ies2rad

Some of you may remember about a week ago I tuned with news of a broken Radiance installation, broken because I was trying to upgrade to 3.5 so I could begin experimenting with ies2rad. Well, my machine is fixed (the HEAD release compiled just fine, thanks for the advice Francesco), so I am playing around with this ies2rad business, and naturally, I'm showing up here once again scratching my head. I'd like to be able to use ies2rad to build a luminaire library of frozen octrees, and then use replmarks to rapidly get CAD layouts into my Radiance models.

I did this:

$ ies2rad -di -t default -u white -m (various) -o dl Dla216.ies

$ oconv -f dl.rad > dl.oct

$ replmarks -i dl.oct room_llights room_llights.tmp > lights.rad

$ oconv mats.rad *.rad > room.oct

Then I looked at the octree in rview. There are glowing discs where the fixtures are, and the intensity of them changes accordingly when you try different multipliers (-m) in ies2rad. Problem is, the room is totally dark. A quick look at the scene files in objview reveals everything to be in order, but when lit with my new lights, I got me one black room. All points in the room are 0L, except for the fixture's luminous aperture. The ies file is just your basic run-of-the-mill Edison Price 6" downlight, fresh off the internet. Ian Ashdown's IES file parser doesn't have any beefs with the file, and indeed it looks fine upon visual inspection.

What the heck am I missing?

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

You can't have frozen octrees of lights and embed them in a scene. Only rad files without light source entities work with octree instances.

Martin Moeck, Penn State

···

-----Original Message-----
  From: Rob Guglielmetti [mailto:[email protected]]
  Sent: Mon 8/25/2003 5:48 PM
  To: radiance-general
  Cc:
  Subject: [Radiance-general] ies2rad
  
  Some of you may remember about a week ago I tuned with news of a broken
  Radiance installation, broken because I was trying to upgrade to 3.5 so
  I could begin experimenting with ies2rad. Well, my machine is fixed
  (the HEAD release compiled just fine, thanks for the advice Francesco),
  so I am playing around with this ies2rad business, and naturally, I'm
  showing up here once again scratching my head. I'd like to be able to
  use ies2rad to build a luminaire library of frozen octrees, and then use
  replmarks to rapidly get CAD layouts into my Radiance models.
  
  I did this:
  
  $ ies2rad -di -t default -u white -m (various) -o dl Dla216.ies
  
  $ oconv -f dl.rad > dl.oct
  
  $ replmarks -i dl.oct room_llights room_llights.tmp > lights.rad
  
  $ oconv mats.rad *.rad > room.oct
  
  Then I looked at the octree in rview. There are glowing discs where the
  fixtures are, and the intensity of them changes accordingly when you try
  different multipliers (-m) in ies2rad. Problem is, the room is totally
  dark. A quick look at the scene files in objview reveals everything to
  be in order, but when lit with my new lights, I got me one black room.
  All points in the room are 0L, except for the fixture's luminous
  aperture. The ies file is just your basic run-of-the-mill Edison Price
  6" downlight, fresh off the internet. Ian Ashdown's IES file parser
  doesn't have any beefs with the file, and indeed it looks fine upon
  visual inspection.
  
  What the heck am I missing?
  
  ----
  
        Rob Guglielmetti
  
  e. [email protected]
  w. www.rumblestrip.org
  
  _______________________________________________
  Radiance-general mailing list
  [email protected]
  http://www.radiance-online.org/mailman/listinfo/radiance-general

Hi Rob,

You must have missed the caveat in the Radiance reference manual relating to instances, which specifically states that light sources do not work within instances. There are a couple of reasons for this, the main one being to discourage people from creating thousands of light sources in their scenes, which is a disaster for the calculation times. If you need to do this, use the -x option of replmarks instead of -i and just create copies of the geometry. If the geometry itself is too complicated, instance all the geometry except for the actual emitting surfaces, and put them in with replmarks -x.

-Greg

···

From: Rob Guglielmetti <[email protected]>
Date: Mon Aug 25, 2003 2:48:13 PM US/Pacific

Some of you may remember about a week ago I tuned with news of a broken Radiance installation, broken because I was trying to upgrade to 3.5 so I could begin experimenting with ies2rad. Well, my machine is fixed (the HEAD release compiled just fine, thanks for the advice Francesco), so I am playing around with this ies2rad business, and naturally, I'm showing up here once again scratching my head. I'd like to be able to use ies2rad to build a luminaire library of frozen octrees, and then use replmarks to rapidly get CAD layouts into my Radiance models.

I did this:

$ ies2rad -di -t default -u white -m (various) -o dl Dla216.ies

$ oconv -f dl.rad > dl.oct

$ replmarks -i dl.oct room_llights room_llights.tmp > lights.rad

$ oconv mats.rad *.rad > room.oct

Then I looked at the octree in rview. There are glowing discs where the fixtures are, and the intensity of them changes accordingly when you try different multipliers (-m) in ies2rad. Problem is, the room is totally dark. A quick look at the scene files in objview reveals everything to be in order, but when lit with my new lights, I got me one black room. All points in the room are 0L, except for the fixture's luminous aperture. The ies file is just your basic run-of-the-mill Edison Price 6" downlight, fresh off the internet. Ian Ashdown's IES file parser doesn't have any beefs with the file, and indeed it looks fine upon visual inspection.

What the heck am I missing?

     Rob Guglielmetti

THanks Greg (and Martin). I'll have a look at this tomorrow. Man, what flexibility Radiance users have! Even though you can't instance light sources, you can separate the geometry from the emitter if you want to, in the case of a complex luminaire, to get the best of both worlds. That's great.

Rob Guglielmetti
[email protected]
www.rumblestrip.org

···

On Monday, August 25, 2003, at 06:13 PM, Greg Ward wrote:

Hi Rob,

You must have missed the caveat in the Radiance reference manual relating to instances, which specifically states that light sources do not work within instances. There are a couple of reasons for this, the main one being to discourage people from creating thousands of light sources in their scenes, which is a disaster for the calculation times. If you need to do this, use the -x option of replmarks instead of -i and just create copies of the geometry. If the geometry itself is too complicated, instance all the geometry except for the actual emitting surfaces, and put them in with replmarks -x.