"Bill Pragnell" <bil### [at] hotmailcom> wrote:
> Bit quiet on p.b.i. at the moment, so here's some shiny things.
>

Very, very beautiful! Especially the lighting and the colored shadows. I tried
something similar with a simple glass sphere and photons a few weeks ago, but
the shadow was much too dark. Maybe the "count" was too low. I'm not so
experienced with glass like objects, because I have frustrating results most of
the time and therefore don't really dare to raytrace them often ...
But now with this proof, that they can look really good, I will try again
sometime in the future.

> These are the Platonic, Archimedean and Catalan solids. How many can you name
> without resorting to wikipedia or wolfram? :-)

.... dodecahedron, icosahedron ... uhm ...

Post a reply to this message

On Wed, 24 Jun 2009 16:45:35 -0400, Bill Pragnell wrote:

> Bit quiet on p.b.i. at the moment, so here's some shiny things.

Shiny!  I like the textures & colours.... :-)

Reminds me a lot of the polyhedral dice I have around ehre somewhere.

Jim

Post a reply to this message

"Simone" <inf### [at] alienenterprisesde> wrote:
> Very, very beautiful! Especially the lighting and the colored shadows.

Thank you!

> I tried
> something similar with a simple glass sphere and photons a few weeks ago, but
> the shadow was much too dark. Maybe the "count" was too low.

The shadows are bright in my image because I used radiosity with a bright
environment, so there are no completely shadowed areas. However, if you have a
sphere and a single light source, a largely dark shadow with a single bright
spot is to be expected. It only looks odd because you don't often get that
lighting setup in reality.

Having said that, count does need to be fairly high to get good results. I
typically use at least 1e6, but 1e4-1e5 could be sufficient for something
really simple like a sphere.

> I'm not so
> experienced with glass like objects, because I have frustrating results most of
> the time and therefore don't really dare to raytrace them often ...
> But now with this proof, that they can look really good, I will try again
> sometime in the future.

Sometimes, the knowledge that something can be done is all the motivation you
need to work out how to do it. :)

Bill

Post a reply to this message

> "Bill Pragnell" <bil### [at] hotmailcom> wrote:
>> Bit quiet on p.b.i. at the moment, so here's some shiny things.
>>
> 
> Very, very beautiful! Especially the lighting and the colored shadows. I tried
> something similar with a simple glass sphere and photons a few weeks ago, but
> the shadow was much too dark. Maybe the "count" was too low. I'm not so
> experienced with glass like objects, because I have frustrating results most of
> the time and therefore don't really dare to raytrace them often ...
> But now with this proof, that they can look really good, I will try again
> sometime in the future.
> 
>> These are the Platonic, Archimedean and Catalan solids. How many can you name
>> without resorting to wikipedia or wolfram? :-)
> 
> .... dodecahedron, icosahedron ... uhm ...
> 
> 
> 
When tracing a sphere with photons, it's normal to get a (very) dark 
shadow with a bright spot in the center. If the sphere is to far away 
from the surface receiving the shadow, the photons will have converged 
to the focal point and spread again, possibly to an area much larger 
than the sphere.

Glass objects, as well as metallic ones, need some environment to look 
good. You also need something to get some light in the shadowed areas. 
It can be a shadowless light source set to NOT cast photons, or it can 
be radiosity.

Photons block for a fill light:
photons{refraction off reflection off}


Alain

Post a reply to this message

Alain wrote:
> Glass objects, as well as metallic ones, need some environment to look
> good. You also need something to get some light in the shadowed areas.
> It can be a shadowless light source set to NOT cast photons, or it can
> be radiosity.

  Or simply use multiple light sources. A single light source,
especially in "interior" scenes, tends to look a bit dull (although it
can be used effectively as well).

Post a reply to this message

Very beautiful, indeed! And surely it did take a lot of work to code the 
shapes. Are they meshes or CSG?

Did you consider posting the polyherda to the object-library? I think a lot 
of people would like to play with these beautiful shapes... me for one ;-)

Can't help it, I am something of a magpie - I love glittering things...

Post a reply to this message

"Bill Pragnell" <bil### [at] hotmailcom> wrote:
> "Simone" <inf### [at] alienenterprisesde> wrote:

> The shadows are bright in my image because I used radiosity with a bright
> environment, so there are no completely shadowed areas. However, if you have a
> sphere and a single light source, a largely dark shadow with a single bright
> spot is to be expected. It only looks odd because you don't often get that
> lighting setup in reality.

Yes, I think I'll have to try with radiosity right from the start. First I
thought I would try it with only a single light and without radiosity and when
it looks halfway decent, to enhance it with radiosity and another light source.
Because having radiosity right from the beginning is always so time consuming
that it doesn't work well with my trial and error approach :))

Alain <aze### [at] qwertyorg> wrote:
> Glass objects, as well as metallic ones, need some environment to look
> good.
Yes, I already had a plane and skysphere. I also tried to put some other
transparent objects around and set them to no_image. It does lighten up the
shadows of the still visible sphere, but at the same time the sphere's shadow
starts too look dull and not colorful. And then when there is supposed to be
only one sphere in the scene, reflections from other objects on the spere's
surface look a little weird, too. On the other hand, if there are no such
reflections, it looks too sterile. Simple things seem to me sometimes the most
difficult.

However, thanks for the various advice. If I'll manage to come up with a decent
result I'll post it here :)

Post a reply to this message

"TC" <do-not-reply@i-do get-enough-spam-already-2498.com> wrote:
> Very beautiful, indeed! And surely it did take a lot of work to code the
> shapes. Are they meshes or CSG?

They are meshes, and can be used in CSG, but I didn't generate the data myself.
The vertex/face data comes from here:

http://www.georgehart.com/virtual-polyhedra/vp.html

It's all in VRML format, but I wrote a tool to convert the files to POV-Ray .inc
files, as a set of macros (with various texture options) as well as a global
array holding the vertices. The only really hard work I've been doing is
tweaking the macros to orient the polyhedra pleasingly.

> Did you consider posting the polyherda to the object-library? I think a lot
> of people would like to play with these beautiful shapes... me for one ;-)

I probably need to contact the author of the above site first to make sure he's
happy with me swiping his vertices (naturally he'd be credited in all the .inc
files), but that was my eventual intention.

> Can't help it, I am something of a magpie - I love glittering things...

That's what raytracing was invented for! :-)

Post a reply to this message

> "Bill Pragnell" <bil### [at] hotmailcom> wrote:
>> "Simone" <inf### [at] alienenterprisesde> wrote:
> 
>> The shadows are bright in my image because I used radiosity with a bright
>> environment, so there are no completely shadowed areas. However, if you have a
>> sphere and a single light source, a largely dark shadow with a single bright
>> spot is to be expected. It only looks odd because you don't often get that
>> lighting setup in reality.
> 
> Yes, I think I'll have to try with radiosity right from the start. First I
> thought I would try it with only a single light and without radiosity and when
> it looks halfway decent, to enhance it with radiosity and another light source.
> Because having radiosity right from the beginning is always so time consuming
> that it doesn't work well with my trial and error approach :))
> 
> Alain <aze### [at] qwertyorg> wrote:
>> Glass objects, as well as metallic ones, need some environment to look
>> good.
> Yes, I already had a plane and skysphere. I also tried to put some other
> transparent objects around and set them to no_image. It does lighten up the
> shadows of the still visible sphere, but at the same time the sphere's shadow
> starts too look dull and not colorful. And then when there is supposed to be
> only one sphere in the scene, reflections from other objects on the spere's
> surface look a little weird, too. On the other hand, if there are no such
> reflections, it looks too sterile. Simple things seem to me sometimes the most
> difficult.
> 
> However, thanks for the various advice. If I'll manage to come up with a decent
> result I'll post it here :)
> 
> 


During the initial stage, use a shadowless fill light. It's brightness 
should be kept low, normaly at less than 0.2, around 0.1.

A second light is much faster than using radiosity, and a shadowless 
light is faster than a regular one because you don't do any shadow test 
for those.

Alain

Post a reply to this message

This is a very beautiful stil image. Well-done!

Post a reply to this message