The curious case of the Ghost Luminaire

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

Hi Pally,

I don't know who or what is setting your rendering parameters, but they are not something I would recommend. The "rad" program is a good way to get sane rendering options if you don't like the defaults. Setting -dt .25 -dc .35 is causing the strange contours around your light sources as well as numerous other inaccuracies. You are telling Radiance that you are fine with 25% error in the direct calculation, meaning everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface that extends over your wall surface(s). I can't tell you for sure because the zip file you uploaded doesn't contain a complete model -- there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

···

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

Greg I'd completely forgotten about this issue & was reminded by a colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files - DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc = 0.95) View file - Monday.rdv Octree file - image5.oct Pic file - image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build from source is a few months old (mid-September IIRC) as I haven't had time to re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

···

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but they are not something I would recommend. The "rad" program is a good way to get sane rendering options if you don't like the defaults. Setting -dt .25 -dc .35 is causing the strange contours around your light sources as well as numerous other inaccuracies. You are telling Radiance that you are fine with 25% error in the direct calculation, meaning everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface that extends over your wall surface(s). I can't tell you for sure because the zip file you uploaded doesn't contain a complete model -- there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

I can't really tell what's going on because I can't reproduce your strange image with the ghost luminaire. The ghost doesn't even match the appearance of the luminaires you've defined, so it makes no sense to me.

You should retry with the latest HEAD version to see if it's still happening for you. If it is, then we have an issue I can investigate further.

Cheers,
-Greg

···

From: Pally Sandher <[email protected]>
Date: February 13, 2013 7:40:44 AM PST

Greg I'd completely forgotten about this issue & was reminded by a colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files - DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc = 0.95) View file - Monday.rdv Octree file - image5.oct Pic file - image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build from source is a few months old (mid-September IIRC) as I haven't had time to re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but they are not something I would recommend. The "rad" program is a good way to get sane rendering options if you don't like the defaults. Setting -dt .25 -dc .35 is causing the strange contours around your light sources as well as numerous other inaccuracies. You are telling Radiance that you are fine with 25% error in the direct calculation, meaning everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface that extends over your wall surface(s). I can't tell you for sure because the zip file you uploaded doesn't contain a complete model -- there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

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

Greg I'm using the snapshot from http://public.kitware.com/gitweb?p=radiance.git;a=tree which doesn't appear to have been changed since the 15th of September. http://radiance-online.org/cgi-bin/viewcvs.cgi/?pathrev=MAIN doesn't appear to have tarball generation enabled.

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

···

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 13 February 2013 19:14
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

I can't really tell what's going on because I can't reproduce your strange image with the ghost luminaire. The ghost doesn't even match the appearance of the luminaires you've defined, so it makes no sense to me.

You should retry with the latest HEAD version to see if it's still happening for you. If it is, then we have an issue I can investigate further.

Cheers,
-Greg

From: Pally Sandher <[email protected]>
Date: February 13, 2013 7:40:44 AM PST

Greg I'd completely forgotten about this issue & was reminded by a colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files - DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc = 0.95) View file - Monday.rdv Octree file - image5.oct Pic file - image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build from source is a few months old (mid-September IIRC) as I haven't had time to re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but they are not something I would recommend. The "rad" program is a good way to get sane rendering options if you don't like the defaults. Setting -dt .25 -dc .35 is causing the strange contours around your light sources as well as numerous other inaccuracies. You are telling Radiance that you are fine with 25% error in the direct calculation, meaning everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface that extends over your wall surface(s). I can't tell you for sure because the zip file you uploaded doesn't contain a complete model -- there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

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

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

Pally,

I just ran your model on a LINUX box. Here is some extra info and
additional steps that were required:

$ getinfo image5.oct
image5.oct:
  #?RADIANCE
  oconv -f Monday.map Monday.rad
  FORMAT=Radiance_octree

$ getinfo image5.pic
image5.pic:
  #?RADIANCE
  oconv -f Monday.map Monday.rad
  OCTREE=image5.oct
  rpict -w -t 5 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu
0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 5 -pt 0.075 -pj 1.0
-dj 0.600 -ds 0.200 -dt 0.025 -dc 0.950 -dr 3 -dp 256 -st 0.300 -ab 2
-aa 0.100 -ar 256 -ad 512 -as 256 -lr 8 -lw 0.002 -av 0.000 0.000
0.000 -e Radiance.log -af temp.amb
  SOFTWARE= RADIANCE 4.2a built by Integrated Environmental Solutions
Limited. (September 2012)
  VIEW= -vtv -vp 6.3 8.1 1.8 -vd -2.8 -3.6 0 -vu 0 0 1 -vh 80 -vv 60
-vo 0 -va 0 -vs 0 -vl 0
  CAPDATE= 2012:11:02 12:18:35
  GMT= 2012:11:02 12:18:35
  FORMAT=32-bit_rle_rgbe
  pfilt -m 0.25 -x /2 -y /2
  EXPOSURE=3.934435e+000

$ gvim Monday.rad (paths)
$ gvim Monday.rdv (prepend 'rvu')
$ cat Monday.rdp
-ps 5 -pt 0.075 -pj 1.0 -dj 0.600 -ds 0.200 -dt 0.025 -dc 0.950 -dr 3
-dp 256 -st 0.300 -ab 2 -aa 0.100 -ar 256 -ad 512 -as 256 -lr 8 -lw
0.002 -av 0.000 0.000 0.000 -e Radiance.log

$ oconv Monday.map Monday.rad > Monday.oct

Test a) - With opts file Monday.rdp
$ rpict -vf Monday.rdv @Monday.rdp Monday.oct > axel.hdr
Test b) - With opts from image5.pic
$ rpict -w -t 5 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu
0 0 1 -vh 80.000 -vv 60.000 -ps 5 -pt 0.075 -pj 1.0 -dj 0.600 -ds
0.200 -dt 0.025 -dc 0.950 -dr 3 -dp 256 -st 0.300 -ab 2 -aa 0.100 -ar
256 -ad 512 -as 256 -lr 8 -lw 0.002 -av 0.000 0.000 0.000 Monday.oct >
image5_axel.hdr

Both images (image5_axel.hdr and axel.hdr) don't show the ghost
luminaires that you have in image5.pic. No idea what's going on.

I also inspected your model in Rhino3d. The surfaces that IESVE
generated can be a bit weird, especially when you don't enable 'inner
volumes'. In your case, the walls at least are good, with no co-planar
surfaces.

Cheers

Axel

···

On 14 February 2013 11:10, Pally Sandher <[email protected]> wrote:

Greg I'm using the snapshot from http://public.kitware.com/gitweb?p=radiance.git;a=tree which doesn't appear to have been changed since the 15th of September. http://radiance-online.org/cgi-bin/viewcvs.cgi/?pathrev=MAIN doesn't appear to have tarball generation enabled.

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 13 February 2013 19:14
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

I can't really tell what's going on because I can't reproduce your strange image with the ghost luminaire. The ghost doesn't even match the appearance of the luminaires you've defined, so it makes no sense to me.

You should retry with the latest HEAD version to see if it's still happening for you. If it is, then we have an issue I can investigate further.

Cheers,
-Greg

From: Pally Sandher <[email protected]>
Date: February 13, 2013 7:40:44 AM PST

Greg I'd completely forgotten about this issue & was reminded by a colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files - DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc = 0.95) View file - Monday.rdv Octree file - image5.oct Pic file - image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build from source is a few months old (mid-September IIRC) as I haven't had time to re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but they are not something I would recommend. The "rad" program is a good way to get sane rendering options if you don't like the defaults. Setting -dt .25 -dc .35 is causing the strange contours around your light sources as well as numerous other inaccuracies. You are telling Radiance that you are fine with 25% error in the direct calculation, meaning everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface that extends over your wall surface(s). I can't tell you for sure because the zip file you uploaded doesn't contain a complete model -- there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

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

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

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

Perhaps this link can help:

  http://ghostbusters.en.softonic.com/

It's supposed to work on Windows, but maybe it will work for luminaires, too.

Cheers,
-Greg

···

Sent from my iPad

On Feb 14, 2013, at 8:33 AM, Axel Jacobs <[email protected]> wrote:

Pally,

I just ran your model on a LINUX box. Here is some extra info and
additional steps that were required:

$ getinfo image5.oct
image5.oct:
   #?RADIANCE
   oconv -f Monday.map Monday.rad
   FORMAT=Radiance_octree

$ getinfo image5.pic
image5.pic:
   #?RADIANCE
   oconv -f Monday.map Monday.rad
   OCTREE=image5.oct
   rpict -w -t 5 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu
0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 5 -pt 0.075 -pj 1.0
-dj 0.600 -ds 0.200 -dt 0.025 -dc 0.950 -dr 3 -dp 256 -st 0.300 -ab 2
-aa 0.100 -ar 256 -ad 512 -as 256 -lr 8 -lw 0.002 -av 0.000 0.000
0.000 -e Radiance.log -af temp.amb
   SOFTWARE= RADIANCE 4.2a built by Integrated Environmental Solutions
Limited. (September 2012)
   VIEW= -vtv -vp 6.3 8.1 1.8 -vd -2.8 -3.6 0 -vu 0 0 1 -vh 80 -vv 60
-vo 0 -va 0 -vs 0 -vl 0
   CAPDATE= 2012:11:02 12:18:35
   GMT= 2012:11:02 12:18:35
   FORMAT=32-bit_rle_rgbe
   pfilt -m 0.25 -x /2 -y /2
   EXPOSURE=3.934435e+000

$ gvim Monday.rad (paths)
$ gvim Monday.rdv (prepend 'rvu')
$ cat Monday.rdp
-ps 5 -pt 0.075 -pj 1.0 -dj 0.600 -ds 0.200 -dt 0.025 -dc 0.950 -dr 3
-dp 256 -st 0.300 -ab 2 -aa 0.100 -ar 256 -ad 512 -as 256 -lr 8 -lw
0.002 -av 0.000 0.000 0.000 -e Radiance.log

$ oconv Monday.map Monday.rad > Monday.oct

Test a) - With opts file Monday.rdp
$ rpict -vf Monday.rdv @Monday.rdp Monday.oct > axel.hdr
Test b) - With opts from image5.pic
$ rpict -w -t 5 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu
0 0 1 -vh 80.000 -vv 60.000 -ps 5 -pt 0.075 -pj 1.0 -dj 0.600 -ds
0.200 -dt 0.025 -dc 0.950 -dr 3 -dp 256 -st 0.300 -ab 2 -aa 0.100 -ar
256 -ad 512 -as 256 -lr 8 -lw 0.002 -av 0.000 0.000 0.000 Monday.oct >
image5_axel.hdr

Both images (image5_axel.hdr and axel.hdr) don't show the ghost
luminaires that you have in image5.pic. No idea what's going on.

I also inspected your model in Rhino3d. The surfaces that IESVE
generated can be a bit weird, especially when you don't enable 'inner
volumes'. In your case, the walls at least are good, with no co-planar
surfaces.

Cheers

Axel

On 14 February 2013 11:10, Pally Sandher <[email protected]> wrote:

Greg I'm using the snapshot from http://public.kitware.com/gitweb?p=radiance.git;a=tree which doesn't appear to have been changed since the 15th of September. http://radiance-online.org/cgi-bin/viewcvs.cgi/?pathrev=MAIN doesn't appear to have tarball generation enabled.

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 13 February 2013 19:14
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

I can't really tell what's going on because I can't reproduce your strange image with the ghost luminaire. The ghost doesn't even match the appearance of the luminaires you've defined, so it makes no sense to me.

You should retry with the latest HEAD version to see if it's still happening for you. If it is, then we have an issue I can investigate further.

Cheers,
-Greg

From: Pally Sandher <[email protected]>
Date: February 13, 2013 7:40:44 AM PST

Greg I'd completely forgotten about this issue & was reminded by a colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files - DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc = 0.95) View file - Monday.rdv Octree file - image5.oct Pic file - image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build from source is a few months old (mid-September IIRC) as I haven't had time to re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but they are not something I would recommend. The "rad" program is a good way to get sane rendering options if you don't like the defaults. Setting -dt .25 -dc .35 is causing the strange contours around your light sources as well as numerous other inaccuracies. You are telling Radiance that you are fine with 25% error in the direct calculation, meaning everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface that extends over your wall surface(s). I can't tell you for sure because the zip file you uploaded doesn't contain a complete model -- there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to generate it at http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the centre left of the image which appears to be generated by the luminaire visible in the centre of the image (turning off the others while leaving it on still shows the ghost while vice-versa doesn't). Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0 0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let me know if the "ghost" luminaire also appears for you it would be most appreciated (I'm using Radiance HEAD built from source on Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

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

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

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

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

(I posted earlier from my personal address, but I don't think I'm actually
subscribed from that one. Apologies if this ends up showing up 2x...)

Hi Pally,

I can help here. The origin trunk (i.e. "master") on that Kitware git repo
does not refresh the way you might think. Kitware has a “combined”
branch that actually has the latest HEAD (updated nightly, I believe at
midnight EST). To add this, issue:

"git checkout -t origin/combined"

"git branch" will tell where you are, and what you've got checked out:

hooper:radiance rpg$ git branch
combined
* master
hooper:radiance rpg$

You can switch between branches with the 'checkout' command:

hooper:radiance rpg$ git checkout combined
Switched to branch 'combined'
hooper:radiance rpg$ git branch
* combined
master
hooper:radiance rpg$

I work off this combined branch for building current packages (speaking of
which a fresh set of them should go live on the NREL site today). I know
this is confusing, but this is a vestige from when Kitware was doing all
the heavy lifting getting the CMake build system implemented across the
source tree, and making frequent changes.

Anyhow, the takeaway here is if you wish to use the Kitware git repo for
your latest HEAD source, use the combined branch for now!

- Rob

P.S.
I know this is a bit confusing. We will hopefully set up our own (NREL)
git mirror that I'll be able to refresh whenever I need to.

Rob Guglielmetti
National Renewable Energy Laboratory (NREL)
Commercial Buildings Research Group
15013 Denver West Parkway MS:RSF202
Golden, CO 80401
303.275.4319
[email protected]

···

On 2/14/13 4:10 AM, "Pally Sandher" <[email protected]> wrote:

Greg I'm using the snapshot from
http://public.kitware.com/gitweb?p=radiance.git;a=tree which doesn't
appear to have been changed since the 15th of September.
http://radiance-online.org/cgi-bin/viewcvs.cgi/?pathrev=MAIN doesn't
appear to have tarball generation enabled.

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No.
SC151456
Registered Office - Helix Building, West Of Scotland Science Park,
Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 13 February 2013 19:14
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

I can't really tell what's going on because I can't reproduce your
strange image with the ghost luminaire. The ghost doesn't even match the
appearance of the luminaires you've defined, so it makes no sense to me.

You should retry with the latest HEAD version to see if it's still
happening for you. If it is, then we have an issue I can investigate
further.

Cheers,
-Greg

From: Pally Sandher <[email protected]>
Date: February 13, 2013 7:40:44 AM PST

Greg I'd completely forgotten about this issue & was reminded by a
colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to
http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files -
DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc =
0.95) View file - Monday.rdv Octree file - image5.oct Pic file -
image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build from
source is a few months old (mid-September IIRC) as I haven't had time to
re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No.
SC151456
Registered Office - Helix Building, West Of Scotland Science Park,
Glasgow G20 0SP
Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but they
are not something I would recommend. The "rad" program is a good way to
get sane rendering options if you don't like the defaults. Setting -dt
.25 -dc .35 is causing the strange contours around your light sources as
well as numerous other inaccuracies. You are telling Radiance that you
are fine with 25% error in the direct calculation, meaning everything
else is worse.

The ghosts you are seeing are due most likely to a glazing surface that
extends over your wall surface(s). I can't tell you for sure because
the zip file you uploaded doesn't contain a complete model -- there are
many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2 (nothing
to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to
generate it at
http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the
centre left of the image which appears to be generated by the luminaire
visible in the centre of the image (turning off the others while
leaving it on still shows the ghost while vice-versa doesn't). Rotating
the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0
0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100
-pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500
-ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000
0.000 -af temp.amb ghost_lum.oct >amb.raw
rpict -w -t 2 -vtv -vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0
0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj
0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa
0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af
temp.amb ghost_lum.oct >ghost_lum.raw
pfilt -m 0.25 -x /2 -y /2 ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let
me know if the "ghost" luminaire also appears for you it would be most
appreciated (I'm using Radiance HEAD built from source on Windows 7
SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

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

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

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

Ah ok I'll try rebuilding using a snapshot of the combined branch & see how I get on.

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No. SC151456
Registered Office - Helix Building, West Of Scotland Science Park, Glasgow G20 0SP
Email Disclaimer

···

-----Original Message-----
From: Guglielmetti, Robert [mailto:[email protected]]
Sent: 15 February 2013 02:04
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

(I posted earlier from my personal address, but I don't think I'm actually subscribed from that one. Apologies if this ends up showing up 2x...)

Hi Pally,

I can help here. The origin trunk (i.e. "master") on that Kitware git repo does not refresh the way you might think. Kitware has a “combined” branch that actually has the latest HEAD (updated nightly, I believe at midnight EST). To add this, issue:

"git checkout -t origin/combined"

"git branch" will tell where you are, and what you've got checked out:

hooper:radiance rpg$ git branch
combined
* master
hooper:radiance rpg$

You can switch between branches with the 'checkout' command:

hooper:radiance rpg$ git checkout combined Switched to branch 'combined'
hooper:radiance rpg$ git branch
* combined
master
hooper:radiance rpg$

I work off this combined branch for building current packages (speaking of which a fresh set of them should go live on the NREL site today). I know this is confusing, but this is a vestige from when Kitware was doing all the heavy lifting getting the CMake build system implemented across the source tree, and making frequent changes.

Anyhow, the takeaway here is if you wish to use the Kitware git repo for your latest HEAD source, use the combined branch for now!

- Rob

P.S.
I know this is a bit confusing. We will hopefully set up our own (NREL) git mirror that I'll be able to refresh whenever I need to.

Rob Guglielmetti
National Renewable Energy Laboratory (NREL) Commercial Buildings Research Group
15013 Denver West Parkway MS:RSF202
Golden, CO 80401
303.275.4319
[email protected]

On 2/14/13 4:10 AM, "Pally Sandher" <[email protected]> wrote:

Greg I'm using the snapshot from
http://public.kitware.com/gitweb?p=radiance.git;a=tree which doesn't
appear to have been changed since the 15th of September.
http://radiance-online.org/cgi-bin/viewcvs.cgi/?pathrev=MAIN doesn't
appear to have tarball generation enabled.

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No.
SC151456
Registered Office - Helix Building, West Of Scotland Science Park,
Glasgow G20 0SP Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 13 February 2013 19:14
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

I can't really tell what's going on because I can't reproduce your
strange image with the ghost luminaire. The ghost doesn't even match
the appearance of the luminaires you've defined, so it makes no sense to me.

You should retry with the latest HEAD version to see if it's still
happening for you. If it is, then we have an issue I can investigate
further.

Cheers,
-Greg

From: Pally Sandher <[email protected]>
Date: February 13, 2013 7:40:44 AM PST

Greg I'd completely forgotten about this issue & was reminded by a
colleague recently as he was seeing this issue intermittently.
He's provided me an archive which I've uploaded to
http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire2.zip
It should contain - Monday.rad, Monday_mod.rad Luminaire Files -
DULCET1203WW.* Parameter file - Monday.rdp (with -dt = 0.025 and -dc =
0.95) View file - Monday.rdv Octree file - image5.oct Pic file -
image5.pic (with ghost luminaire)

We're running on Windows 7 x64. The radiance code I used to build
from source is a few months old (mid-September IIRC) as I haven't had
time to re-download & re-run Cmake etc. if there's been any updates since then.

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

**Design, Simulate + Innovate with the <Virtual Environment>**
Integrated Environmental Solutions Limited. Registered in Scotland No.
SC151456
Registered Office - Helix Building, West Of Scotland Science Park,
Glasgow G20 0SP Email Disclaimer

-----Original Message-----
From: Gregory J. Ward [mailto:[email protected]]
Sent: 01 November 2012 16:33
To: code development
Subject: Re: [Radiance-dev] The curious case of the Ghost Luminaire

Hi Pally,

I don't know who or what is setting your rendering parameters, but
they are not something I would recommend. The "rad" program is a good
way to get sane rendering options if you don't like the defaults.
Setting -dt
.25 -dc .35 is causing the strange contours around your light sources
as well as numerous other inaccuracies. You are telling Radiance that
you are fine with 25% error in the direct calculation, meaning
everything else is worse.

The ghosts you are seeing are due most likely to a glazing surface
that extends over your wall surface(s). I can't tell you for sure
because the zip file you uploaded doesn't contain a complete model --
there are many referenced files that are missing.

What system are you using to assemble all this?

-Greg

From: Pally Sandher <[email protected]>
Date: November 1, 2012 9:15:20 AM PDT

Hi,
We've noticed some strange "ghost" luminaires in Radiance 4.2
(nothing to do with Halloween yesterday).
I've uploaded a file with a sample .pic file and the files used to
generate it at
http://upload.iesve.com/downloads/radiance_issues/ghost_luminaire.zip

As you can see from the .pic file there's a "ghost" luminaire to the
centre left of the image which appears to be generated by the
luminaire visible in the centre of the image (turning off the others
while leaving it on still shows the ghost while vice-versa doesn't).
Rotating the view to the right also removes the "ghost".
The commands we're running to generate the image are as follows:

oconv -f Monday.map Monday.rad > ghost_lum.oct rpict -w -t 2 -vtv
-vp 6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0
0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -x 64 -y 64 -ps 6 -pt 0.100
-pj 1.0 -dj 0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st
0.500 -ab 2 -aa 0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000
0.000
0.000 -af temp.amb ghost_lum.oct >amb.raw rpict -w -t 2 -vtv -vp
6.300 8.100 1.800 -vd -2.800 -3.600 0.000 -vu 0
0 1 -vh 80.000 -vv 60.000 -x 1600 -y 1600 -ps 6 -pt 0.100 -pj 1.0 -dj
0.350 -ds 0.300 -dt 0.250 -dc 0.350 -dr 2 -dp 128 -st 0.500 -ab 2 -aa
0.150 -ar 64 -ad 128 -as 64 -lr 6 -lw 0.005 -av 0.000 0.000 0.000 -af
temp.amb ghost_lum.oct >ghost_lum.raw pfilt -m 0.25 -x /2 -y /2
ghost_lum.raw >ghost_lum.pic

If anyone has a few minutes to run this image on their systems & let
me know if the "ghost" luminaire also appears for you it would be
most appreciated (I'm using Radiance HEAD built from source on
Windows 7 SP1).

Thanks,

Palbinder Sandher
Software Platform Engineer
T: +44 (0) 141 945 8500
F: +44 (0) 141 945 8501
http://www.iesve.com

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

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

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

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

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