 |
|
|
|
|
|
| |
| |
|
|
From: Tim Kaulmann
Subject: Speeding up glass-raytracing / Pentium optimization
Date: 19 Sep 1999 16:42:27
Message: <37e54ab3@news.povray.org>
|
|
|
| |
| |
|
|
Hello there !
Is there a way to speed up the calculation of a scene that has 3
glass-objects with reflection and refraction in it ?
At a certain point my POV3.1 for Win slows down to 2 pps on my Pentium MMX
233 (128 MB RAM) :-(
On the other hand the output of this tracer ist so great that I go
completely crazy :-)=)
Is there a version that is optimized for Pentium MMX only ?
Greetings, -= Tim =-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Tim Kaulmann wrote:
>
> Hello there !
>
> Is there a way to speed up the calculation of a scene that has 3
> glass-objects with reflection and refraction in it ?
Not really. You can try lowering the reflection component which is usualy
what causes the greatest slow down due to internal reflection calculations.
A high ior value in the interior statement will also slow you down but not
as much as the reflection value will. Also make sure that you have a -
max_trace_level specified in the global settings statement. Start with
something like 10 - 15 to begin with. It may not speed it up but it does
help clear refractive objects come out looking better in many cases.
> At a certain point my POV3.1 for Win slows down to 2 pps on my Pentium MMX
> 233 (128 MB RAM) :-(
> On the other hand the output of this tracer ist so great that I go
> completely crazy :-)=)
Welcome to the wonderful world of raytracing :)
> Is there a version that is optimized for Pentium MMX only ?
None that I am aware of and the instruction set for the MMX is not really
an aid to raytracing anyway. If you are using the pentium optimized version
of the program under windows you are most likely seeing all of the speed
you are likely to get. If you are not already using the MSVC6 version as
opposed to the watcom compile you may wish to download that from povray.org.
It does give some speed imporvements over the watcom compile of the program.
--
Ken Tyler
See my 1000+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html
Post a reply to this message
|
|
| |
| |
|
|
From: Nieminen Juha
Subject: Re: Speeding up glass-raytracing / Pentium optimization
Date: 20 Sep 1999 06:41:46
Message: <37e60f6a@news.povray.org>
|
|
|
| |
| |
|
|
Also try increasing the adc_bailout value from the default 1/128 (if
I remember right).
--
main(i,_){for(_?--i,main(i+2,"FhhQHFIJD|FQTITFN]zRFHhhTBFHhhTBFysdB"[i]
):5;i&&_>1;printf("%s",_-70?_&1?"[]":" ":(_=0,"\n")),_/=2);} /*- Warp -*/
Post a reply to this message
|
|
| |
| |
|
|
From: Ty Matthews
Subject: Re: Speeding up glass-raytracing / Pentium optimization
Date: 20 Sep 1999 15:36:47
Message: <37e68ccf@news.povray.org>
|
|
|
| |
| |
|
|
Tim Kaulmann <T.K### [at] Zentral de> wrote in message
news:37e54ab3@news.povray.org...
> At a certain point my POV3.1 for Win slows down to 2 pps on my Pentium MMX
> 233 (128 MB RAM) :-(
I can tell I'm in for a fun ride. I only have a Pentium 150.
Just out of curiosity, though. Who holds the record for the longest time
letting their computer render an image and where is the image? It would
take mine probably a week and a half to render a 1280x1024 image, adding an
extra three and a half days over Tim's 2 pps.
Ty Matthews
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ty Matthews wrote:
>
> Tim Kaulmann <T.K### [at] Zentralde> wrote in message
> news:37e54ab3@news.povray.org...
>
> > At a certain point my POV3.1 for Win slows down to 2 pps on my Pentium MMX
> > 233 (128 MB RAM) :-(
>
> I can tell I'm in for a fun ride. I only have a Pentium 150.
Say good night Ty :)
> Just out of curiosity, though. Who holds the record for the longest time
> letting their computer render an image and where is the image? It would
> take mine probably a week and a half to render a 1280x1024 image, adding an
> extra three and a half days over Tim's 2 pps.
>
> Ty Matthews
I have no idea what the record is nor where the image may be. I would easily
speculate though that some renders in the early days of the program when
people were running on 286's some lasted in the range of months. I currently
have a render going on a p200 that is now 52% complete at 29 hrs. with not
a single glass object in sight.
--
Ken Tyler
See my 1000+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html
Post a reply to this message
|
|
| |
| |
|
|
From: TonyB
Subject: Re: Speeding up glass-raytracing / Pentium optimization
Date: 20 Sep 1999 18:22:56
Message: <37e6b3c0@news.povray.org>
|
|
|
| |
| |
|
|
>I have no idea what the record is nor where the image may be. I would
easily
>speculate though that some renders in the early days of the program when
>people were running on 286's some lasted in the range of months.
I know this fellow who 'jumped' from a 286 to a 233 Pentium with MMX... he
said he was "waiting for the advancement of technology to slow down before
he got on the bandwagon".
>I currently
>have a render going on a p200 that is now 52% complete at 29 hrs. with not
>a single glass object in sight.
It had better be spectacular... and posted within the first 5 minutes of
being done rendering. ;)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
TonyB wrote:
> >I currently
> >have a render going on a p200 that is now 52% complete at 29 hrs. with not
> >a single glass object in sight.
>
> It had better be spectacular... and posted within the first 5 minutes of
> being done rendering. ;)
At 1 pps it will take a while longer but yes I was planning on posting it
when done. It's using the method 2 media. Faster ? Hmmm...
--
Ken Tyler
See my 1000+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html
Post a reply to this message
|
|
| |
| |
|
|
From: TonyB
Subject: Re: Speeding up glass-raytracing / Pentium optimization
Date: 20 Sep 1999 22:48:19
Message: <37e6f1f3@news.povray.org>
|
|
|
| |
| |
|
|
> At 1 pps it will take a while longer but yes I was planning on posting it
>when done. It's using the method 2 media. Faster ? Hmmm...
Have you tried it with lower intervals and sampling? If you put them up to
high, it will be just as slow as the other one...
Post a reply to this message
|
|
| |
| |
|
|
From: Bob Hughes
Subject: Re: Speeding up glass-raytracing / Pentium optimization
Date: 21 Sep 1999 01:41:03
Message: <37e71a6f@news.povray.org>
|
|
|
| |
| |
|
|
Using 'scattering' and 'absorption' seem typically slow things as
well. Not as slow as before I guess but of course there's still going
to be that slowdown involved.
Bob
TonyB <ben### [at] panamaphoenixnet> wrote in message
news:37e6f1f3@news.povray.org...
> > At 1 pps it will take a while longer but yes I was planning on
posting it
> >when done. It's using the method 2 media. Faster ? Hmmm...
>
> Have you tried it with lower intervals and sampling? If you put them
up to
> high, it will be just as slow as the other one...
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Bob Hughes wrote:
>
> Using 'scattering' and 'absorption' seem typically slow things as
> well. Not as slow as before I guess but of course there's still going
> to be that slowdown involved.
>
> Bob
The media I am using probably is a worse case scenario but I wanted to
see if the media would turn out non-grainy and so far it's looking good.
Other secrets to be revealed later...
--
Ken Tyler
See my 1000+ Povray and 3D Rendering and Raytracing Links at:
http://home.pacbell.net/tylereng/index.html
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
