POV-Ray : Newsgroups : povray.general : glass cylinder on plane Server Time
15 Nov 2024 21:17:33 EST (-0500)
  glass cylinder on plane (Message 1 to 4 of 4)  
From: Micha Riser
Subject: glass cylinder on plane
Date: 29 Nov 1998 13:21:58
Message: <98112919275000.05226@micha>
I tried to make a glass cylinder standing on a plane.
If the cylinder's bottom is exactly on the plane there is the problem of
coincident surfaces. But what is the correct solution: to set the something
above plane or somthing in plane?

And how to solve it when water an glass (different iors) objects touches each
other?

Here's an example scene (of the first question):

---------

#include "colors.inc"
#include "textures.inc"
#include "glass.inc"

camera{
 location <0,.9,-2.5>
 look_at <0,0.6,0>
}

light_source{
 <3,10,-10>
 color White*2
 }

plane{y,0 pigment{White_Marble}}

#declare my_cyl=cylinder{<0,0,0><0,1.5,0>,.3
texture{T_Glass1}
interior{I_Glass}
}

object{my_cyl translate<-1,0.0001,0>}
object{my_cyl translate<0,0,0>}
object{my_cyl translate<1,-0.0001,0>}
background{Blue}

---------

 Micha

---------------------------------------------------
visit my homepage:
 http://www.geocities.com/SiliconValley/Pines/7992/
the POVRay Objects Collection:
 http://twysted.net/povobjects/
---------------------------------------------------


Post a reply to this message

From: Ken
Subject: Re: glass cylinder on plane
Date: 29 Nov 1998 13:45:57
Message: <366195D0.7758E1C1@pacbell.net>
Hi Micha,

   I have run into this problem on occasion and what I do is
for the plane problem I extend it slightly into the plane. This
helps get rid of multiple reflections between the bottom of
the cylinder and the surface of the plane.
   For water in a glass I always scale the water a little larger.
If you make it smaller you have an unnatural reflection situation
between the glass surface and the water surface. Just keep your
over scaling to a minimum amount - something like .0001 is
enough. I believe this is the recommended method somewhere
in the docs but I'm unsure which section. Probably under cgs
operations.

  Other people may handle these situations differently but
they have worked well for me.

Ken Tyler

Micha Riser wrote:

> I tried to make a glass cylinder standing on a plane.
> If the cylinder's bottom is exactly on the plane there is the problem of
> coincident surfaces. But what is the correct solution: to set the something
> above plane or somthing in plane?
>
> And how to solve it when water an glass (different iors) objects touches each
> other?


Post a reply to this message

From: Margus Ramst
Subject: Re: glass cylinder on plane
Date: 30 Nov 1998 08:47:17
Message: <3662A135.6835E40A@peak.edu.ee>
I have always followed the crude logic that such objects can either
touch or be or bit apart, but they can never intersect (in real life).
In the glass and water example it probably doesn't matter much (Ken, I
believe the water and glass would interreflect either way?)
I think the cylinder should not extend into the plane (esp. when the
plane is opaque and the cylinder is transparent but tinted) Otherwise
the part of the cylinder inside the plane is invisible and its colour is
not substracted. This looks unnatural.

Margus

Ken wrote:
> 
> Hi Micha,
> 
>    I have run into this problem on occasion and what I do is
> for the plane problem I extend it slightly into the plane. This
> helps get rid of multiple reflections between the bottom of
> the cylinder and the surface of the plane.
>    For water in a glass I always scale the water a little larger.
> If you make it smaller you have an unnatural reflection situation
> between the glass surface and the water surface. Just keep your
> over scaling to a minimum amount - something like .0001 is
> enough. I believe this is the recommended method somewhere
> in the docs but I'm unsure which section. Probably under cgs
> operations.
> 
>   Other people may handle these situations differently but
> they have worked well for me.
> 
> Ken Tyler


Post a reply to this message

From: Micha Riser
Subject: Re: glass cylinder on plane
Date: 30 Nov 1998 12:46:29
Message: <3662DBE5.322F30FF@usa.net>
Fact is that there is a clearly different result if you make the two
objects overlap (should I use merge to do that or not?) or if you have a
little space between the objects.

Maybe we've to have a look the reality, which of the two solutions comes
closer to nature...


Margus Ramst wrote:
> 
> I have always followed the crude logic that such objects can either
> touch or be or bit apart, but they can never intersect (in real life).
> In the glass and water example it probably doesn't matter much (Ken, I
> believe the water and glass would interreflect either way?)
> I think the cylinder should not extend into the plane (esp. when the
> plane is opaque and the cylinder is transparent but tinted) Otherwise
> the part of the cylinder inside the plane is invisible and its colour is
> not substracted. This looks unnatural.
> 
> Margus
> 
> Ken wrote:
> >
> > Hi Micha,
> >
> >    I have run into this problem on occasion and what I do is
> > for the plane problem I extend it slightly into the plane. This
> > helps get rid of multiple reflections between the bottom of
> > the cylinder and the surface of the plane.
> >    For water in a glass I always scale the water a little larger.
> > If you make it smaller you have an unnatural reflection situation
> > between the glass surface and the water surface. Just keep your
> > over scaling to a minimum amount - something like .0001 is
> > enough. I believe this is the recommended method somewhere
> > in the docs but I'm unsure which section. Probably under cgs
> > operations.
> >
> >   Other people may handle these situations differently but
> > they have worked well for me.
> >
> > Ken Tyler

-- 
---------------------------------------------------
visit my homepage:
 http://www.geocities.com/SiliconValley/Pines/7992/
the POVRay Objects Collection:
 http://twysted.net/povobjects/
---------------------------------------------------


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.