Hi all,
I am trying to generate a tensor tree BSDF for the 5-phase simulation, but
I get an error.
Here is what I type:
genBSDF +f +b -geom meter -dim 1 7 0.8 1.8 -.2 0 -t4 5 bsdf/windows-v80.rad
bsdf/windows-v80_t45.xml
And it returns this error repeatedly:
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
.
.
.
.
Does anyone know what might be wrong?
Attached is my windows-v80.rad file.
Best,
Leyla
windows-v80.rad (52.8 KB)
The tensor-tree code in genBSDF hasn't yet been ported to Windows. So, if you're under Windows, that is your problem.
-Greg
windows-v80.rad (52.8 KB)
···
From: Leyla Sanati <[email protected]>
Date: September 6, 2013 2:58:02 PM PDT
Hi all,
I am trying to generate a tensor tree BSDF for the 5-phase simulation, but I get an error.
Here is what I type:
genBSDF +f +b -geom meter -dim 1 7 0.8 1.8 -.2 0 -t4 5 bsdf/windows-v80.rad > bsdf/windows-v80_t45.xml
And it returns this error repeatedly:
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
.
.
.
.
Does anyone know what might be wrong?
Attached is my windows-v80.rad file.
Best,
Leyla
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
Hi Greg,
Thank you for your prompt reply.
I am doing my simulations on my mac, and I have the latest version of
Radiance.
If you may remember from my presentation, I modeled a room with four
openings on the south wall: two clerestories, and two windows.
Similar to my 3-phase simulation, I am trying to generate separate BSDFs
for windows and clerestories, as they have different and dynamic shading
systems. The room is positioned 3 meters above the ground (on second floor).
For the windows, I generated venetian blinds using genblind. I used -a
option in xform to duplicate the venetian blinds. Then I used xform to
position the South wall, the windows and venetian blinds on xy plane with
the interior wall facing up at z=0. In genBSDF script, I set -dim command
to cover the two windows and the wall between them. Since I am new to
genBSDF, I am not sure what is causing the error. Does the clerestory
openings on the south wall need to be filled when creating the window BSDFs?
#objects/venetain80.rad
void plastic blindslat
0
0
5 0.7 0.7 0.7 0.1 0
!genblinds blindslat blinds .02 2 1 83 80 +r .2 | xform -rz -90 -t 5.0 3.95
3.8 -a 2 -t 4 0 0
$ xform -rz 180 -rx -90 objects/glazing_W.rad objects/venetian80.rad
materials/room.mat objects/Swall.rad > bsdf/windows-v80-1.rad
$ xform -t 12 -3 -4 bsdf/windows-v80-1.rad > bsdf/windows-v80.rad
$ genBSDF +f +b -geom meter -dim 1 7 0.8 1.8 -.2 0 -t4 5
bsdf/windows-v80.rad > bsdf/windows-v80_t45.xml
Please let me know if you need to look at any of my files.
Thanks,
Leyla
···
On Fri, Sep 6, 2013 at 5:12 PM, Greg Ward <[email protected]> wrote:
The tensor-tree code in genBSDF hasn't yet been ported to Windows. So, if
you're under Windows, that is your problem.
-Greg
*From: *Leyla Sanati <[email protected]>
*Date: *September 6, 2013 2:58:02 PM PDT
*
*
Hi all,
I am trying to generate a tensor tree BSDF for the 5-phase simulation, but
I get an error.
Here is what I type:
genBSDF +f +b -geom meter -dim 1 7 0.8 1.8 -.2 0 -t4 5
bsdf/windows-v80.rad > bsdf/windows-v80_t45.xml
And it returns this error repeatedly:
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
.
.
.
.
Does anyone know what might be wrong?
Attached is my windows-v80.rad file.
Best,
Leyla
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
Hi Leyla,
I don't think you are doing anything wrong. In fact, the errors you were getting were due to a floating point round-off problem that up until now had never manifested. I guess your geometry tripped it, somehow. I have checked in a fixed version, which should eventually show up on the downloads sites. Meanwhile, you can manually change line # 184 in your copy of genBSDF to read:
in_disk_rgn = floor((.999995*in_disk_phi + PI/4)/(PI/2)) + 1;
The new .999995 multiplier avoids the round-off problem at phi=315 degrees.
It is unusual to model a system consisting of two windows. Normally, you would model each window system, or just one of the two if they are identical, then re-use the generated BSDF. Do include the window surrounds in your model as far as the sills go, but not the walls around them. Unless you have some special reason for doing that.
Cheers,
-Greg
···
From: Leyla Sanati <[email protected]>
Date: September 7, 2013 8:54:58 AM PDT
Hi Greg,
Thank you for your prompt reply.
I am doing my simulations on my mac, and I have the latest version of Radiance.
If you may remember from my presentation, I modeled a room with four openings on the south wall: two clerestories, and two windows.
Similar to my 3-phase simulation, I am trying to generate separate BSDFs for windows and clerestories, as they have different and dynamic shading systems. The room is positioned 3 meters above the ground (on second floor).
For the windows, I generated venetian blinds using genblind. I used -a option in xform to duplicate the venetian blinds. Then I used xform to position the South wall, the windows and venetian blinds on xy plane with the interior wall facing up at z=0. In genBSDF script, I set -dim command to cover the two windows and the wall between them. Since I am new to genBSDF, I am not sure what is causing the error. Does the clerestory openings on the south wall need to be filled when creating the window BSDFs?
#objects/venetain80.rad
void plastic blindslat
0
0
5 0.7 0.7 0.7 0.1 0
!genblinds blindslat blinds .02 2 1 83 80 +r .2 | xform -rz -90 -t 5.0 3.95 3.8 -a 2 -t 4 0 0
$ xform -rz 180 -rx -90 objects/glazing_W.rad objects/venetian80.rad materials/room.mat objects/Swall.rad > bsdf/windows-v80-1.rad
$ xform -t 12 -3 -4 bsdf/windows-v80-1.rad > bsdf/windows-v80.rad
$ genBSDF +f +b -geom meter -dim 1 7 0.8 1.8 -.2 0 -t4 5 bsdf/windows-v80.rad > bsdf/windows-v80_t45.xml
Please let me know if you need to look at any of my files.
Thanks,
Leyla
On Fri, Sep 6, 2013 at 5:12 PM, Greg Ward <[email protected]> wrote:
The tensor-tree code in genBSDF hasn't yet been ported to Windows. So, if you're under Windows, that is your problem.
-Greg
From: Leyla Sanati <[email protected]>
Date: September 6, 2013 2:58:02 PM PDT
Hi all,
I am trying to generate a tensor tree BSDF for the 5-phase simulation, but I get an error.
Here is what I type:
genBSDF +f +b -geom meter -dim 1 7 0.8 1.8 -.2 0 -t4 5 bsdf/windows-v80.rad > bsdf/windows-v80_t45.xml
And it returns this error repeatedly:
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
rcontrib: select: domain error
rcontrib: floor: domain error
.
.
.
.
Does anyone know what might be wrong?
Attached is my windows-v80.rad file.
Best,
Leyla
Hello,
I'm new to the radiance thing so my question might seem an easy one.
I tried generating an .xml file for a window.rad file containing the following:****************************************************void glass windowglass003 .1 .1 .1
windowglass polygon window00120 .625 -1.3750 1.375 -1.3750 1.375 -.6250 .625 -.625****************************************************And in the genBSDF file I used the following command:
****************************************************genBSDF.exe srcwindow.rad > srcwindowbsdf.xml****************************************************
however I get the following error:
***************************************************Illegal division by zero at script/genBSDF.pl line 158.***************************************************
I tried looking any forums related to this error but I couldn't find any.
I would be grateful if anyone can guide me through the process of creating a genBSDF.xml file and adding it as a material.
Regards.
Islam Mashaly.
Islam,
I checked line 158 and it seems to be different from yours. Could you post
yours here? It should be in ray/src/util/genBSDF.pl
THANKS
German
···
2013/9/30 Islam Mashaly <[email protected]>
Hello,
I'm new to the radiance thing so my question might seem an easy one.
I tried generating an .xml file for a window.rad file containing the
following:
****************************************************
void glass windowglass
0
0
3 .1 .1 .1
windowglass polygon window
0
0
12
0 .625 -1.375
0 1.375 -1.375
0 1.375 -.625
0 .625 -.625
****************************************************
And in the genBSDF file I used the following command:
****************************************************
genBSDF.exe srcwindow.rad > srcwindowbsdf.xml
****************************************************
however I get the following error:
***************************************************
Illegal division by zero at script/genBSDF.pl line 158.
***************************************************
I tried looking any forums related to this error but I couldn't find any.
I would be grateful if anyone can guide me through the process of creating
a genBSDF.xml file and adding it as a material.
Regards.
Islam Mashaly.
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
Hello Islam,
As Germán points out, it is difficult to pinpoint your problem based on a compiled, older version of genBSDF. However, I can guess the cause based on your input. Normally, a sample surface should be laying in the Z-plane, but you have put yours in the X-plane. Furthermore, your system has essentially zero thickness. There is a line in genBSDF where it divides by the X and Y dimensions, which should never be zero, but the width in X is zero in your case.
Beyond that, I should point out that it is neither necessary nor recommended to create an XML file to describe a pane of clear glass. Even the tensor tree format has difficulty representing perfectly specular systems such as this, and in fact all the default algorithms in WINDOW6 and Radiance work quite well with a more exact description. Can you explain why you even think you need to do this?
Cheers,
-Greg
···
From: Germán Molina Larrain <[email protected]>
Date: September 30, 2013 7:16:06 AM PDT
Islam,
I checked line 158 and it seems to be different from yours. Could you post yours here? It should be in ray/src/util/genBSDF.pl
THANKS
German
2013/9/30 Islam Mashaly <[email protected]>
Hello,
I'm new to the radiance thing so my question might seem an easy one.
I tried generating an .xml file for a window.rad file containing the following:
****************************************************
void glass windowglass
0
0
3 .1 .1 .1
windowglass polygon window
0
0
12
0 .625 -1.375
0 1.375 -1.375
0 1.375 -.625
0 .625 -.625
****************************************************
And in the genBSDF file I used the following command:
****************************************************
genBSDF.exe srcwindow.rad > srcwindowbsdf.xml
****************************************************
however I get the following error:
***************************************************
Illegal division by zero at script/genBSDF.pl line 158.
***************************************************
I tried looking any forums related to this error but I couldn't find any.
I would be grateful if anyone can guide me through the process of creating a genBSDF.xml file and adding it as a material.
Regards.
Islam Mashaly.
One scenario I can think of would be using the 3- or 5-phase method to look at a window with clear glazing and a shade, and wanting to select the appropriate BSDF based on whether the shade is up or down (at runtime). But you're right Greg, of course; given the difficulty of even a high resolution BSDF to represent purely specular transmittance or reflectance, the best way to handle these situations is to do simply calculate the view matrix with the whole model - including the glazing, modeled as glass or trans or whatever -- and then do another round where you do the VMX to the glass and the DMX to the sky, and just have the one BSDF for the CFS. With gendaymtx, the generation of the matrices is so fast now it makes more sense to just do them both (for this scenario) and then at runtime just select the appropriate result.
- Rob
···
From: Greg Ward [mailto:[email protected]]
Sent: Monday, September 30, 2013 10:29 AM
To: Radiance general discussion
Subject: Re: [Radiance-general] genBSDF error
Beyond that, I should point out that it is neither necessary nor recommended to create an XML file to describe a pane of clear glass. Even the tensor tree format has difficulty representing perfectly specular systems such as this, and in fact all the default algorithms in WINDOW6 and Radiance work quite well with a more exact description. Can you explain why you even think you need to do this?
You're right, Rob. I remember I even wrote a script at some point for HMG to create clear glazing XML files, since it was faster than running genBSDF. Faster still would be using WINDOW6 to create the appropriate XML files.
-Greg
···
From: "Guglielmetti, Robert" <[email protected]>
Date: September 30, 2013 11:28:24 AM PDT
One scenario I can think of would be using the 3- or 5-phase method to look at a window with clear glazing and a shade, and wanting to select the appropriate BSDF based on whether the shade is up or down (at runtime). But you’re right Greg, of course; given the difficulty of even a high resolution BSDF to represent purely specular transmittance or reflectance, the best way to handle these situations is to do simply calculate the view matrix with the whole model – including the glazing, modeled as glass or trans or whatever -- and then do another round where you do the VMX to the glass and the DMX to the sky, and just have the one BSDF for the CFS. With gendaymtx, the generation of the matrices is so fast now it makes more sense to just do them both (for this scenario) and then at runtime just select the appropriate result.
- Rob
From: Greg Ward [mailto:[email protected]]
Sent: Monday, September 30, 2013 10:29 AM
To: Radiance general discussion
Subject: Re: [Radiance-general] genBSDF error
Beyond that, I should point out that it is neither necessary nor recommended to create an XML file to describe a pane of clear glass. Even the tensor tree format has difficulty representing perfectly specular systems such as this, and in fact all the default algorithms in WINDOW6 and Radiance work quite well with a more exact description. Can you explain why you even think you need to do this?
Yeah, we were gonna go a similar route with OpenStudio, including a simple library of "clear BSDFs" to be used with models, but in the end we are going to do what I described, doing a two pass daylight coefficient deal, once for the VMX with glazing, and once as a three phase pair of VMX/DMX. This was all thanks to Andy for plotting my stupid "clear BSDFs".
Which I guess means there is now technically a four phase method. =)
-Rob
···
On Sep 30, 2013, at 4:21 PM, Greg Ward <[email protected]> wrote:
You're right, Rob. I remember I even wrote a script at some point for HMG to create clear glazing XML files, since it was faster than running genBSDF. Faster still would be using WINDOW6 to create the appropriate XML files.
-Greg
From: "Guglielmetti, Robert" <[email protected]>
Date: September 30, 2013 11:28:24 AM PDT
One scenario I can think of would be using the 3- or 5-phase method to look at a window with clear glazing and a shade, and wanting to select the appropriate BSDF based on whether the shade is up or down (at runtime). But you’re right Greg, of course; given the difficulty of even a high resolution BSDF to represent purely specular transmittance or reflectance, the best way to handle these situations is to do simply calculate the view matrix with the whole model – including the glazing, modeled as glass or trans or whatever -- and then do another round where you do the VMX to the glass and the DMX to the sky, and just have the one BSDF for the CFS. With gendaymtx, the generation of the matrices is so fast now it makes more sense to just do them both (for this scenario) and then at runtime just select the appropriate result.
- Rob
From: Greg Ward [mailto:[email protected]]
Sent: Monday, September 30, 2013 10:29 AM
To: Radiance general discussion
Subject: Re: [Radiance-general] genBSDF error
Beyond that, I should point out that it is neither necessary nor recommended to create an XML file to describe a pane of clear glass. Even the tensor tree format has difficulty representing perfectly specular systems such as this, and in fact all the default algorithms in WINDOW6 and Radiance work quite well with a more exact description. Can you explain why you even think you need to do this?
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
That makes sense... actualy, what I am doing in mkSchedule is loading the
V, D and all the Ts matrices, but the simulations is made by choosing DC
(=VTD) for each... this avoids re-multiplying VxTxD again and again. It was
much more efficient.
···
2013/9/30 Rob Guglielmetti <[email protected]>
Yeah, we were gonna go a similar route with OpenStudio, including a simple
library of "clear BSDFs" to be used with models, but in the end we are
going to do what I described, doing a two pass daylight coefficient deal,
once for the VMX with glazing, and once as a three phase pair of VMX/DMX.
This was all thanks to Andy for plotting my stupid "clear BSDFs".
Which I guess means there is now technically a four phase method. =)
-Rob
On Sep 30, 2013, at 4:21 PM, Greg Ward <[email protected]> wrote:
You're right, Rob. I remember I even wrote a script at some point for HMG
to create clear glazing XML files, since it was faster than running
genBSDF. Faster still would be using WINDOW6 to create the appropriate XML
files.
-Greg
*From: *"Guglielmetti, Robert" <[email protected]>
*Date: *September 30, 2013 11:28:24 AM PDT
*
*
One scenario I can think of would be using the 3- or 5-phase method to
look at a window with clear glazing and a shade, and wanting to select the
appropriate BSDF based on whether the shade is up or down (at runtime). But
you’re right Greg, of course; given the difficulty of even a high
resolution BSDF to represent purely specular transmittance or reflectance,
the best way to handle these situations is to do simply calculate the view
matrix with the whole model – including the glazing, modeled as glass or
trans or whatever -- and then do another round where you do the VMX to the
glass and the DMX to the sky, and just have the one BSDF for the CFS. With
gendaymtx, the generation of the matrices is so fast now it makes more
sense to just do them both (for this scenario) and then at runtime just
select the appropriate result. ****
** **
- Rob****
** **
*From:* Greg Ward [mailto:[email protected] <[email protected]>]
*Sent:* Monday, September 30, 2013 10:29 AM
*To:* Radiance general discussion
*Subject:* Re: [Radiance-general] genBSDF error****
** **
Beyond that, I should point out that it is neither necessary nor
recommended to create an XML file to describe a pane of clear glass. Even
the tensor tree format has difficulty representing perfectly specular
systems such as this, and in fact all the default algorithms in WINDOW6 and
Radiance work quite well with a more exact description. Can you explain
why you even think you need to do this?
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
I did not connect my comment to the topic... what I mean, is that a loading
to mkSchedule or something a "blinds-up clear DC" could be useful for
applying this same method.
···
2013/9/30 Germán Molina Larrain <[email protected]>
That makes sense... actualy, what I am doing in mkSchedule is loading the
V, D and all the Ts matrices, but the simulations is made by choosing DC
(=VTD) for each... this avoids re-multiplying VxTxD again and again. It was
much more efficient.
2013/9/30 Rob Guglielmetti <[email protected]>
Yeah, we were gonna go a similar route with OpenStudio, including a
simple library of "clear BSDFs" to be used with models, but in the end we
are going to do what I described, doing a two pass daylight coefficient
deal, once for the VMX with glazing, and once as a three phase pair of
VMX/DMX. This was all thanks to Andy for plotting my stupid "clear BSDFs".
Which I guess means there is now technically a four phase method. =)
-Rob
On Sep 30, 2013, at 4:21 PM, Greg Ward <[email protected]> wrote:
You're right, Rob. I remember I even wrote a script at some point for
HMG to create clear glazing XML files, since it was faster than running
genBSDF. Faster still would be using WINDOW6 to create the appropriate XML
files.
-Greg
*From: *"Guglielmetti, Robert" <[email protected]>
*Date: *September 30, 2013 11:28:24 AM PDT
*
*
One scenario I can think of would be using the 3- or 5-phase method to
look at a window with clear glazing and a shade, and wanting to select the
appropriate BSDF based on whether the shade is up or down (at runtime). But
you’re right Greg, of course; given the difficulty of even a high
resolution BSDF to represent purely specular transmittance or reflectance,
the best way to handle these situations is to do simply calculate the view
matrix with the whole model – including the glazing, modeled as glass or
trans or whatever -- and then do another round where you do the VMX to the
glass and the DMX to the sky, and just have the one BSDF for the CFS. With
gendaymtx, the generation of the matrices is so fast now it makes more
sense to just do them both (for this scenario) and then at runtime just
select the appropriate result. ****
** **
- Rob****
** **
*From:* Greg Ward [mailto:[email protected] <[email protected]>]
*Sent:* Monday, September 30, 2013 10:29 AM
*To:* Radiance general discussion
*Subject:* Re: [Radiance-general] genBSDF error****
** **
Beyond that, I should point out that it is neither necessary nor
recommended to create an XML file to describe a pane of clear glass. Even
the tensor tree format has difficulty representing perfectly specular
systems such as this, and in fact all the default algorithms in WINDOW6 and
Radiance work quite well with a more exact description. Can you explain
why you even think you need to do this?
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
Hello, I'm using genBSDF in Windows and I get the following error:
"Could not wrap BSDFs data."
In the Radiance mailing list I saw that others have already inquired about this error (http://comments.gmane.org/gmane.comp.graphics.radiance.general/5234) and that the genBSDF script has been fixed. I wanted to know if I can have access to the fixed script.
Thank you very much, Ayelén Villalba.
···
--
Dra. D.I. Ayelén María Villalba
Laboratorio de Ambiente Humano y Vivienda
INCIHUSA - CCT-Mendoza - CONICET
Av. Ruiz Leal s/n, Parque Gral. San Martín
Mendoza - Argentina
Tel 0054-261-5244322
The latest installers on the NREL mirror include the updated script:
https://github.com/NREL/Radiance/releases/tag/5.0.a.8
You could also download the latest genBSDF.pl directly from this link:
https://github.com/NREL/Radiance/blob/combined/src/util/genBSDF.pl
Have fun!
- Rob
···
On Feb 16, 2016, at 7:07 AM, Ayelen Maria Villalba <[email protected]> wrote:
Hello, I'm using genBSDF in Windows and I get the following error:
"Could not wrap BSDFs data."
In the Radiance mailing list I saw that others have already inquired about this error (http://comments.gmane.org/gmane.comp.graphics.radiance.general/5234) and that the genBSDF script has been fixed. I wanted to know if I can have access to the fixed script.
Thank you very much, Ayelén Villalba.
--
Dra. D.I. Ayelén María Villalba
Laboratorio de Ambiente Humano y Vivienda
INCIHUSA - CCT-Mendoza - CONICET
Av. Ruiz Leal s/n, Parque Gral. San Martín
Mendoza - Argentina
Tel 0054-261-5244322
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general