Ken <tyl### [at] pacbellnet> wrote:
> You probably have a "plane" object in your scene

  A really VFAQ? :)

  One might think *why* does a plane need to be defined as solid (ie.
having an interior, which is actually half of the universe). One
naturally thinks a plane is just an infinitely thin sheet with no
volume.
  If one thinks this way, one forgets that planes are quite handy in
CSG (eg. for cutting parts of an object out) and would be quite useless
for this handy purpose if it was a non-solid object (like a bicubic
patch is).

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

Ken wrote:
> If you are not using fog or media ignore the warning. You probably have
> a "plane" object in your scene and your camera is inside the normal of
> the plane. It could also be that you are using a sphere to mimic the sky
> which would also lead to that warning. In either case you can add the
> "inverse" keyword to the plane or sphere and that will get rid of that
> warning.

"hollow" should work too.

Post a reply to this message

Philippe Debar wrote:
> 
> Ken wrote:
> > If you are not using fog or media ignore the warning. You probably have
> > a "plane" object in your scene and your camera is inside the normal of
> > the plane. It could also be that you are using a sphere to mimic the sky
> > which would also lead to that warning. In either case you can add the
> > "inverse" keyword to the plane or sphere and that will get rid of that
> > warning.
> 
> "hollow" should work too.

It does, but it also confuses the issue of surface normals so I chose not
to use it in my example.

-- 
Ken Tyler

Post a reply to this message

Philippe Debar <phd### [at] wanadoobe> wrote:
> "hollow" should work too.

  IMHO it's not the best solution.
  You should understand how plane works and it should be completely crystal
clear which side of the plane you are defining is inside and which is
outside. "Fixing" the warning by putting 'hollow' in the plane definition
is more like hiding the problem, not fixing it. It can teach you wrong
principles. You should have a good, real reason to use 'hollow', and IMHO
getting rid of the warning is not one.
  Defining the plane so that the scene is located *outside* it is the best
solution. It helps you visualizing better what's happening.

  A quick recapitulation: The first parameter of 'plane' defines its normal
vector. The normal vector points outside the plane.
  That is, if you for example define a 'plane { z,0 }', then everything
which is located anywhere with a positive z coordinate is outside the plane,
while things located at a negative z coordinate are inside the plane.
'plane { -z,0 }' inverts this.
  The second parameter tells an offset. The plane is translated by this
amount along the normal vector (ie. in the direction of the first parameter).
Usually you want to move the plane backwards, ie. in the opposite direction
of the normal vector (very common when defining floors and walls), which
naturally means that this parameter should be negative.
  Thus, eg. a floor is usually defined like 'plane { y, -5 }', which creates
a plane at y=-5 so that everything above it is outside.

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

"Ken" <tyl### [at] pacbellnet> wrote in message
news:3E23DCD8.B37F4AFB@pacbell.net...
> > "hollow" should work too.
>
> It does, but it also confuses the issue of surface normals so I chose not
> to use it in my example.

Yes, the "hollow" keyword is confusing. But I believed the normals were
unaffected. Aren't they? What I understood is that the object still has a
clearly determined outside and inside - hence you can use interior{}and csg,
but you need to hollow the object to fill it with media.

Is there any planned change for that behaviour?

I do not understand why you need "hollow" to use interior{media} and not
(iirc) interior{ior} (And I really mean that I do not understand, not
anything else - there are probably technical reasons, I just do not know
them) . Is there any case where there is an advantage in using an
interior{media} in a non hollow object? BTW I'd be happy to be able to have
an object inside a media to be filled either by it's own media only or with
it's and the surrounding medias without having to use tricky csg.


Povingly,


Philippe

Post a reply to this message

Stepping in on this little discussion...

"Philippe Debar" <phd### [at] wanadoobe> wrote in message
news:3e24ff65@news.povray.org...
>
> "Ken" <tyl### [at] pacbellnet> wrote in message
> news:3E23DCD8.B37F4AFB@pacbell.net...
> > > "hollow" should work too.
> >
> > It does, but it also confuses the issue of surface normals so I chose
not
> > to use it in my example.
>
> Yes, the "hollow" keyword is confusing. But I believed the normals were
> unaffected. Aren't they? What I understood is that the object still has a
> clearly determined outside and inside - hence you can use interior{}and
csg,
> but you need to hollow the object to fill it with media.

Right.

> Is there any planned change for that behaviour?

Not that I know of, except I'm sure this general subject must have been
talked about in the past. Something of a hierarchy for combining or
separating medias. Those two words would label it well I think, combine or
separate. I haven't looked at the media source code but it doesn't sound
that easy to change. It would mean per object media mixing as is already
done within interiors having more than one media statement present, so it
would need to check all other object interiors.

> I do not understand why you need "hollow" to use interior{media} and not
> (iirc) interior{ior} (And I really mean that I do not understand, not
> anything else - there are probably technical reasons, I just do not know
> them) . Is there any case where there is an advantage in using an
> interior{media} in a non hollow object? BTW I'd be happy to be able to
have
> an object inside a media to be filled either by it's own media only or
with
> it's and the surrounding medias without having to use tricky csg.

It's really a pre-media thing, when 'atmosphere' came along it applied to
that as well; and finally media too, since it could also be global. I
couldn't tell you the exact reasoning but that's probably why 'hollow' is
the defacto way to produce a surface shell rather than the usual solid
object.

As to what you asked there, "Is there any case where there is an advantage
in using an interior{media} in a non-hollow object?", I guess you are saying
again that hollow is redundant if media exists at all on a per object basis.
And you answer your own question in the sense that it would be good to mix
and match medias with more control, and it's CSG or global media that
becomes tricky as it stands now.

--
Farewell,
Bob

Post a reply to this message

Philippe Debar <phd### [at] wanadoobe> wrote:
> Is there any planned change for that behaviour?

  Nothing have been planned, but there has been some discussion about the
issue.

  My opinion is that the keyword 'hollow' should be completely removed.
Even though it's a logical name *in its context*, it's *way* too confusing
for new users, and even for some more advanced ones.

  The best replacement keyword I can think of is 'allow_media', optionally
followed by a boolean value.
  (Also the warning would become more clear: "Camera is inside an object
which does not allow media inside it... etc etc".

  Also its behaviour could be changed a bit:
  For objects with no media block defined its default value would be 'off',
while for objects with media block defined the default value would be 'on'.

  There have been other opinions about this behavioral thing as well (but
these require more drastical changes to the code).
  One good idea is that an object with media will always show its own media,
but whether it will show external media (which includes media of other
objects, global media and fog) or not would depend on the 'allow_media'
value.
  This way an object could have its own media, but still be "solid" (in the
sense that other media will not penetrate it).

-- 
plane{-x+y,-1pigment{bozo color_map{[0rgb x][1rgb x+y]}turbulence 1}}
sphere{0,2pigment{rgbt 1}interior{media{emission 1density{spherical
density_map{[0rgb 0][.5rgb<1,.5>][1rgb 1]}turbulence.9}}}scale
<1,1,3>hollow}text{ttf"timrom""Warp".1,0translate<-1,-.1,2>}//  - Warp -

Post a reply to this message

"Warp" <war### [at] tagpovrayorg> wrote in message
news:3e25696c@news.povray.org...
>   One good idea is that an object with media will always show its own
media,
> but whether it will show external media (which includes media of other
> objects, global media and fog) or not would depend on the 'allow_media'
> value.
>   This way an object could have its own media, but still be "solid" (in
the
> sense that other media will not penetrate it).

Except that then allow_media would be confused since a media might reside in
the object already. I think something more like combine_media on|off might
make sense to everyone. Would be interesting to see how it would affect
things.

Post a reply to this message

I got a similar situation and would like your advice:  I have an image_map
applied to a plane, of which I'd like to be able to view from either side.
Now even if I use the inverse, one of the sides will be the "inside" giving
the error message.  I tried adding: bounded_by { box { <0,0,0> <1,1,0> } },
but PovRAY gives the message this is unnecessary and removes it.  Ok, I
agree it should be unnecessary, but I was hoping it'd kill off the "inside"
aspect of the plane.

--Eric

Ken <tyl### [at] pacbellnet> wrote:
> "DaveS
> >
> > ...fog & participating media may not work as expected.
> >
> > Does anyone know why I get this error message? The camera is not inside
> > anything and I can't figure out whats wrong.
>
> If you are not using fog or media ignore the warning. You probably have
> a "plane" object in your scene and your camera is inside the normal of
> the plane. It could also be that you are using a sphere to mimic the sky
> which would also lead to that warning. In either case you can add the
> "inverse" keyword to the plane or sphere and that will get rid of that
> warning.
>
> --
> Ken Tyler

Post a reply to this message

eric914 wrote:
> I got a similar situation and would like your advice:  I have an image_map
> applied to a plane, of which I'd like to be able to view from either side.
> Now even if I use the inverse, one of the sides will be the "inside" giving
> the error message.  I tried adding: bounded_by { box { <0,0,0> <1,1,0> } },
> but PovRAY gives the message this is unnecessary and removes it.  Ok, I
> agree it should be unnecessary, but I was hoping it'd kill off the "inside"
> aspect of the plane.

1) Reverse the surface normal of the plane.
2) If you truely want it to allow media on either side of the surface, 
add hollow to the definition of the plane.

-- 
~Mike

Things! Billions of them!

Post a reply to this message