IES file fun

I was creating a light source the other day, from an IES file. Turns out this was to be my first encounter with an IESNA LM-63-95-formatted file, on a directional, linear fixture. Well, the first time, using Radiance that is.

I ran the *non*-directional downlight file thru ies2rad and then checked it in ltview, and it looked like this:

http://www.rumblestrip.org/site-img/rad/linear.jpg

So far so good.

Then I ran the *directional* downlight file thru ies2rad and checked it in ltview, and it looked like this:

http://www.rumblestrip.org/site-img/rad/assym.jpg

Eureka! I have defied physics! It was after looking at this image that I re-checked the ies file, and discovered that it's the LM-63-95 spec. This is a problem, because the spec calls for the 0-180 vertical plane to be oriented parallel to the lamp axis, whereas it's more common to orient it perpendicular. ies2rad can't even have a check for this, because the spec is so loosely written. This is why checking the distribution of my converted luminaires is so important to me, prior using them in a calc. I always did it because I don't trust myself, but clearly too much faith in the quality of your input files is also a bad thing.

Anyhoo, I simply reversed the length and width values for the luminous opening parameters in the ies file, and got this:

http://www.rumblestrip.org/site-img/rad/assym90.jpg

Now, I have a useable fixture, albeit one with a default plan orientation that is 90 degrees off from the norm. I guess I could "wrap" the whole thing in an xform file so that I could use it in a standard library.

Is there a better way to "correct" a funky IES file for ies2rad? What do you all do when you encounter this?

- Rob G.

Il 11Jun 16:08, Rob Guglielmetti scrisse:

I was creating a light source the other day, from an IES file. Turns

First, are you sure the ies data are correct?

[cut]

http://www.rumblestrip.org/site-img/rad/linear.jpg

So far so good.

It seems a symmetric fitting.

Then I ran the *directional* downlight file thru ies2rad and checked it
in ltview, and it looked like this:

http://www.rumblestrip.org/site-img/rad/assym.jpg

This time it seems an asymmetric fitting. What you meen with
*directional*?

Eureka! I have defied physics! It was after looking at this image that
I re-checked the ies file, and discovered that it's the LM-63-95 spec.
This is a problem, because the spec calls for the 0-180 vertical plane
to be oriented parallel to the lamp axis, whereas it's more common to
orient it perpendicular. ies2rad can't even have a check for this,

It's right, the 0-180 plane normally must be oriented parallel to the
lamp axis. This is a general rule to point right the fitting.

The IES specifies:

[...]
This integer value indicates the type of photometric web used for the
photometric measurements as follows:

  Value Description

    1 Type C photometry
    2 Type B photometry
    3 Type A photometry

Luminaires are photometered by locating the luminaire at the center of
an imaginary sphere and measuring the light intensity (candela) values
at grid points (the "photometric web') on the sphere's surface. The
orientation of the luminaire's axes relative to that of the sphere
determines the photometric type.

Type C photometry is normally used for architectural and roadway
luminaires. The polar axis of the photometric web coincides with the
vertical axis of the luminaire, and the 0-180 degree photometric plane
coincides with the luminaire's major axis (length).

Type B photometry is normally used for adjustable outdoor area and
sports lighting luminaires. The polar axis of the luminaire coincides
with the minor axis (width) of the luminaire, and the 0-180 degree
photometric plane coinicides with the luminaire's vertical axis.

Type A photometry is normally used for automotive headlights and signal
lights. The polar axis of the luminaire coincides with the major axis
(length) of the luminaire, and the 0-180 degree photometric plane
coinicides with the luminaire's vertical axis.
[...]

Normally the major axis in the type C (the most used) coincides whit
the lengt of the lamps. If the lamp is vertical in the fitting, like a
downlight, the luminous flux is normaly simmetric so is not important
which is the 0-180 plane.

If the fitting has an asymmetric reflector, then the direction of the
max flux coincides at 90� (0� 3 o'clock, 90� 12 o'clock)

because the spec is so loosely written. This is why checking the

I have a good text about the IESNA Photometric Data File (about 33K).
If you need it, could I attach it to the ML?

[cut]

Is there a better way to "correct" a funky IES file for ies2rad? What do
you all do when you encounter this?

If the rad is wrong most likely you have a bad IES file.

Ciao

···

--
Stefano Callegari <[email protected]>
Omnilux Srl
Via Frassanedo, 2 - I 35020 Villatora di Saonara (PD)
+39 049 8792281
GnuPG Public Key Server: pgp.mit.edu

Ciao Stefano!

First, are you sure the ies data are correct?

It looks fine to me in two different photometric file viewers, as well as in plain text...

It seems a symmetric fitting.

It is. No problems there. That was just for example.

http://www.rumblestrip.org/site-img/rad/assym.jpg

This time it seems an asymmetric fitting. What you meen with
*directional*?

Sorry, I mean exactly the same thing, asymmetric.

Eureka! I have defied physics! It was after looking at this image that
I re-checked the ies file, and discovered that it's the LM-63-95 spec.
This is a problem, because the spec calls for the 0-180 vertical plane
to be oriented parallel to the lamp axis, whereas it's more common to
orient it perpendicular. ies2rad can't even have a check for this,

It's right, the 0-180 plane normally must be oriented parallel to the
lamp axis. This is a general rule to point right the fitting.

Depends on which version of the IES spec you reference, and this is the whole problem. There is a good bit about it in Ian Ashdown's Lightfair presentation "Thinking Photometrically". A link to the course notes is at the bottom of the page referenced by this url (there is also an excellent summary of the IES file format in there): http://www.helios32.com/resources.htm

I have cribbed from these course notes the following section 3.6.2, "Luminaire Orientation and Position":

"Given a luminaire with an asymmetric luminous flux distribution (such as a fluorescent wall-washer), it is essential to orient the luminaire correctly in a CAD model. Unfortunately, this can be more difficult than you might expect.

"The first problem is that the various IESNA LM-series documents provide contradictory specifications on how the photometric web is to be oriented with respect to the physical outline of a luminaire. For example, IESNA LM-63-95 implies (but does not specify) that the 0–180 degree vertical plane of the photometric web is oriented parallel to the lamp axis of linear fluorescent luminaires. If the luminous flux distribution is bilaterally symmetric about the 90–270 degree vertical plane (that is, perpendicular to the lamp axis), then the horizontal angles must be reported from 90 to 270 degrees.

"However, IESNA LM-41-98 recommends (but does not specify) that the photometric web for such luminaires be oriented perpendicular to the lamp axis, with 0 degrees being the “beam side.” Most (but not all) fluorescent lighting manufacturers have ignored IESNA LM-63-95 and followed the recommendation of IESNA LM-41-98. This means that the lighting design software program user must manually examine the IESNA LM-63 text file to see whether the photometric web is oriented parallel or perpendicular to the lamps axis. If this information is not stated in the file header, the user may have to contact the luminaire manufacturer."
   - From Ian Ashdown's "Thinking Photometrically"

So, I guess we could call it a faulty file, but really it seems to me to be a difference in opinion as to how the photometric web is supposed to be arranged, between the manufacturer that supplied this IES file, and the IES spec. Or am I missing something (else)?

-Rob G.

···

On Jun 11, 2004, at 6:51 PM, Stefano Callegari wrote:

Rob Guglielmetti wrote:

IESNA LM-63-95 implies (but does not specify) that the 0?180
degree vertical plane of the photometric web is oriented parallel to
the lamp axis of linear fluorescent luminaires.

As I understand my copy of the standards document, this is not
just "implied" but clearly spelled out. At least the phrasing
left me with no doubt about the intended goal (I'm not at the
office, so I can't cite the original right now). nevertheless,
I consider this requirement a conceptual defect in the
specification.

The practical result is that with asymmetric linear fixtures, the
C0 plane will show the least interesting curve of the complete
dataset. For this reason, many manufacturers "cheat". They create
their files so that the width is given a larger value than the
length. That way, they can place the most interesting
(asymmetric) curve in the C0 plane, without violating the letter
of the standard (of course, they still violate its intention).

Most software reading IES files knows about those tricks, and
will apply the necessary corrections. I do the same in Rayfront,
but I didn't check if ies2rad does. In either case, I don't know
if this is the reason for the troubles described here, but it
might...

FWIW, the Eulumdat format specifies the orientation of the C0
plane in the sensible way, so that it will contain the most
interesting curve for asymmetric distributions.

-schorsch

···

--
Georg Mischler -- simulations developer -- schorsch at schorsch com
+schorsch.com+ -- lighting design tools -- http://www.schorsch.com/

I can attest that ies2rad doesn't know about these "tricks," as I didn't know about these tricks. I don't think I'll have time to fix it in the immediate future, and given that I still don't understand the subtleties involved, I wouldn't want to attempt a quick hack in this case. It sounds like Schorsch is definitely onto something, though...

-Greg

···

From: Georg Mischler <[email protected]>
Date: June 12, 2004 3:49:18 AM PDT

Rob Guglielmetti wrote:

IESNA LM-63-95 implies (but does not specify) that the 0?180
degree vertical plane of the photometric web is oriented parallel to
the lamp axis of linear fluorescent luminaires.

As I understand my copy of the standards document, this is not
just "implied" but clearly spelled out. At least the phrasing
left me with no doubt about the intended goal (I'm not at the
office, so I can't cite the original right now). nevertheless,
I consider this requirement a conceptual defect in the
specification.

The practical result is that with asymmetric linear fixtures, the
C0 plane will show the least interesting curve of the complete
dataset. For this reason, many manufacturers "cheat". They create
their files so that the width is given a larger value than the
length. That way, they can place the most interesting
(asymmetric) curve in the C0 plane, without violating the letter
of the standard (of course, they still violate its intention).

Most software reading IES files knows about those tricks, and
will apply the necessary corrections. I do the same in Rayfront,
but I didn't check if ies2rad does. In either case, I don't know
if this is the reason for the troubles described here, but it
might...

FWIW, the Eulumdat format specifies the orientation of the C0
plane in the sensible way, so that it will contain the most
interesting curve for asymmetric distributions.

-schorsch

Methinks you are right. I will have to look at the file(s) tomorrow when I get back into the office.

···

On Jun 12, 2004, at 6:49 AM, Georg Mischler wrote:

Most software reading IES files knows about those tricks, and
will apply the necessary corrections. I do the same in Rayfront,
but I didn't check if ies2rad does. In either case, I don't know
if this is the reason for the troubles described here, but it
might...

=================
    Rob Guglielmetti
www.rumblestrip.org