If I make a cylinder like this:

cylinder { -y*2,y*2,1 material { Glass } photons { ... } }

and put it on a plane, I get a nice caustic highlight on the plane, but if I
make the cylinder hollow this way:

difference
{ cylinder { -y*2,y*2,1 }
  cylinder { -y*1.8,y*2.01,.8 }
  material { Glass }
  photons { ... }
}

I don't get any caustic highlight anymore.

  This may:
  a) just work fine. I don't get a caustic higlight on the plane because I
shouldn't, because this kind of object doesn't refract the light in that way.
  b) be a bug in the program.
  c) be because I haven't understood how photons work.

  Which one of those answers is correct and why?

-- 
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

Check
 a) max_trace_level... more surfaces to go through (this should be OK,
      though)
 b) autostop angle.  If autostop is zero and the first few photons go through
      the hole in the middle, it will stop right there.  setting autostop
      to a higher angle forces the patch to keep going in the spiral pattern
      until at least the autostop angle is reached (try 45 degrees).
      You'll know if autostop is the problem if the "Building photon
      maps..." step takes almost no time.

Are there no photons (black shadow) or just no highlight (normal looking
shadow)?

-Nathan

Nieminen Mika wrote:
> 
>   If I make a cylinder like this:
> 
> cylinder { -y*2,y*2,1 material { Glass } photons { ... } }
> 
> and put it on a plane, I get a nice caustic highlight on the plane, but if I
> make the cylinder hollow this way:
> 
> difference
> { cylinder { -y*2,y*2,1 }
>   cylinder { -y*1.8,y*2.01,.8 }
>   material { Glass }
>   photons { ... }
> }
> 
> I don't get any caustic highlight anymore.
> 
>   This may:
>   a) just work fine. I don't get a caustic higlight on the plane because I
> shouldn't, because this kind of object doesn't refract the light in that way.
>   b) be a bug in the program.
>   c) be because I haven't understood how photons work.
> 
>   Which one of those answers is correct and why?
> 
> --
> 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

Nathan Kopp <Nat### [at] koppcom> wrote:
: Are there no photons (black shadow) or just no highlight (normal looking
: shadow)?

  No highlight. It looks just like the photons are dispersed so that they
don't make a highlight instead of being grouped in a very little area like
with the solid cylinder. Perhaps that's just the way it works...

  Another question:

  Do the program stop sending more photons after certain number of
reflections/refractions are performed and new ones will not affect
noticeably the illumination of the scene? What I mean is that if there
is some 'confidence' value when calculating photons, just like povray
does when sending rays from the camera (I'm talking about the ADC_Bailout
value in global_settings or 'confidence' in focal blurred camera).

  I ask this because I'm continuously running out of memory even with
the simplest scenes (I have 64 megs of ram).
  I have a scene rendering right now. It have been rendering for over a
day (ie. 24 hours) now. It spends about 99% of the time just swapping
(the image is rendered in less than a half of an hour with less photons,
but it looks blotchy).

  I think that it would be a very good idea if you could set the photons
'confidence' value in the global_settings block, just like with the focal
blur settings in the camera.
  This way you could set a very high max_trace_level and also a very high
confidence in photons and you shouldn't run out of memory.

  Btw, I hope you will compile a version of the patch with pov 3.1d/e soon
because rendering lathes with the current version is painly slow... :)

-- 
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

Nieminen Mika wrote:
> 
> Nathan Kopp <Nat### [at] koppcom> wrote:
> : Are there no photons (black shadow) or just no highlight (normal looking
> : shadow)?
> 
>   No highlight. It looks just like the photons are dispersed so that they
> don't make a highlight instead of being grouped in a very little area like
> with the solid cylinder. Perhaps that's just the way it works...

I think it's just the way it works.

>   Another question:
> 
>   Do the program stop sending more photons after certain number of
> reflections/refractions are performed and new ones will not affect
> noticeably the illumination of the scene? What I mean is that if there
> is some 'confidence' value when calculating photons, just like povray
> does when sending rays from the camera (I'm talking about the ADC_Bailout
> value in global_settings or 'confidence' in focal blurred camera).

The patch right now just uses the max_trace_level of the normal trace
loop.  I want to set it up so you get a photon-specific max_trace and
adc_bailout.

>   I ask this because I'm continuously running out of memory even with
> the simplest scenes (I have 64 megs of ram).
>   I have a scene rendering right now. It have been rendering for over a
> day (ie. 24 hours) now. It spends about 99% of the time just swapping
> (the image is rendered in less than a half of an hour with less photons,
> but it looks blotchy).

That's no good.  I really think you have too many photons.  ;-)

>   I think that it would be a very good idea if you could set the photons
> 'confidence' value in the global_settings block, just like with the focal
> blur settings in the camera.
>   This way you could set a very high max_trace_level and also a very high
> confidence in photons and you shouldn't run out of memory.
> 
>   Btw, I hope you will compile a version of the patch with pov 3.1d/e soon
> because rendering lathes with the current version is painly slow... :)

Hmmm... I thought of that... unfortunately, it takes a lot of work to move
all of my various changes (especially UV mapping) into a new set of files.
Maybe it would simply be easier for me to update the lathe rendering code.
Does anyone know which lathe-intersection (or polynomial-solving) functions
changed?  (Ron, does the superpatch have the fast lathe intersections?)

-Nathan

Post a reply to this message

On Tue, 23 Mar 1999 10:58:35 -0500, Nathan Kopp <Nat### [at] Koppcom> wrote:
>Hmmm... I thought of that... unfortunately, it takes a lot of work to move
>all of my various changes (especially UV mapping) into a new set of files.
>Maybe it would simply be easier for me to update the lathe rendering code.
>Does anyone know which lathe-intersection (or polynomial-solving) functions
>changed?  (Ron, does the superpatch have the fast lathe intersections?)

I don't know.  My personal copy has been merged with the generic 3.1e 
source, but I haven't taken the time to see what changed since 3.1a.  I 
expect that it will, whenever I get it put online.  I have all the patch 
files at home, so I can easily check to see what changed in the lathe 
routines if you'd like.

Post a reply to this message

Nathan Kopp <Nat### [at] koppcom> wrote:
: Hmmm... I thought of that... unfortunately, it takes a lot of work to move
: all of my various changes (especially UV mapping) into a new set of files.
: Maybe it would simply be easier for me to update the lathe rendering code.
: Does anyone know which lathe-intersection (or polynomial-solving) functions
: changed?  (Ron, does the superpatch have the fast lathe intersections?)

  Note that lathe (and sor and...) speed is not the only feature. The
bug fixes are also very important.

-- 
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

Ron Parker wrote:
> 
> I don't know.  My personal copy has been merged with the generic 3.1e
> source, but I haven't taken the time to see what changed since 3.1a.  I
> expect that it will, whenever I get it put online.  I have all the patch
> files at home, so I can easily check to see what changed in the lathe
> routines if you'd like.

I sent you an email about photon mapping + superpatch.  If you email the
superpatch source (with 3.1e changes, if possible) to nko### [at] megsinetcom,
I'll try to merge photon mapping into it.  (Let me know if you mail it
to that address, since I almost never check it... and don't mail it to
my regular address, please... I'm running out of space there, and it's
not POP3).

-Nathan

Post a reply to this message