one, but I seem to remember that
the wall itself didn't take that long after I remmed out the grass and
trees. There is a lot of detail in both of them, and the more detail, the
longer the computer has to think before it can plot. The more detailed the
scene, the longer it's going to take.
Greybeard
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thanks!
Since the scene takes very long even without everything except the walls, I
should try the rounded boxes. If I ever need a wall again.. ;-) Right now
it renders and since it renders for hours now, I will not stop it! ;-)
But can anyone explain, WHY the rendering of superellipsoids takes so long?
Thanks again,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> But can anyone explain, WHY the rendering of superellipsoids takes so
long?
Well I cant help you there but... You might have a look at the macro and
try substituting a Mesh for each SE. Create one mesh and scale it etc in
the same way the SE's are scaled. Might be quicker.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Josh" <som### [at] microsoftcom> wrote:
> > But can anyone explain, WHY the rendering of superellipsoids takes so
> long?
>
> Well I cant help you there but... You might have a look at the macro and
> try substituting a Mesh for each SE. Create one mesh and scale it etc in
> the same way the SE's are scaled. Might be quicker.
Thanks! I will try this when I need a new wall...
Regards,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"frozen" <fro### [at] thefrozenno-ipcom> wrote in message
news:web.414eecf4bf1ad50e102354040@news.povray.org...
> Thanks!
>
> Since the scene takes very long even without everything except the walls,
I
> should try the rounded boxes. If I ever need a wall again.. ;-) Right now
> it renders and since it renders for hours now, I will not stop it! ;-)
>
> But can anyone explain, WHY the rendering of superellipsoids takes so
long?
>
Somethings wrong, I removed the grass and trees, 1M, 37S to render.
2.4ghz Celeron, 256M memory, 800 X 600. Time to render with trees and
grass, 1 Hr 17 M.
Greybeard
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"GreyBeard" <r.b### [at] sbcglobalnet> wrote:
> "frozen" <fro### [at] thefrozenno-ipcom> wrote in message
> news:web.414eecf4bf1ad50e102354040@news.povray.org...
> > Thanks!
> >
> > Since the scene takes very long even without everything except the walls,
> I
> > should try the rounded boxes. If I ever need a wall again.. ;-) Right now
> > it renders and since it renders for hours now, I will not stop it! ;-)
> >
> > But can anyone explain, WHY the rendering of superellipsoids takes so
> long?
> >
> Somethings wrong, I removed the grass and trees, 1M, 37S to render.
>
> 2.4ghz Celeron, 256M memory, 800 X 600. Time to render with trees and
> grass, 1 Hr 17 M.
>
> Greybeard
I suppose, you used the same example scene, which uses mur.inc? With the
same lightsource, the same sky and the changed walls?
I can't understand, how this is possible? Maybe it is because I use windows
2k at the moment?
Thank you very much,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>
> I suppose, you used the same example scene, which uses mur.inc? With the
> same lightsource, the same sky and the changed walls?
Nothing changed from original.
>
> I can't understand, how this is possible? Maybe it is because I use
windows
> 2k at the moment?
>
Haven't tried it on my 2K machine, I ran it under 98SE. TImes I got are
about the same on my XP Home edition machine, but that's a 2gig, should run
slightly slower. I think times were similar under SuSe 9.1 Linux, but
haven't played with that one for a while. (The Win2K machine is dual boot,
20 gig allotted to Win, 100 gig given to Linux. 2.6ghz Pentium 4c with 1gig
memory. Yeah, it's fast.)
Greybeard
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hey Greybeard,
thank you very much for your efforts on testing it! I don't have any idea,
what might cause these very slow renderings. I guess, I will have to do
some more tests, when the picture it's rendering now is done.. This will
still take a while.. 1d2h working, 43% reached... :-(
Regards,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"frozen" <fro### [at] thefrozenno-ipcom> wrote in message
news:web.414f7cacbf1ad50e102354040@news.povray.org...
> Hey Greybeard,
>
> thank you very much for your efforts on testing it! I don't have any idea,
> what might cause these very slow renderings. I guess, I will have to do
> some more tests, when the picture it's rendering now is done.. This will
> still take a while.. 1d2h working, 43% reached... :-(
>
One more test, well, two actually. Arbre with grass and trees remmed out,
19 sec.
Original file as downloaded, 33 min 6 sec,
Win2k, 2.6 ghz Pentium 4c, 1 gig memory. P-4c has the 800mhz buss, again
rendered 800 X 600 AA 0.3
POV-Ray 3.5 though, don't remember which version you're running.
Greybeard
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> One more test, well, two actually. Arbre with grass and trees remmed out,
> 19 sec.
> Original file as downloaded, 33 min 6 sec,
> Win2k, 2.6 ghz Pentium 4c, 1 gig memory. P-4c has the 800mhz buss, again
> rendered 800 X 600 AA 0.3
>
MY ERROR! No AA. Sorry bout that.
Greybeard
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> I guess, the time can't be reduced, but I would like to understand, WHY it
> takes so long?!
Can't help you right now, but make sure you safe the render statistics
POV spits out at the end of the render (the whole table). Post them
here, they often tell you what's going wrong ;)
Florian
--
This may seem a bit weird, but that's okay,
because it is weird. (perldata)
[------------ http://www.torfbold.com - POV-Ray gallery ------------]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
@greybeard:
Thanks again! So you have rather fast renderings, even with AA and I have
very slow ones, even without AA at much smaller resolutions.. This is
something I don't understand..
I'm using povray 3.6.1
@florian:
Do you know, how to save these messages using the windows version? If I can
save them, I will post these stats here, when the pic is done.. 68% are
done after 1d14h...
Thanking you all,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
@florian:
Here are the latest stats:
Remove bounds........On
Split unions.........Off
Library paths:
i:ProgrammePOV-Ray for Windows v3.6INCLUDE
C:WINNTFonts
Output Options
Image resolution 800 by 600 (rows 1 to 600, columns 1 to 800).
Output file: I:downloadsarbrechess.bmp, 24 bpp (system format)
Graphic display......On (gamma: 2.2)
Mosaic preview.......Off
CPU usage histogram..Off
Continued trace......Off
Tracing Options
Quality: 9
Bounding boxes.......On Bounding threshold: 3
Light Buffer.........On
Vista Buffer.........On Draw Vista Buffer....Off
Antialiasing.........On (Method 1, Threshold 0.100, Depth 3, Jitter 0.50)
Clock value: 0.000 (Animation off)
File: mgrass.pov Line: 205
Parse Warning: fgrass2.inc mesh file creation starts
File: mgrass.pov Line: 240
Parse Warning: fgrass2.inc file created
Cleanup Parse Warning: This rendering uses the following experimental
feature(s): radiosity. The
design and implementation of these features is likely to change in future
versions of POV-Ray. Full backward compatibility with the current
implementation is NOT
guaranteed.
Scene Statistics
Finite objects: 156821
Infinite objects: 3
Light sources: 1
Total: 156825
Possible Rendering Error: Maximum trace level reached! If your scene
contains black spots read
more about the max_trace_level setting in the documentation!
Render Statistics
Image Resolution 800 x 600
Pixels: 481405 Samples: 3058125 Smpls/Pxl: 6.35
Rays: 7019005 Saved: 1579 Max Level: 5/5
Ray->Shape Intersection Tests Succeeded Percentage
Box 639759 543867 85.01
Cone/Cylinder 17402 7559 43.44
CSG Intersection 8701 2259 25.96
Mesh 24525945 6537704 26.66
Plane 12966658 3319793 25.60
Sphere 25949631 19836331 76.44
Superellipsoid 4011692 1897812 47.31
Triangle 595668 32490 5.45
Bounding Box 349230262435 540399421 0.15
Light Buffer 427137499824 427137499824 100.00
Vista Buffer 159623429 120567531 75.53
Calls to Noise: 18065575 Calls to DNoise: 70823911
Shadow Ray Tests: 9515610 Succeeded: 5499837
Transmitted Rays: 1297848
Radiosity samples calculated: 81206 (1.34 %)
Radiosity samples reused: 5987420
Smallest Alloc: 18 bytes
Largest Alloc: 75497488 bytes
Peak memory used: 370198107 bytes
Total Scene Processing Times
Parse Time: 0 hours 29 minutes 28 seconds (1768 seconds)
Photon Time: 0 hours 0 minutes 0 seconds (0 seconds)
Render Time: 43 hours 43 minutes 25 seconds (157405 seconds)
Total Time: 44 hours 12 minutes 53 seconds (159173 seconds)
CPU time used: kernel 104.26 seconds, user 135555.17 seconds, total
135659.43 seconds
Render averaged 3.54 PPS over 480000 pixels
POV-Ray finished
Do you see anything, which explains the looooong time?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Possible Rendering Error: Maximum trace level reached! If your scene
> contains black spots read
>
> more about the max_trace_level setting in the documentation!
You say you have superellipsoids stacked one on top of another. What is
their texture? Are they reflective or transparent? Both?
- Slime
[ http://www.slimeland.com/ ]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Slime" <fak### [at] emailaddress> wrote:
> You say you have superellipsoids stacked one on top of another. What is
> their texture? Are they reflective or transparent? Both?
#declare StoneWallStoneTexture=
texture
{
T_Stone8
normal { dents scale 0.2 }
finish { phong 0 specular 0 crand 0.1 }
}
#declare StoneWallRGBlow = <0.7,0.5,0.35>;
#declare StoneWallRGBhigh = <1,0.9,.7>;
#declare StonePillarStoneTexture=
texture
{
T_Stone9
normal { dents scale 0.2 }
finish { phong 0 specular 0 crand 0.1 }
}
I hope, this answers your question? ;-)
Thanks,
frozen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"frozen" <fro### [at] thefrozenno-ipcom> wrote in message
news:web.4150a14ebf1ad50e9c688d540@news.povray.org...
> @florian:
>
> Antialiasing.........On (Method 1, Threshold 0.100, Depth 3, Jitter
0.50)
The threshold isn't helping any, as far as render times. I normally use .3,
very rarely, .2 (Renders to check progress, no AA, it can speed things
considerably.) I don't know how much this is really slowing it down, maybe
someone else can venture a guess.
Greybeard
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Ray->Shape Intersection Tests Succeeded Percentage
> Bounding Box 349230262435 540399421 0.15
> Do you see anything, which explains the looooong time?
I could be wrong, but 0.15 percentage of successful bounding box tests
looks like a bounding problem to me. Apart from manual bounding, these
are often caused by differences: Check the ones in your scene (you don't
need many to slow things down a lot) and try to bound them manually. Or
check if some methods/tips mentioned in the Holes Tutorial
(http://www.econym.demon.co.uk/holetut/index.htm) help (render the image
in a low resolution with no antialiasing to test these methods). The
tutorial also explains what's the problem with differences in POV-Ray.
If none of that helps, I usually try to spot the problem by removing
things from the scene step by step until I find out which objects slow
the render down a lot. Once you found these, it's often a lot easier to
pin down the problem.
HTH,
Florian
--
Minimum Requirement #8: Emacs is for experts. Joe is for beginners. VI
is a disease. (WEB.DE help-wanted ad: Linux Guru)
[------------ http://www.torfbold.com - POV-Ray gallery ------------]
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
are you performing any "differences" with the wall? that could slow things
down considerably. I've written wall type macros that use superellipsoids,
rounded boxes, or regular boxes, and while superellipsoids were the slowest,
it never took several hours even on my 800Mhz duron. Differencing a doorway
or a window in the wall would spike the render times into the several hours
range for sure.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
