 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
> Anthony D. Baye wrote:
>
>>
>> So my question is: Is there an eficient way to model Optical
>> fibers, so that they not only glow, but transmit the light along the
>> length as well.
>> Naturally, I'll continue my own experimentations. I would however
>> appreciate any thoughts people might have on the subject.
>
>
> The transmission of light is nothing that makes sense to simulate in
> POV-Ray since it has no effect on the image POV-Ray renders (light
> itself isn't visible). Therefore using emitting media for the glow and
> placing a spotlight at the end is the way to go.
>
> Christoph
>
Hey Christoph,
Thanks for the input. The only problem that I see with your solution
is that, if I were trying to model one of those fiber optic displays
(f/ex. The flowers with the FO strands in them) I would have hundreds of
spotlights in my scene. The light calculations would take forever.
Now I'll admit that what I'm trying to do may not be possible, but that
doesn't mean it's not worth investigating.
Regards,
A.D.B.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Bob Hughes wrote:
> "Christoph Hormann" <chr### [at] gmx de> wrote in message
> news:dcn984$3f2$1@chho.imagico.de...
>
>>The transmission of light is nothing that makes sense to simulate in
>>POV-Ray since it has no effect on the image POV-Ray renders (light itself
>>isn't visible). Therefore using emitting media for the glow and placing a
>>spotlight at the end is the way to go.
>
>
> I don't think I can debate you on that, especially since I'm not sure what
> you meant by "no effect on the image", but never say never. Not sure what
> all Anthony knows about fiber optics but I know only enough to describe it
> to people with no knowledge of it in the first place. ;)
>
> As I might have said elsewhere already, I was once able to get a simulated
> fiber optic cable made in POV and at least partially succeeded (maybe). It
> was done by using photons and reflection on the inside of an S-shaped
> sphere_sweep. I think it is total internal reflection effect, I believe,
> that is the operative factor IRL. Meaning, the boundary of the tube with the
> air is supposed to cause the light to stay within the tube walls because of
> a specific limiting angle whereby light only reflects instead of breaks
> through.
>
> My test scene wasn't very good for anything as I recall, and I was counting
> on faking it with reflection, not refraction. I'm not good enough with
> photons to be sure if it really works or not anyway. I guess there would be
> trouble due to how the photons are designed to map out into the scene, not
> very sure of anything about it though. Probably was never designed to bounce
> around inside of curving tubes.
>
> Sure hope you can manage to do something with your idea, Anthony. I just
> found the test scene I had done and rendered it as-is. Only got a bunch of
> red spots everywhere, mostly at the ground beyond the opposite end of the
> tube. Probably wouldn't help to see the scene yourself, it's basically just
> a narrow spotlight close to one end of that sphere_sweep (hollowed out with
> a smaller twin), which then has refraction and reflection including
> interior_texture.
>
> I just looked for info about TIR and fiber optics and there's mention of
> critical angles for the light entering the tube and the bends in the tube.
> Gets beyond me but this one web page allows for calculating the critical
> incidence angle for different ior, if you're interested.
>
> http://hyperphysics.phy-astr.gsu.edu/hbase/phyopt/totint.html
>
> Bob
>
>
Hi Bob,
I don't really know that much about them, only what a quick google
search could tell me.
That being said, I did find out that the coating layer has a lower IOR
than the fiber core. This is explained by that link you posted. When
passing through one medium strikes a medium with a lesser IOR, "the ray
is bent away from the normal." Thus, the angle of reflection will be
greater than the incident angle. (I can picture this. That doesn't mean
that I completely understand it.)
I do however see Christoph's point about max_trace_level.
Thanks for the link.
Regards,
A.D.B.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Cool idea!
> >>The transmission of light is nothing that makes sense to simulate in
> >>POV-Ray since it has no effect on the image POV-Ray renders (light itself
> >>isn't visible). Therefore using emitting media for the glow and placing a
> >>spotlight at the end is the way to go.
A MUCH faster way to get the "points of light" at the ends of the fibers
(like in a flower sculpture) might be to attach spheres at the end the same
diameter as the cable. Give them a high ambience.
>
> That being said, I did find out that the coating layer has a lower IOR
> than the fiber core. This is explained by that link you posted. When
> passing through one medium strikes a medium with a lesser IOR, "the ray
> is bent away from the normal." Thus, the angle of reflection will be
> greater than the incident angle. (I can picture this. That doesn't mean
> that I completely understand it.)
This might actually be made to work with photons, although be prepared to
have your swap file limit increased to an astonomical size and a LONG
render time! I'd love to experiment just now but MY machine is 15 hours
into a major render with photons and my swap file just expanded to over two
gigs :o( One potential problem I forsee is the dreaded co-incident surface
bug. Photons only refract when entering (or leaving, depending on your
point of view) an object. An object can only have one ior in it's
interior. So to simulate two ior's I'd put a solid sphere sweep with high
ior inside a larger, hollowed out (NOT hollow) sphere sweep with a lower
one. Make sure the inner one is a little bigger than the hollow in the
outer one is, which prevents a co-incident surface, and then merge the two
together to remove interior surfaces. A lot of this is personal conjecture
on my part but it's where I'll start next week when I get POV back again.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Lonnie wrote:
> Cool idea!
>
>
>>>>The transmission of light is nothing that makes sense to simulate in
>>>>POV-Ray since it has no effect on the image POV-Ray renders (light itself
>>>>isn't visible). Therefore using emitting media for the glow and placing a
>>>>spotlight at the end is the way to go.
>
>
> A MUCH faster way to get the "points of light" at the ends of the fibers
> (like in a flower sculpture) might be to attach spheres at the end the same
> diameter as the cable. Give them a high ambience.
>
>
>> That being said, I did find out that the coating layer has a lower IOR
>>than the fiber core. This is explained by that link you posted. When
>>passing through one medium strikes a medium with a lesser IOR, "the ray
>>is bent away from the normal." Thus, the angle of reflection will be
>>greater than the incident angle. (I can picture this. That doesn't mean
>>that I completely understand it.)
>
>
> This might actually be made to work with photons, although be prepared to
> have your swap file limit increased to an astonomical size and a LONG
> render time!
>
>
What's a swap file?
I'm on a Mac. maybe it's called something else.
A.D.B.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
> Anthony D. Baye wrote:
>
>>
>> So my question is: Is there an eficient way to model Optical
>> fibers, so that they not only glow, but transmit the light along the
>> length as well.
>> Naturally, I'll continue my own experimentations. I would however
>> appreciate any thoughts people might have on the subject.
>
>
> The transmission of light is nothing that makes sense to simulate in
> POV-Ray since it has no effect on the image POV-Ray renders (light
> itself isn't visible). Therefore using emitting media for the glow and
> placing a spotlight at the end is the way to go.
>
> Christoph
>
Something else I just thought of:
Light is radiation in the visible spectrum.
In the real world, you can't have radiation in the visible spectrum
without an object present to give off said radiation.
so when you look at a light bulb, it's not really the light you're
seeing, but the glowing filament, and the light bulb of course.
But this is fairly elementary stuff.
In the case of Optical Fibers, it's not really the light that's being
seen, but the reflection off the interior surface of the fiber.
I'll admit, Lonnie has a point about 'gluing' ambient spheres to the
edges of an object. I hadn't thought of this, because I have absolutely
no idea how to use the trace macro. (I assume I'd need it.)
It's nice that this is the subject of so much discussion though.
Regards,
A.D.B.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Anthony D. Baye" <Sha### [at] hotmailcom> wrote in message
news:42f0342b$1@news.povray.org...
<snip>
> so when you look at a light bulb, it's not really the light you're
> seeing, but the glowing filament, and the light bulb of course.
That's not exactly correct. It is the light/photons/radiation that you are
seeing, not the object.
The object is only seen because photons are being reflected off it, or because
it is producing photons. After all, our eyes can't sense objects (we use touch
for that), they can only sense light (and the polarisation of light).
:)
Lance.
thezone - thezone.firewave.com.au
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Anthony D. Baye" <Sha### [at] hotmailcom> wrote in message
news:42efcdf9$1@news.povray.org...
> Christoph Hormann wrote:
> > Anthony D. Baye wrote:
> >
> >>
> >> So my question is: Is there an eficient way to model Optical
> >> fibers, so that they not only glow, but transmit the light along the
> >> length as well.
> >> Naturally, I'll continue my own experimentations. I would however
> >> appreciate any thoughts people might have on the subject.
> >
> >
> > The transmission of light is nothing that makes sense to simulate in
> > POV-Ray since it has no effect on the image POV-Ray renders (light
> > itself isn't visible). Therefore using emitting media for the glow and
> > placing a spotlight at the end is the way to go.
> >
> > Christoph
> >
> Hey Christoph,
>
> Thanks for the input. The only problem that I see with your solution
> is that, if I were trying to model one of those fiber optic displays
> (f/ex. The flowers with the FO strands in them) I would have hundreds of
> spotlights in my scene. The light calculations would take forever.
How many strands are you trying to simulate? If you're talking more than just a
few, the number of calculations to properly simulate the total internal
reflection of fibre optics using photons would be *astronomical*.
It doesn't mean it can't be done, but it will take so much time that I suspect
it won't be worth doing.
High quality fibre optics don't glow along the entire fibre, either - if it was
to glow along the entire fibre it means that total internal reflection isn't
being achieved, and therefore the fibre is being *very* inefficient (it's losing
most of its light out the sides of the fibre, rather than out the end). The
light should only be escaping from the end of the fibre.
Lance.
thezone - thezone.firewave.com.au
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Anthony D. Baye nous apporta ses lumieres en ce 2005-08-02 22:46:
> Lonnie wrote:
>
> What's a swap file?
>
> I'm on a Mac. maybe it's called something else.
>
> A.D.B.
Virtual memory. A file on disk where the OS puts data when it runs short of physical
memory. It's
called swap because you swap chunks of data and programms between it and the RAM.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Anthony D. Baye" <Sha### [at] hotmailcom> wrote:
> Lonnie wrote:
> > This might actually be made to work with photons, although be prepared to
> > have your swap file limit increased to an astonomical size and a LONG
> > render time!
> >
> >
>
> What's a swap file?
>
> I'm on a Mac. maybe it's called something else.
>
> A.D.B.
A "swap file" or more correctly a page file, is what a Windows computer uses
when it runs out of RAM. Data is "swapped" in and out of memory to the
hard disk as needed. While it works OK for background tasks that only need
to access their data occasionally, things like a big photon map will slow a
system to a crawl. The CPU is spending far more time waiting on the hard
disk than it is crunching numbers. I'm not sure how a Mac does this, but
it has to be something similiar.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <web.42f89fab22ec88d03b3a698d0@news.povray.org>,
lon### [at] yahoocom says...
> "Anthony D. Baye" <Sha### [at] hotmailcom> wrote:
> > Lonnie wrote:
> > > This might actually be made to work with photons, although be prepared to
> > > have your swap file limit increased to an astonomical size and a LONG
> > > render time!
> > >
> > >
> >
> > What's a swap file?
> >
> > I'm on a Mac. maybe it's called something else.
> >
> > A.D.B.
>
> A "swap file" or more correctly a page file, is what a Windows computer uses
> when it runs out of RAM. Data is "swapped" in and out of memory to the
> hard disk as needed. While it works OK for background tasks that only need
> to access their data occasionally, things like a big photon map will slow a
> system to a crawl. The CPU is spending far more time waiting on the hard
> disk than it is crunching numbers. I'm not sure how a Mac does this, but
> it has to be something similiar.
>
Probably more efficiently though. Unless they fixed it in XP, the problem
was that it started paging once a certain percentage of memory was used
at one shot, even if you still had a significant amount left. Something
stupid like 1MB or something on the older versions of Windows. This is
what made games lag. If you moved around more than X megs in one shot,
the memory manager would start paging out, even if you still had 90% of
your physical RAM available. Or at least that is what I read anyway, when
someone described the problem.
--
void main () {
call functional_code()
else
call crash_windows();
}
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
