Radiance, objline, objpict etc..

I was running test renders to determine some electric lighting levels at a back wall: I had run ies2rad for the fixtures and associated the distribution to simple flat geometry to the same plane as the ceiling. When I render I don't 'see' the lights so I ran objline and objpict to check that the geometry of fixtures are there
and get the following errors:

$ objline Scene_1.rad | x11meta
fatal - cannot open file "/usr/local/lib/ray/meta/vchars.mta", mode "r"

$ objpict Scene_1.rad | ximage
oconv: system - cannot open scene file "/usr/local/lib/ray/lib/testroom": No such file or directory

SO I guess I have a two part question:

1- My setenv variables don't point to the directories shown above..so why is objline and objpict calling those locations
and
2-Could someone clarify the visibility of lighting distributions in radiance. below is the fixture rad file

···

#
_________________________________________________________________

Here is the missing lighting fixture rad file (omitted in previous email)

# From: [email protected]

···

To: [email protected]
Subject: RE: [Radiance-general] Radiance, objline, objpict etc..
Date: Wed, 23 Jan 2008 19:33:04 -0500

I was running test renders to determine some electric lighting levels at a back wall: I had run ies2rad for the fixtures and associated the distribution to simple flat geometry to the same plane as the ceiling. When I render I don't 'see' the lights so I ran objline and objpict to check that the geometry of fixtures are there
and get the following errors:

$ objline Scene_1.rad | x11meta
fatal - cannot open file "/usr/local/lib/ray/meta/vchars.mta", mode "r"

$ objpict Scene_1.rad | ximage
oconv: system - cannot open scene file "/usr/local/lib/ray/lib/testroom": No such file or directory

SO I guess I have a two part question:

1- My setenv variables don't point to the directories shown above..so why is objline and objpict calling those locations
and
2-Could someone clarify the visibility of lighting distributions in radiance. below is the fixture rad file

#
_________________________________________________________________

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

_________________________________________________________________

I was running test renders to determine some electric lighting
levels at a back wall: I had run ies2rad for the fixtures and
associated the distribution to simple flat geometry

ies2rad does that for you. You don't have to add 'geometry'
to make the light appear.

to the same plane as the ceiling.

Bad move. Try to keep at least a tiny distance to the ceiling.
Actually, ies2rad creates the light sources a few mm below the
origin of the light distribution. If you place your fixture
scene file (the .rad file that's created by ies2rad) at 2.5m
the disc or rectangle will be at 2.4975m; enough for Radiance
to keep light source and ceiling apart.

When I render I don't 'see' the lights

Did you use the -i option for ies2rad? This would use an 'illum'
type for the source which is invisible in a picture. The default
light source should be visible, though.

Do you miss the bright light source itself or is there no light
at all in your scene?

so I ran objline and
objpict to check that the geometry of fixtures are there
and get the following errors:

$ objline Scene_1.rad | x11meta
fatal - cannot open file "/usr/local/lib/ray/meta/vchars.mta", mode "r"

quick guess: this path is hardcoded in one of the scripts that are used
by objline or x11meta. I think these tools are fairly uncommon these days
when everyone just fires up rvu to get a preview.

$ objpict Scene_1.rad | ximage
oconv: system - cannot open scene file "/usr/local/lib/ray/lib/testroom": No such file or directory

On my system I have '/usr/local/lib/ray/lib' hardcoded in the
script 'objpict'. I use the /usr/local/ prefix so I can not say
if it's adjusted by the installer or not. Your error suggests it's
not so you could adjust the script yourself. The path is defined
in the first line after the comments.

2-Could someone clarify the visibility of lighting distributions
in radiance.

Typically light sources from IES files are visible, as disc,
rectangle or box. You can use the '-i' option to allow a fancy
fitting geometry to appear and still have the light distribution
for the room. Also, the '-dv' option in rpict switches off the
rendering of sources as bright objects (they will appear black).

Hth,
Thomas

···

On 24 Jan 2008, at 00:33, steve michel wrote:

Hi Steve,

This is additional information to what Thomas wrote. You can manually set the MDIR environment variable to the actual path on your system. (Note that this variable only takes a single path directory, which must end with '/'.) The objpict script does seem to be hard-coded, which is a mistake you can correct as Thomas suggested by editing the file.

Best,
-Greg

···

From: steve michel <[email protected]>
Date: January 23, 2008 4:33:04 PM PST

I was running test renders to determine some electric lighting levels at a back wall: I had run ies2rad for the fixtures and associated the distribution to simple flat geometry to the same plane as the ceiling. When I render I don't 'see' the lights so I ran objline and objpict to check that the geometry of fixtures are there
and get the following errors:

$ objline Scene_1.rad | x11meta
fatal - cannot open file "/usr/local/lib/ray/meta/vchars.mta", mode "r"

$ objpict Scene_1.rad | ximage
oconv: system - cannot open scene file "/usr/local/lib/ray/lib/testroom": No such file or directory

SO I guess I have a two part question:

1- My setenv variables don't point to the directories shown above..so why is objline and objpict calling those locations
and
2-Could someone clarify the visibility of lighting distributions in radiance. below is the fixture rad file

Hi Thomas,

There is light in the scene but the lighting polygon is black not even lit. As you said I placed the polygons an inch blow the ceiling plane to no effect. It's important that the sources be visible for presentation purposes.

here is the conversion syntax I used
ies2rad -df -t default -m .85 2X2TROFFER2T8.ies

and these rpict options without the -dv switch.

F syncfile -x 1024 -y 1024 -af ambfile -t 60 -vf views/Scene_1.vf -ps 3
-pt .08
-dp 512 -ar 309 -ms 0.072 -ds 0 -dj .5 -dt 0 -dc .5 -dr 1 -sj .7 -st .1
-aa .15 -ad 1500 -as 128 -av 0.01 0.01 0.01 -lr 8 -lw .002 -ab 7 -i -w -o
output.pic Scene_1.oct

I will redo the whole run with the -i option in ies2rad...

Thanks for the tip for objpict...

regards
Steve

···

----------------------------------------

From: [email protected]
Subject: Re: [Radiance-general] Radiance, objline, objpict etc..
Date: Thu, 24 Jan 2008 19:29:05 +0000
To: [email protected]

On 24 Jan 2008, at 00:33, steve michel wrote:

I was running test renders to determine some electric lighting
levels at a back wall: I had run ies2rad for the fixtures and
associated the distribution to simple flat geometry

ies2rad does that for you. You don't have to add 'geometry'
to make the light appear.

to the same plane as the ceiling.

Bad move. Try to keep at least a tiny distance to the ceiling.
Actually, ies2rad creates the light sources a few mm below the
origin of the light distribution. If you place your fixture
scene file (the .rad file that's created by ies2rad) at 2.5m
the disc or rectangle will be at 2.4975m; enough for Radiance
to keep light source and ceiling apart.

When I render I don't 'see' the lights

Did you use the -i option for ies2rad? This would use an 'illum'
type for the source which is invisible in a picture. The default
light source should be visible, though.

Do you miss the bright light source itself or is there no light
at all in your scene?

so I ran objline and
objpict to check that the geometry of fixtures are there
and get the following errors:

$ objline Scene_1.rad | x11meta
fatal - cannot open file "/usr/local/lib/ray/meta/vchars.mta", mode
"r"

quick guess: this path is hardcoded in one of the scripts that are used
by objline or x11meta. I think these tools are fairly uncommon these
days
when everyone just fires up rvu to get a preview.

$ objpict Scene_1.rad | ximage
oconv: system - cannot open scene file "/usr/local/lib/ray/lib/
testroom": No such file or directory

On my system I have '/usr/local/lib/ray/lib' hardcoded in the
script 'objpict'. I use the /usr/local/ prefix so I can not say
if it's adjusted by the installer or not. Your error suggests it's
not so you could adjust the script yourself. The path is defined
in the first line after the comments.

2-Could someone clarify the visibility of lighting distributions
in radiance.

Typically light sources from IES files are visible, as disc,
rectangle or box. You can use the '-i' option to allow a fancy
fitting geometry to appear and still have the light distribution
for the room. Also, the '-dv' option in rpict switches off the
rendering of sources as bright objects (they will appear black).

Hth,
Thomas

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

_________________________________________________________________

I may have missed this, but what kind of light fixture is it? If it's a low brightness downlight or something similar with good shielding, It may just be that your viewing angle to the light source is shallow enough that you cannot see any luminance. Radiance will render the disc according to what the ies file says is appropriate for that angle. It does tend to appear "all or nothing", that is, if you're looking straight up into the fixture the disc is all white and very shallow angles appear all black. If you want to model the appearance of the light fixture itself, you will have to get a little fancier and model the geometry, include a glow material for the lamp brightness, and enclose the whole thing in an illum sphere and apply the luminance distribution to that. This is well detailed in RwR and in various posts on this list. I have little experience doing this without the help of front ends like Rayfront, but several people around here are. For my needs, the default rendering of the light sources is generally adequate.

- Rob Guglielmetti
www.rumblestrip.org

···

On Jan 25, 2008, at 5:34 PM, steve michel wrote:

Hi Thomas,

There is light in the scene but the lighting polygon is black not even lit. As you said I placed the polygons an inch blow the ceiling plane to no effect. It's important that the sources be visible for presentation purposes.

Rob,

The fixture is a recessed fluorescent with a flat translucent lens; therefore my needs are simply to show a glow for the lens on the same plane as the ceiling. I applied the ies distribution to a simple polygon. But anyways, for purposes of rigor I think I'll look for the latest edition of RwR. Can someone post the link to the best method of obtaining the book??..

regards
Steve

···

----------------------------------------

From: [email protected]
Subject: Re: [Radiance-general] Radiance, objline, objpict etc..
Date: Sat, 26 Jan 2008 08:51:21 -0700
To: [email protected]

On Jan 25, 2008, at 5:34 PM, steve michel wrote:

Hi Thomas,

There is light in the scene but the lighting polygon is black not
even lit. As you said I placed the polygons an inch blow the
ceiling plane to no effect. It's important that the sources be
visible for presentation purposes.

I may have missed this, but what kind of light fixture is it? If
it's a low brightness downlight or something similar with good
shielding, It may just be that your viewing angle to the light source
is shallow enough that you cannot see any luminance. Radiance will
render the disc according to what the ies file says is appropriate
for that angle. It does tend to appear "all or nothing", that is, if
you're looking straight up into the fixture the disc is all white and
very shallow angles appear all black. If you want to model the
appearance of the light fixture itself, you will have to get a little
fancier and model the geometry, include a glow material for the lamp
brightness, and enclose the whole thing in an illum sphere and apply
the luminance distribution to that. This is well detailed in RwR and
in various posts on this list. I have little experience doing this
without the help of front ends like Rayfront, but several people
around here are. For my needs, the default rendering of the light
sources is generally adequate.

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

_________________________________________________________________

Hi Steve,

"Rendering with Radiance" is available again on Amazon:

  http://www.amazon.com/Rendering-Radiance-Science-Lighting-Visualization/dp/0974538108/ref=pd_bbs_sr_1?ie=UTF8&s=books&qid=1201307677&sr=8-1

Let us know if you have any trouble ordering it.

-Greg

···

From: steve michel <[email protected]>
Date: January 26, 2008 10:50:58 AM PST

Rob,

The fixture is a recessed fluorescent with a flat translucent lens; therefore my needs are simply to show a glow for the lens on the same plane as the ceiling. I applied the ies distribution to a simple polygon. But anyways, for purposes of rigor I think I'll look for the latest edition of RwR. Can someone post the link to the best method of obtaining the book??..

regards
Steve

Rob,

The fixture is a recessed fluorescent with a flat translucent lens;

Being recessed the fixture will have no uplight component. Did you
check the orientation the IES file will have in the scene? Typically
major output direction will point downwards (-z). For a downlight you
should not need any rotation, just lift (xform -t) the *.rad file to
the right height and place in your scene.

therefore my needs are simply to show a glow for the lens on the same
plane as the ceiling. I applied the ies distribution to a simple polygon.

Did you create the polygon yourself? Ies2rad (without -i option!) will
do that and you don't need any further polygons. They will only interfere
with you're distribution. The only problem is that the polygon will have
the size of the luminous area not the of the actual fitting.

Regards,
Thomas

···

On 26 Jan 2008, at 18:50, steve michel wrote:

Hi Thomas,

MARK THIS SOLVED

I had created a 'glow' material and applying it to the same polygon in my ies.rad file. SO I got a render with my fixtures showing but still no illumination! I was trying to overlap or layer a glow material on a distribution's polygon. So you are right that the two might interfere. (But I can see where that might be useful to avoid the extra step of two distinct polygon geometry for the same thing (which I have done before with erco fixtures 3d geometry))

BUT The simpler method of deleting the -i option in ies2rad worked...I see the fixture and the light distribution...

The next challenge is to use a single light distribution material and apply to multiple similar geometry in the scene... To isolate the problem, I've been hand editing each fixture geometry and duplicating the same ies.rad parameters to each...

thanks to all
Steve

···

----------------------------------------

From: [email protected]
Subject: Re: [Radiance-general] Radiance, objline, objpict etc..
Date: Sun, 27 Jan 2008 11:31:48 +0000
To: [email protected]

On 26 Jan 2008, at 18:50, steve michel wrote:

Rob,

The fixture is a recessed fluorescent with a flat translucent lens;

Being recessed the fixture will have no uplight component. Did you
check the orientation the IES file will have in the scene? Typically
major output direction will point downwards (-z). For a downlight you
should not need any rotation, just lift (xform -t) the *.rad file to
the right height and place in your scene.

therefore my needs are simply to show a glow for the lens on the same
plane as the ceiling. I applied the ies distribution to a simple
polygon.

Did you create the polygon yourself? Ies2rad (without -i option!) will
do that and you don't need any further polygons. They will only
interfere
with you're distribution. The only problem is that the polygon will have
the size of the luminous area not the of the actual fitting.

Regards,
Thomas

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

_________________________________________________________________

Hi Thomas,

MARK THIS SOLVED

Good. But see below ...

I had created a 'glow' material and applying it to the same polygon in my ies.rad file. SO I got a render with my fixtures showing but still no illumination! I was trying to overlap or layer a glow material on a distribution's polygon. So you are right that the two might interfere. (But I can see where that might be useful to avoid the extra step of two distinct polygon geometry for the same thing (which I have done before with erco fixtures 3d geometry))

BUT The simpler method of deleting the -i option in ies2rad worked...I see the fixture and the light distribution...

I'm still confused about your way of using ies2rad. You'd only
want to use the '-i' option if you want to replace the crude IES
geometry with a geometry file you created before and you don't
want the geometry to interfere with the distribution. A recessed
fitting has no visible geometry except for a bright disc or rectangle
and that's already covered in the IES file.

The next challenge is to use a single light distribution material
and apply to multiple similar geometry in the scene...

If you want to use the same IES file a few times just use the
*.rad file that's created by ies2rad a few times in your scene.

Example:

First create the description for the light fitting:

## this is typed at the command line
$ ies2rad -m 0.8 -t default aero.ies

This will create 'aero.dat' and 'aero.rad'. 'aero.dat' is
a converted description of the light distribution. This file
never changes. The 'aero.dat' file contains the material
definition and the geometry of the luminous area. This geometry
is necessary to create light in your scene. Whenever you call
ies2rad with different options but with the same IES file as
argument it will create the same .dat file but the material
definition in the .rad file will change (brighter or darker,
more or less white).

Then position the aero.rad file in your scene as required:

## these lines are part of the scene description
!xform -t 0 0 2.4 aero.rad
!xform -t 0 2 2.4 aero.rad
!xform -t 0 4 2.4 aero.rad

This will place three fittings at 2.4m height spaces at 2m along
the y-axis. You can create simple arrays with xform in one line
but that's not as illustrative right now. All of these fittings
will have the same output and distribution.

If you want to change the maintenance factor (-m in ies2rad) or
the lamptype (-t) you have to create a second file and use that
instead of the first 'aero.rad'.

Another more advanced way is to use ies2rad as an inline command
in the scene file and use a 'pipe' to transform the geometry:

## these lines combine the two separate steps above
!ies2rad -s -m 0.80 -t default aero.ies | xform -t 0 0 2.4
!ies2rad -s -m 0.80 -t default aero.ies | xform -t 0 2 2.4
!ies2rad -s -m 0.80 -t default aero.ies | xform -t 0 4 2.4

Note the additional '-s' in the ies2rad command.

When you run these lines through oconv the 'aero.ies' file has
to be present in the directory or you will get an error. After
the conversion, there will be one 'aero.dat' file while all
the material definitions and light polygons have been passed
through xform instead of being written to a .rad file.

This method has the advantage that you don't have to create
a new 'aero.rad' file for every dimming value you want to
try. Just modify the '-m' argument:

## same fitting, different dimming values:
!ies2rad -s -m 0.80 -t default aero.ies | xform -t 0 0 2.4
!ies2rad -s -m 0.60 -t default aero.ies | xform -t 0 2 2.4
!ies2rad -s -m 0.40 -t default aero.ies | xform -t 0 4 2.4

I hope this helps to solve your problems.

Thomas

···

On 27 Jan 2008, at 14:42, steve michel wrote:

Hi,

I wanted to add a few comments on this.

The "-i <radius>" option to ies2rad creates a sphere of radius specified by -i. This sphere has a an illum type applied to it to produce light. The illum is modified by a brightfunc specifying the output distribution based on the converted photometry data. However there are a few things to note:

   1. an illum is invisible - however it gives off light, if you are
      having problems a good first test is to convert your ies data
      without the -i option so you will get a "light" material defined
      rather than an "illum" material
   2. ies2rad -i creates a sphere, you need to understand where the
      sphere is relative to other surface geometry such as a ceiling,
      you may have problems if your sphere cuts through other geometry
   3. an illum can specifiy an alternate material to show up in its
      place (however if you have a sphere defined via ies2rad where in
      fact there should be a ring then this will not look right):

        void glow my.glowing.source
        0
        4 1 1 1 0

        void brightfunc my.illum.distribution
        <.....>
        0
        1 1

        my.illum.distribution illum my.illum.source
        1 my.glowing.source
        0
        3 1 1 1

I have always advocated that whenever anyone is using ies photometric data, one should create a simple test box to place the output of ies2rad into in order to quickly view what is going on. This way you can see if the things are behaving the way they need and/or if they need additional transforms to get things in the correct location/orientation.

In the case where you want to create more detailed geometry representing the fixture and perhaps a glowing source, then the thing to do is to create this geometry and use a glow for the source surface and then either enclose this in an illum sphere or other enclosing geometry or float a quad or ring in front of the glow (assuming relatively directional output) and again apply the illum distribution. However be aware of what Rob G. has pointed out that there will be some distributions that will "cut off" dependent on view angle.

To deploy multiple instances of a light into a scene you can use xform as I think Thomas has outlined. This will work when thing are generally evenly spaced. If you have a more irregular deployment then you need to take a look at "replmarks" (man replmarks), which will replace a triangular marker with whatever you want (light or other kind of instanced geometry).

I hope this helps.

Regards,

-Jack de Valpine

steve michel wrote:

···

Hi Thomas,

MARK THIS SOLVED

I had created a 'glow' material and applying it to the same polygon in my ies.rad file. SO I got a render with my fixtures showing but still no illumination! I was trying to overlap or layer a glow material on a distribution's polygon. So you are right that the two might interfere. (But I can see where that might be useful to avoid the extra step of two distinct polygon geometry for the same thing (which I have done before with erco fixtures 3d geometry))

BUT The simpler method of deleting the -i option in ies2rad worked...I see the fixture and the light distribution...

The next challenge is to use a single light distribution material and apply to multiple similar geometry in the scene... To isolate the problem, I've been hand editing each fixture geometry and duplicating the same ies.rad parameters to each...

thanks to all
Steve

----------------------------------------
  

From: [email protected]
Subject: Re: [Radiance-general] Radiance, objline, objpict etc..
Date: Sun, 27 Jan 2008 11:31:48 +0000
To: [email protected]

On 26 Jan 2008, at 18:50, steve michel wrote:

Rob,

The fixture is a recessed fluorescent with a flat translucent lens;
      

Being recessed the fixture will have no uplight component. Did you
check the orientation the IES file will have in the scene? Typically
major output direction will point downwards (-z). For a downlight you
should not need any rotation, just lift (xform -t) the *.rad file to
the right height and place in your scene.

therefore my needs are simply to show a glow for the lens on the same
plane as the ceiling. I applied the ies distribution to a simple polygon.
      

Did you create the polygon yourself? Ies2rad (without -i option!) will
do that and you don't need any further polygons. They will only interfere
with you're distribution. The only problem is that the polygon will have
the size of the luminous area not the of the actual fitting.

Regards,
Thomas

_______________________________________________
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

--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

I wholeheartedly agree. I wrote a quick script (well actually it's just a modified copy of the objview script) to do this a few years ago; the link has been bused for a while now, but I just fixed it:

http://www.rumblestrip.org/downloads/ltview.tgz

ltview takes a radiance luminaire description (output from ies2rad) and places it centered in a 4' square box with one side missing, and fires up rvu (and cranks up -ds) to display the distribution. It has come in handy more than once. Here is a link to my original blog post announcing its availability and explaining its use:

http://www.rumblestrip.org/2004/05/10/ltview-a-radiance-utility

And be advised that it's geared to imperial units, so if you converted your ies file with meters in mind, you'll need to adjust the -bs parameter to scale your box down so it's tighter to the fixture. I think Aksel had problems with this when I first released it; setting -bs .63 should give you a similarly sized box for a metric-converted fixture.

- Rob Guglielmetti
www.rumblestrip.org

···

On Jan 27, 2008, at 5:31 PM, Jack de Valpine wrote:

I have always advocated that whenever anyone is using ies photometric data, one should create a simple test box to place the output of ies2rad into in order to quickly view what is going on. This way you can see if the things are behaving the way they need and/or if they need additional transforms to get things in the correct location/orientation.