POV-Ray : Newsgroups : povray.general : Photon mapping problem Server Time
7 Nov 2024 00:18:55 EST (-0500)
  Photon mapping problem (Message 1 to 2 of 2)  
From: Skip Talbot
Subject: Photon mapping problem
Date: 26 Feb 2004 18:13:51
Message: <403e7daf$1@news.povray.org>
Here is the setup:  I've got about 20 diamonds that are photon targets, and
I'm shooting 30 million photons at them.  It takes about 30 minutes to map
the photons, which is no problem.  However when I add in several hundred
very small diamond objects, the photon mapping process pretty much stalls.
At a count of 30 million it has been mapping now for over 19 hours.  I have
taken lengths to try and prevent the added objects from interfering with the
photon process.  All diamonds have collect off.  The smaller diamonds do not
have the target keyword and pass_through is on.  Because of this I thought
the photons would basically ignore these new objects but apparently not.
Any insight or suggestions available?  Am I looking at simplifying my scene?

Skip


Post a reply to this message

From: Tom Melly
Subject: Re: Photon mapping problem
Date: 27 Feb 2004 04:33:21
Message: <403f0ee1$1@news.povray.org>
"Skip Talbot" <sta### [at] uiucedu> wrote in message
news:403e7daf$1@news.povray.org...
> Here is the setup:  I've got about 20 diamonds that are photon targets, and
> I'm shooting 30 million photons at them.  It takes about 30 minutes to map
> the photons, which is no problem.  However when I add in several hundred
> very small diamond objects, the photon mapping process pretty much stalls.
> At a count of 30 million it has been mapping now for over 19 hours.  I have
> taken lengths to try and prevent the added objects from interfering with the
> photon process.  All diamonds have collect off.  The smaller diamonds do not
> have the target keyword and pass_through is on.  Because of this I thought
> the photons would basically ignore these new objects but apparently not.
> Any insight or suggestions available?  Am I looking at simplifying my scene?
>

It *sounds* as though the small diamonds shouldn't have an impact on the
photon-mapping part of the render - I wondered if the problem was less with the
photons and more with the refraction of the caustics as viewed through the
smaller diamonds. But that seems unlikely since, afaik, you would notice that
when the scene was actually being rendered rather than in any pre-render
calculations. Can you post a sample scene? (assuming a nice neat macro generates
your diamonds).

For testing, why not reduce your photon count and compare with/without small
diamonds?


Post a reply to this message

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