 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Greetings Mr. Nathan Kopp Sir,
A couple of quick questions in regards to the Photon feature of your
newest patch. There are statistics available when using the photon feature.
The first is a counter in the lower left of the window frame. If you are
lucky enough to get your scene description set up correctly it starts to
display two pieces of information when parsing the scene. One number to
the left for the photon map and a second number in parenthesis. What are
the two numbers and what may be learned from them as they are displayed ?
Second question is in regards to the statistics reported in the message
window when a render is finished. Here you have as many as four statistics
returned - that is the most I have seen so far anyway.
For example I just finished a render with these reported results:
Photons:
Shot - 245,000
Stored - 117,000
Priority Que:
Insert - 124,000
Removed - 165
I would like to know what this is telling me in regards to the
numbers returned. Do I need to adjust my choice of parameters,
is it is in the ballpark for the process, or this is just useless
statistics put there to make the process look more important but
actually tells you nothing ? (just kidding on the last part :) ).
Thirdly in regards to using the reflection feature of the Photon
process I would like to know what kind of intensity return to
expect from a reflected ray after one bounce. I don't believe I
am seeing anything close to 100% coming off of a smooth surface
with a reflection value of 1 and landing on a surface with a
diffuse 1 ambient 0 finish statement.
The reflected light seems more on the order of something closer
to 60% - 70% of the original. Is this again a result of user
chosen parameters or is this a known characteristic of the process ?
Not that I am complaining mind you. Even 15% returned of the
reflected light intensity is better than nothing at all.
Thank you for taking the time to read this and I am really enjoying
the new art coming from my computer.
You dun guud things Mr. Nathan Kopp Sir.
--
Ken Tyler
mailto://tylereng@pacbell.net
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote:
>
> Greetings Mr. Nathan Kopp Sir,
>
> A couple of quick questions in regards to the Photon feature of your
> newest patch. There are statistics available when using the photon feature.
>
> The first is a counter in the lower left of the window frame. If you are
> lucky enough to get your scene description set up correctly it starts to
> display two pieces of information when parsing the scene. One number to
> the left for the photon map and a second number in parenthesis. What are
> the two numbers and what may be learned from them as they are displayed ?
This counter gets updated every 500 photons that are shot. The first number is
how many photons have been shot at the current object. The number in
parentheses is how many total photons have been stored in the scene. Thus, the
first number will keep getting reset (as POV finishes one light/object
combination and moves on to another). For example, if you have three lights
shooting photons at three objects, you'll see this first counter start at zero
3x3=9 times. The second counter will just keep accumulating. This is useful
in a few ways. First, if you notice that one object gets a lot more photons
than another object, you may need to adjust the settings (well, unless one is
tiny and the other object is huge). Secondly, you can see if the photons
getting shot are actually being stored or not. They will not be stored if
either a) they miss the object they were shot at, or b) they hit the object but
never hit anything else in the scene (bounce off into infinity).
> Second question is in regards to the statistics reported in the message
> window when a render is finished. Here you have as many as four statistics
> returned - that is the most I have seen so far anyway.
>
> For example I just finished a render with these reported results:
>
> Photons:
> Shot - 245,000
> Stored - 117,000
Like I explained above, not all photons shot are stored. The photons are shot
at the object's bounding box, so some are bound to miss the object. The ratio
of shot to stored that you have here is pretty good.
>
> Priority Que:
> Insert - 124,000
> Removed - 165
This one was for some profiling/debugging that I was doing. When photons are
gathered, they are stored in a priority queue, so that only the ones closest to
the gather-point are kept. For example, if you use "gather 20,100" and more
than 100 photons are found in the search radius, some will have to be discarded
(removed from the priority queue). So in this scene, you very rarely gathered
more than the maximum number of photons. This is a good thing for render time.
> Thirdly in regards to using the reflection feature of the Photon
> process I would like to know what kind of intensity return to
> expect from a reflected ray after one bounce. I don't believe I
> am seeing anything close to 100% coming off of a smooth surface
> with a reflection value of 1 and landing on a surface with a
> diffuse 1 ambient 0 finish statement.
> The reflected light seems more on the order of something closer
> to 60% - 70% of the original. Is this again a result of user
> chosen parameters or is this a known characteristic of the process ?
> Not that I am complaining mind you. Even 15% returned of the
> reflected light intensity is better than nothing at all.
If you use reflection 1, you should get 100% reflection. Actually,
I just ran a test comparing a real (photon) reflection with a fake
reflection (mad with a light shining through a hole) and they
were of the same brightness. I'll cut&paste the scene below.
> Thank you for taking the time to read this and I am really enjoying
> the new art coming from my computer.
Your welcome.
> You dun guud things Mr. Nathan Kopp Sir.
Thanks!
-Nathan Kopp
================== reflection test code ===================
#include "colors.inc"
#include "textures.inc"
#include "stones.inc"
/*
This scene has a floor plane, a wall, and two 'mirrors'.
The mirror on the right is a real mirror that uses photon mapping.
The 'mirror' on the left is really just a hole cut in the wall
so that a light source on the other side can shine through so
that it looks like a mirror.
*/
#declare phd=.6; // lots of photons for sharp edge
global_settings{
ambient_light 0
photons {
gather 20, 100
radius .05*phd*sqrt(20)/2, 1 // only one gather size
// (all photons are the same density)
autostop 0
jitter .4
}
}
camera
{
location <6,2.8,0>
look_at <0,0,0>
}
// here's the 'real' light source
light_source{
<5,5,0>
color White
}
// place another light source on the other side
// of the wall to do a fake reflection through
// a hole in the wall
light_source{
<-5,5,0>
color White
photons{reflection off refraction off} // don't bother
}
// flore plane
plane{y,0 pigment{checker}}
// the wall
difference{
box{-1,1 scale<.05,100,100>}
box{ // hole for fake mirror
-1,1
translate <0,1,-1.5>
}
box{ // hole for real mirror
-1,1
translate <0,1,1.5>
}
pigment{checker Red,White}
}
box{ // real mirror
-1,1
scale<.05,1,1>
translate <0,1,1.5>
pigment{ color White }
finish{ diffuse 0 ambient 0 reflection 1}
photons {
density 0.02*phd//*1.5
reflection on
ignore_photons
}
}
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
A gather size of
radius .09*phd*sqrt(20)/2, 1 // only one gather size
// (all photons are the same density)
works better on that test scene I posted.
If your gather size is too small, your caustics (both reflective and
refractive) may be a bit darker than they should be. I'll have
to test this more.
-Nathan
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Spider jumps in here.
Ken wrote:
>
> Greetings Mr. Nathan Kopp Sir,
>
> A couple of quick questions in regards to the Photon feature of your
> newest patch. There are statistics available when using the photon feature.
Hmm, I'd like to see them even If I don't finish the render, taht would be
useful to tweak the settings and don't wait for eternety.
> The first is a counter in the lower left of the window frame. If you are
> lucky enough to get your scene description set up correctly it starts to
> display two pieces of information when parsing the scene. One number to
> the left for the photon map and a second number in parenthesis. What are
> the two numbers and what may be learned from them as they are displayed ?
As far as I know..
Theese go once for every object that photons affect.
left number is shot photons(increasing by 500) and right number is stored
photons.
theese go for all objects.
> Second question is in regards to the statistics reported in the message
> window when a render is finished. Here you have as many as four statistics
> returned - that is the most I have seen so far anyway.
I like stats :-)
> For example I just finished a render with these reported results:
>
> Photons:
> Shot - 245,000
> Stored - 117,000
It means that about 45% of all photons actually hit an object and therefore is
stored.
> Priority Que:
> Insert - 124,000
> Removed - 165
noo idea.
> I would like to know what this is telling me in regards to the
> numbers returned. Do I need to adjust my choice of parameters,
> is it is in the ballpark for the process, or this is just useless
> statistics put there to make the process look more important but
> actually tells you nothing ? (just kidding on the last part :) ).
Well, the first are relevant for your settings, if they are to big
difference(100 000 / 10 000 ratio) you really need to tweak. :-)
<snip number 3>
> Thank you for taking the time to read this and I am really enjoying
> the new art coming from my computer.
same here :-)
> You dun guud things Mr. Nathan Kopp Sir.
and here :-)
--
//Spider
[ spi### [at] bahnhof se ]-[ http://www.bahnhof.se/~spider/ ]
What I can do and what I could do, I just don't know anymore
"Marian"
By: "Sisters Of Mercy"
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Spider wrote:
>
> (good answers)
Look at the times our messages were posted. A tie! hehehe. :-)
-Nathan
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Nathan Kopp wrote:
>
> Spider wrote:
> >
> > (good answers)
>
> Look at the times our messages were posted. A tie! hehehe. :-)
>
> -Nathan
No, wait... you beat me by 22 seconds.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
hehe, but I think you have better answers..
Nathan Kopp wrote:
>
> Nathan Kopp wrote:
> >
> > Spider wrote:
> > >
> > > (good answers)
> >
> > Look at the times our messages were posted. A tie! hehehe. :-)
> >
> > -Nathan
>
> No, wait... you beat me by 22 seconds.
--
//Spider
[ spi### [at] bahnhofse ]-[ http://www.bahnhof.se/~spider/ ]
What I can do and what I could do, I just don't know anymore
"Marian"
By: "Sisters Of Mercy"
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Nathan Kopp wrote:
>
> A gather size of
>
> radius .09*phd*sqrt(20)/2, 1 // only one gather size
> // (all photons are the same density)
>
> works better on that test scene I posted.
>
> If your gather size is too small, your caustics (both reflective and
> refractive) may be a bit darker than they should be. I'll have
> to test this more.
>
> -Nathan
Thank you for your replies to my questions. In regards to the
term caustics as used above I have one more little question.
With the photon patch is it necessary to use the caustics interior
modifier for the caustic reflections ? It doesn't seem likely to me
but I have not heard explicitly one way or another on the this and
would like a clarification for sake of proceedural accuracy.
Again thanks and that disco ball is almost finished. :)
--
Ken Tyler
mailto://tylereng@pacbell.net
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote:
>
> Thank you for your replies to my questions. In regards to the
> term caustics as used above I have one more little question.
> With the photon patch is it necessary to use the caustics interior
> modifier for the caustic reflections ? It doesn't seem likely to me
> but I have not heard explicitly one way or another on the this and
> would like a clarification for sake of proceedural accuracy.
The 'caustics' keyword will have no effect on photon mapping.
-Nathan
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Actually, that keyword only get in the way, sometimes hiding the "true"
caustics..
Nathan Kopp wrote:
>
> Ken wrote:
> >
> > Thank you for your replies to my questions. In regards to the
> > term caustics as used above I have one more little question.
> > With the photon patch is it necessary to use the caustics interior
> > modifier for the caustic reflections ? It doesn't seem likely to me
> > but I have not heard explicitly one way or another on the this and
> > would like a clarification for sake of proceedural accuracy.
>
> The 'caustics' keyword will have no effect on photon mapping.
>
> -Nathan
--
//Spider
[ spi### [at] bahnhofse ]-[ http://www.bahnhof.se/~spider/ ]
What I can do and what I could do, I just don't know anymore
"Marian"
By: "Sisters Of Mercy"
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Nathan Kopp wrote:
>
> Ken wrote:
> >
> > Thank you for your replies to my questions. In regards to the
> > term caustics as used above I have one more little question.
> > With the photon patch is it necessary to use the caustics interior
> > modifier for the caustic reflections ? It doesn't seem likely to me
> > but I have not heard explicitly one way or another on the this and
> > would like a clarification for sake of proceedural accuracy.
>
> The 'caustics' keyword will have no effect on photon mapping.
>
> -Nathan
I suspected as much but wanted it confirmed ono the less. Thanks.
I just posted your request for a photon patch, user generated
disco like mirrored ball, reflecting multiple beams of colored
light, which appear as spots of multicoloured lights illuminating
the surrounding walls, into the povray.binaries.images group.
I would just like to go on record after this little exercise as
saying it sure would have been easier had I been able to use spot
lights for the light sources. Trying to hide 60 colored lights with
an intensity of 1.8 is a challenge unto itself without the aid of
the programs spotlight light_source modifier. Had I not hidden them
they would have washed out the desired effects. In consequence they
had to be obscured from everything but the reflective surfaces they
were pointed at. Not insurmountable but but uniquely challenging to
say the least.
If anyone wants the source for this to do an animation with I am more
that willing to hand it off to someone else to peruse. A 600x400 with
the current settings took 15.6 minutes to render and maybe 1.5 to parse
on my P-200 machine.
--
Ken Tyler
mailto://tylereng@pacbell.net
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken <tyl### [at] pacbellnet> wrote:
: If anyone wants the source for this to do an animation with I am more
: that willing to hand it off to someone else to peruse. A 600x400 with
: the current settings took 15.6 minutes to render and maybe 1.5 to parse
: on my P-200 machine.
I can make that animation in my PII-350 (with a limitation: I don't have
net connection at home, so I will only carry the resulting mpeg in
disquettes, not all the frames...)
--
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote:
>
> I would just like to go on record after this little exercise as
> saying it sure would have been easier had I been able to use spot
> lights for the light sources. Trying to hide 60 colored lights with
> an intensity of 1.8 is a challenge unto itself without the aid of
> the programs spotlight light_source modifier. Had I not hidden them
> they would have washed out the desired effects. In consequence they
> had to be obscured from everything but the reflective surfaces they
> were pointed at. Not insurmountable but but uniquely challenging to
> say the least.
60 lights!!! wow! You could have used spotlights, by the way. It's
just that the photon part treats them as point lights... meaning that,
while during the regular trace part they'll act like spotlights, photons
can shoot out of them in any direction. Of course, this won't always
be the case... I just didn't code that part yet.
-Nathan
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
