Warp wrote:
>   And what exactly what prohibit you from using a sphere right next to
> the mesh if radiosity was used?

The image will look horrible if each object is rendered with different
lighting.

Would the sphere reflect diffuse light to the mesh?

Post a reply to this message

Nicolas Alvarez <nic### [at] gmailcom> wrote:
> The image will look horrible if each object is rendered with different
> lighting.

  If you want the sphere to have the same lighting as the mesh, then use
a mesh instead of a sphere.

  It's not like this would be the only primitive-specific feature in POV-Ray.
Not being able to implement some feature for all primitives has never stopped
anyone from implementing them for specific primitives anyways.

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
>   What do you mean "cannot be used"? Of course it can be used even if your
> scene contains something else than meshes. It's just that the radiosity
> lighting is limited to appear on the meshes. The other primitive will have
> to do with regular lighting.

.... and that will probably look crappy, unless you're actually aiming at this
type of effect.

Post a reply to this message

> -----Original Message-----
> From: Warp [mailto:war### [at] tagpovrayorg]
> Posted At: Sunday, December 28, 2008 10:37 AM
> Posted To: povray.beta-test
> Conversation: Radiosity: status & SMP idea
> Subject: Re: Radiosity: status & SMP idea
> 
> Nicolas Alvarez <nic### [at] gmailcom> wrote:
> > The image will look horrible if each object is rendered with
> different
> > lighting.
> 
>   If you want the sphere to have the same lighting as the mesh, then
> use
> a mesh instead of a sphere.

Warp, that statement makes my point.

I think we're operating under different assumptions here.  I'm under the
impression that a scene would look best when the entire scene uses the
same lighting model.

You're saying that different parts of the scene can use different
lighting models.

While technically you are correct, the results do *not* look good.
Remember those 80s movies, where you would have a scene shot with one
light, and an actor filmed separately in different lighting conditions,
and the actor was superimposed onto the scene?  Remember how fake and
horrible it looked?

Now, if you're *trying* to make a scene that looks fake and horrible,
then by all means go ahead.  Most of us here don't want that.

And as far as triangle meshes go, I think it would be a good thing for
POV to support them (it currently doesn't, you know, unless you run it
through a converter anyway), BUT they should fit in the scene... meaning
they need the same texturing & lighting models that the rest of POV
uses.

...Ben Chambers
www.pacificwebguy.com

A render isn't slow unless it won't finish until after your next
birthday.

Post a reply to this message

BTW, it's the first time I see Warp defending polygon meshes, rather than his
preferred perfectly mathematical primitives, parametric and iso surfaces.  And
he get scorned for it.  Poor unlucky fellow... :P

Post a reply to this message

Warp nous illumina en ce 2008-12-28 13:37 -->
> Nicolas Alvarez <nic### [at] gmailcom> wrote:
>> The image will look horrible if each object is rendered with different
>> lighting.
> 
>   If you want the sphere to have the same lighting as the mesh, then use
> a mesh instead of a sphere.
> 
>   It's not like this would be the only primitive-specific feature in POV-Ray.
> Not being able to implement some feature for all primitives has never stopped
> anyone from implementing them for specific primitives anyways.
> 
Whatever lighting model is used, it ABSOLUTELY MUST apply to any element of any 
scene. It just can't apply to only one kind of object. Any POV-Ray scene can 
contain simple primitives, isosurfaces and meshes. Any lighting model that don't 
work with any one of those just can't be used at all.

-- 
Alain
-------------------------------------------------
You know you've been raytracing too long when your ophthalmologist examines you 
for complaints of eye strain and blurred vision and asks you why the words Pov, 
#declare, #include, sphere, translate, rotate, texture, and pigment are 
permanently burned into your retina.
Ken Tyler

Post a reply to this message

Chambers <ben### [at] pacificwebguycom> wrote:
> >   If you want the sphere to have the same lighting as the mesh, then
> > use
> > a mesh instead of a sphere.

> Warp, that statement makes my point.

  Which I disagree with. *My* point is that a lighting model which applies
only to meshes can be useful given how common mesh-based scenes are
(especially when importing from other modellers/renderers to POV-Ray).

  The counter-argument seems to be that since the algorithm can only be
applied to meshes, it should not be provided *at all* (even though providing
it wouldn't hurt anyone). I find this reasoning incomprehensible. In the
exact same way you could argue that UV-mapping should not be provided
because it cannot be applied to all primitives. It makes the exact same
amount of sense.

  If you don't like UV-mapping because it cannot be applied to everything
then the solution is rather simple: Don't use UV-mapping. Likewise if you
don't like radiosity because it can only be applied to meshes, the same
solution applies: Don't use it. Nobody gets hurt.

  However, depriving people who *could* find it useful from the tool
doesn't make sense.

> I think we're operating under different assumptions here.  I'm under the
> impression that a scene would look best when the entire scene uses the
> same lighting model.

  Have you ever heard of light groups?

  The entire scene being lighted in the exact same way is in no way a
prerequisite. On the contrary, sometimes you even *want* some parts not
being illuminated in the same way as the others. That's why light groups
were introduced in the first place.

  Not that this has anything to do with radiosity+meshes. I just wanted
to point out that your argument that a lighting model *must* apply to all
possible primitives or else it's useless does not necessarily hold true.

> You're saying that different parts of the scene can use different
> lighting models.

  Of course they can. I don't see the problem. Light groups already provide
a way of achieving that.

> While technically you are correct, the results do *not* look good.

  Then don't use the proposed radiosity (assuming it was actually
implemented). It's that simple.

> Now, if you're *trying* to make a scene that looks fake and horrible,
> then by all means go ahead.  Most of us here don't want that.

  Another alternative is to make the entire scene with triangle meshes
and have superb, fast and viewpoint-independent global illumination.
You want to deprive people from having that?

> And as far as triangle meshes go, I think it would be a good thing for
> POV to support them (it currently doesn't, you know, unless you run it
> through a converter anyway)

  I think you are trying to say something else than what you wrote there,
but I'm not sure what.

-- 
                                                          - Warp

Post a reply to this message

nemesis <nam### [at] gmailcom> wrote:
> BTW, it's the first time I see Warp defending polygon meshes, rather than his
> preferred perfectly mathematical primitives, parametric and iso surfaces.  And
> he get scorned for it.  Poor unlucky fellow... :P

  Well, it's one thing to appreciate the work put into a scene made of
varied POV-Ray primitives, and another thing to admit the reality of the
computer graphics world: That 99.9% of scenes out there are made of
triangle meshes (or of easily tesselable primitives), and that most
rendering algorithms have been optimized for them.

-- 
                                                          - Warp

Post a reply to this message

Alain <ele### [at] netscapenet> wrote:
> Whatever lighting model is used, it ABSOLUTELY MUST apply to any element of any 
> scene.

  Says who? That's like saying that whatever texturing elements are used,
they ABSOLUTELY MUST apply to any element of any scene. Clearly UV-mapping
doesn't do this, yet nobody is complaining.

  I see no problem in triangle meshes supporting lightmaps, and these
lightmaps being pre-calculated using the algorithm called "radiosity".
(In fact, lightmaps are not very far from UV-mapping. It's just an
additional layer on top of the texture layer.)

> It just can't apply to only one kind of object. Any POV-Ray scene can 
> contain simple primitives, isosurfaces and meshes. Any lighting model that don't 
> work with any one of those just can't be used at all.

  And why, exactly? I see no technical, ideological nor philosophical
problem with it.

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> > And as far as triangle meshes go, I think it would be a good thing for
> > POV to support them (it currently doesn't, you know, unless you run it
> > through a converter anyway)
>
>   I think you are trying to say something else than what you wrote there,
> but I'm not sure what.

I guess he was talking about .obj format files or something similar.

Post a reply to this message