Strange problem using -vtl to cut sections

Westling_Robert · July 11, 2002, 6:14pm

Hi,

I am cutting plan and sectional views through a model for radiance
images, illuminance images using falsecolor, and point data output using
rtrace and have found the following problem: light is showing up inside
of fully closed boxes. For instance in plan view using -vtl and -o to
get a plan cut 4' off the floor I find that light appears to be
propagating inside of volumes with no holes. Running rtrace in these
areas also shows that a small amount of light is bouncing around inside
these solid shapes.

The solution I plan on using is to model a flat plane at the bottom of
each of these spaces and give it a completely black plastic finish.
However it is a little disconcerting that light is entering the inside
of these boxes.

Any ideas on the subject would be greatly appreciated.

Thank you,

Robert Westling
steinerwestling

Carsten_Bauer2 · July 11, 2002, 7:48pm

Hi Robert

when reading your posting I just got a spontaneous idea - maybe the
light stems from the constant ambient term (-av ...) which gets
activated on every rayhit. Although your boxes are closed, de facto
they're open because of the cutoff-plane, and rays hit inside and will
get assigned the constant ambient term (in the same manner as a scene
with no sources present). Maybe, maybe ..

Carsten

Andrew_Bissell · July 12, 2002, 12:46pm

If you set -av at 0 0 0 do you still get measurable light levels in these
closed rooms.

Regards

Andrew

···

-----Original Message-----
From: [email protected] [SMTP:cbauer-@t-online.de]
Sent: 11 July 2002 20:49
To: [email protected]
Subject: Re: [Radiance-general] Strange problem using -vtl to cut
sections

Hi Robert

when reading your posting I just got a spontaneous idea - maybe the
light stems from the constant ambient term (-av ...) which gets
activated on every rayhit. Although your boxes are closed, de facto
they're open because of the cutoff-plane, and rays hit inside and will
get assigned the constant ambient term (in the same manner as a scene
with no sources present). Maybe, maybe ..

Carsten
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

Westling_Robert · July 12, 2002, 12:49pm

Hi,

I found the error causing the bounces within solids.

The model had been created in formZ with the different materials on
different layers, then exported as separate rad files. Normally not a
problem, but as a result of this modeling technique at the bottom of
solid shapes where they are sitting on another shape, like the floor of
a room, the bottom of the solid was coincidental with the top surface of
the other shape. Radiance sees the shape below and is sampleing that
surface into the base of the solid shape. By moving the bottom face of
the solid off the surface of the floor, the problem went away and no
light is propagating within the solid. The next step will be to cut
away the floor at these intersections so that Radiance cannot 'see' into
the base of a solid shape.

Robert Westling
steinerwestling

···

--__--__--

Message: 1
Date: Thu, 11 Jul 2002 14:14:32 -0400
From: "Westling, Robert" <[email protected]>
To: <[email protected]>
Subject: [Radiance-general] Strange problem using -vtl to cut sections
Reply-To: [email protected]

This is a multi-part message in MIME format.

------_=_NextPart_001_01C22906.CE315366
Content-Type: text/plain;
charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi,
=20
I am cutting plan and sectional views through a model for radiance
images, illuminance images using falsecolor, and point data output using
rtrace and have found the following problem: light is showing up inside
of fully closed boxes. For instance in plan view using -vtl and -o to
get a plan cut 4' off the floor I find that light appears to be
propagating inside of volumes with no holes. Running rtrace in these
areas also shows that a small amount of light is bouncing around inside
these solid shapes. =20
=20
The solution I plan on using is to model a flat plane at the bottom of
each of these spaces and give it a completely black plastic finish.
However it is a little disconcerting that light is entering the inside
of these boxes.
=20
Any ideas on the subject would be greatly appreciated.
=20
Thank you,
=20
Robert Westling
steinerwestling

------_=_NextPart_001_01C22906.CE315366
Content-Type: text/html;
charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<head>
<meta http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<meta name=3DProgId content=3DWord.Document>
<meta name=3DGenerator content=3D"Microsoft Word 10">
<meta name=3DOriginator content=3D"Microsoft Word 10">
<link rel=3DFile-List href=3D"cid:[email protected]">

<style>

</style>

</head>

Hi,<o:p></o:p>

<o:p> </o:p>

I am cutting plan and sectional views through a model

for radiance
images, illuminance images using falsecolor,
and point data output using rtrace and have

found the
following problem:  light =
is showing
up inside of fully closed boxes. 
For instance in plan view using –vtl and
–o to get a plan cut 4’ off the floor I find that light =
appears to
be propagating inside of volumes with no holes.  Running rtrace
in these areas also shows that a small amount of light is bouncing =
around
inside these solid shapes.  =
<o:p></o:p>

<o:p> </o:p>

The solution I plan on using is to model a flat plane

at the
bottom of each of these spaces and give it a completely black plastic
finish.  However it is a =
little
disconcerting that light is entering the inside of these =
boxes.<o:p></o:p>

<o:p> </o:p>

Any ideas on the subject would be greatly =
appreciated.<o:p></o:p>

<o:p> </o:p>

Thank you,<o:p></o:p>

<o:p> </o:p>

Robert Westling<o:p></o:p>

steinerwestling<o:p></o:p>

</div>

</body>

</html>
=00
------_=_NextPart_001_01C22906.CE315366--

--__--__--

Message: 2
Date: Thu, 11 Jul 2002 21:48:58 +0200
From: [email protected] (Carsten Bauer)
To: [email protected]
Subject: Re: [Radiance-general] Strange problem using -vtl to cut
sections
Reply-To: [email protected]

Hi Robert

when reading your posting I just got a spontaneous idea - maybe the
light stems from the constant ambient term (-av ...) which gets
activated on every rayhit. Although your boxes are closed, de facto
they're open because of the cutoff-plane, and rays hit inside and will
get assigned the constant ambient term (in the same manner as a scene
with no sources present). Maybe, maybe ..

Carsten

--__--__--

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

End of Radiance-general Digest

Carsten_Bauer2 · July 12, 2002, 3:03pm

Hi,

hmm, somehow this sounds strange ...

Carsten

Greg_Ward1 · July 12, 2002, 6:16pm

From: "Westling, Robert" <[email protected]>
To: <[email protected]>
Subject: [Radiance-general] Strange problem using -vtl to cut sections

Hi,

I am cutting plan and sectional views through a model for radiance
images, illuminance images using falsecolor, and point data output using
rtrace and have found the following problem: light is showing up inside
of fully closed boxes. For instance in plan view using -vtl and -o to
get a plan cut 4' off the floor I find that light appears to be
propagating inside of volumes with no holes. Running rtrace in these
areas also shows that a small amount of light is bouncing around inside
these solid shapes.

The solution I plan on using is to model a flat plane at the bottom of
each of these spaces and give it a completely black plastic finish.
However it is a little disconcerting that light is entering the inside
of these boxes.

Any ideas on the subject would be greatly appreciated.

Thank you,

Robert Westling
steinerwestling

This is disturbing, but nothing really to worry about. In most lighting calculations, there will be some small amount of light leakage, especially across infinitely thin boundaries. And, as Carsten points out, the ambient term will give you light even where there is none. If you want to avoid all such leakage problems, you would need to set -ar 0 and -av 0 0 0. However, the image would take longer to compute, and setting -av 0 0 0 rather than to a good guess of the ambient level results in a loss of accuracy over most of the scene.

(There is a script in the 3.4 release called compamb, which computes an ambient value for a rad input file automatically.)

-Greg

Nicholas_Lander · July 11, 2002, 11:31pm

Hi Robert,

Sometimes, depending on the method used to build the model, you can get
light leakage where planes meet: very small floating point errors can cause
otherwise unnoticeable gaps, especially when the number of nodes (defined
points) on one plane edge doesn't equal the number of nodes on the adjoining
edge. (Although this may just be a problem with the geometry builder I use.)

Nick L.

Strange problem using -vtl to cut sections

<p class=3DMsoNormal><font size=3D2 face=3DArial><span =
style=3D'font-size:10.0pt;
font-family:Arial'>I am cutting plan and sectional views through a model

for radiance
images, <span class=3DSpellE>illuminance</span> images using <span =
class=3DSpellE>falsecolor</span>,
and point data output using <span class=3DSpellE>rtrace</span> and have

<p class=3DMsoNormal><font size=3D2 face=3DArial><span =
style=3D'font-size:10.0pt;
font-family:Arial'>The solution I plan on using is to model a flat plane

Strange problem using -vtl to cut sections

<p class=3DMsoNormal><font size=3D2 face=3DArial><span = style=3D'font-size:10.0pt; font-family:Arial'>I am cutting plan and sectional views through a model

for radiance images, <span class=3DSpellE>illuminance</span> images using <span = class=3DSpellE>falsecolor</span>, and point data output using <span class=3DSpellE>rtrace</span> and have

<p class=3DMsoNormal><font size=3D2 face=3DArial><span = style=3D'font-size:10.0pt; font-family:Arial'>The solution I plan on using is to model a flat plane

<p class=3DMsoNormal><font size=3D2 face=3DArial><span =
style=3D'font-size:10.0pt;
font-family:Arial'>I am cutting plan and sectional views through a model

for radiance
images, <span class=3DSpellE>illuminance</span> images using <span =
class=3DSpellE>falsecolor</span>,
and point data output using <span class=3DSpellE>rtrace</span> and have

<p class=3DMsoNormal><font size=3D2 face=3DArial><span =
style=3D'font-size:10.0pt;
font-family:Arial'>The solution I plan on using is to model a flat plane