Error: Rendering with Radiance p. 319 Luminaire Modeling

Hi Randoph,

Could you be more specific? The radius of a sphere whose *projected* area is 1.0 for any distant point is sqrt(1/pi) as Chas derived it. This is in fact what ies2rad needs. This is not the same as a sphere whose total surface area is 1.0, which is the value you derived.

Is there an error in ies2rad.c, somewhere?

-Greg

···

From: "Randolph M. Fritz" <[email protected]>
Date: February 6, 2012 3:59:10 PM PST

I just had occasion to go through Chas. Ehrlich's luminaire modeling procedure given in *Rendering with Radiance* and there is an error on page 319 of the text. The radius of a unit sphere is 0.282095, not 0.56419. Both the number in the text and the ies2rad command are wrong.

Ouch!
--
Randolph M. Fritz

Hi Randoph,

Could you be more specific? The radius of a sphere whose *projected* area is 1.0 for any distant point is sqrt(1/pi) as Chas derived it. This is in fact what ies2rad needs. This is not the same as a sphere whose total surface area is 1.0, which is the value you derived.

Is there an error in ies2rad.c, somewhere?

Then maybe I've done something else wrong. When I ran a simulation which checks the luminaire model against the IES file by simulating a goniophotometer the numbers lined up. I'll take a closer look at my scripts and see if I can find a factor of two somewhere.

Randolph

···

On 2012-02-07 00:12:18 +0000, Greg Ward said:

-Greg

From: "Randolph M. Fritz" <[email protected]>
Date: February 6, 2012 3:59:10 PM PST

I just had occasion to go through Chas. Ehrlich's luminaire modeling procedure given in *Rendering with Radiance* and there is an error on page 319 of the text. The radius of a unit sphere is 0.282095, not 0.56419. Both the number in the text and the ies2rad command are wrong.

Ouch!
--
Randolph M. Fritz

--
Randolph M. Fritz

Hi Randoph,

Could you be more specific? The radius of a sphere whose *projected*
area is 1.0 for any distant point is sqrt(1/pi) as Chas derived it.
This is in fact what ies2rad needs. This is not the same as a sphere
whose total surface area is 1.0, which is the value you derived.

Is there an error in ies2rad.c, somewhere?

Then maybe I've done something else wrong. When I ran a simulation
which checks the luminaire model against the IES file by simulating a
goniophotometer the numbers lined up. I'll take a closer look at my
scripts and see if I can find a factor of two somewhere.

The same check script worked correctly with an untouched ies2rad output file with both the given geometry and a spherical substitute.

I then ran the procedure with -i 0.56419 and the numbers I am getting back are the same and correct.

I am now thoroughly confused. Does the radius even matter in that procedure? Or do I have a case of a luminaire where it doesn't matter?

Randolph

···

On 2012-02-07 00:18:26 +0000, Randolph M. Fritz said:

On 2012-02-07 00:12:18 +0000, Greg Ward said:

Randolph

-Greg

From: "Randolph M. Fritz" <[email protected]>
Date: February 6, 2012 3:59:10 PM PST

I just had occasion to go through Chas. Ehrlich's luminaire modeling
procedure given in *Rendering with Radiance* and there is an error on
page 319 of the text. The radius of a unit sphere is 0.282095, not
0.56419. Both the number in the text and the ies2rad command are wrong.

Ouch!
--
Randolph M. Fritz

--
Randolph M. Fritz

Hi Randolph,

Whatever radius you give, ies2rad should adjust the brightness to keep the total lumen output the same.

Best,
-Greg

···

From: "Randolph M. Fritz" <[email protected]>
Date: February 6, 2012 4:45:53 PM PST

On 2012-02-07 00:18:26 +0000, Randolph M. Fritz said:

On 2012-02-07 00:12:18 +0000, Greg Ward said:

Hi Randoph,
Could you be more specific? The radius of a sphere whose *projected*
area is 1.0 for any distant point is sqrt(1/pi) as Chas derived it.
This is in fact what ies2rad needs. This is not the same as a sphere
whose total surface area is 1.0, which is the value you derived.
Is there an error in ies2rad.c, somewhere?

Then maybe I've done something else wrong. When I ran a simulation
which checks the luminaire model against the IES file by simulating a
goniophotometer the numbers lined up. I'll take a closer look at my
scripts and see if I can find a factor of two somewhere.

The same check script worked correctly with an untouched ies2rad output file with both the given geometry and a spherical substitute.

I then ran the procedure with -i 0.56419 and the numbers I am getting back are the same and correct.

I am now thoroughly confused. Does the radius even matter in that procedure? Or do I have a case of a luminaire where it doesn't matter?

Randolph

But then, why did Chas say to specify the radius of a projected unit sphere in RwR? The radius of the sphere does not seem to matter, at least in the case of the photometry I am working with. Does it make a difference with other sorts of photometry? Or…?

What am I missing?

Randolph

···

On 2012-02-07 01:40:29 +0000, Greg Ward said:

Hi Randolph,

Whatever radius you give, ies2rad should adjust the brightness to keep the total lumen output the same.

Best,
-Greg

From: "Randolph M. Fritz" <[email protected]>
Date: February 6, 2012 4:45:53 PM PST

On 2012-02-07 00:18:26 +0000, Randolph M. Fritz said:

On 2012-02-07 00:12:18 +0000, Greg Ward said:

Hi Randoph,
Could you be more specific? The radius of a sphere whose *projected*
area is 1.0 for any distant point is sqrt(1/pi) as Chas derived it.
This is in fact what ies2rad needs. This is not the same as a sphere
whose total surface area is 1.0, which is the value you derived.
Is there an error in ies2rad.c, somewhere?

Then maybe I've done something else wrong. When I ran a simulation
which checks the luminaire model against the IES file by simulating a
goniophotometer the numbers lined up. I'll take a closer look at my
scripts and see if I can find a factor of two somewhere.

The same check script worked correctly with an untouched ies2rad output file with both the given geometry and a spherical substitute.

I then ran the procedure with -i 0.56419 and the numbers I am getting back are the same and correct.

I am now thoroughly confused. Does the radius even matter in that procedure? Or do I have a case of a luminaire where it doesn't matter?

Randolph

--
Randolph M. Fritz

I guess you would have to ask Chas that one. Maybe he wanted the RGB values to come out (close to) 1.0 for some reason.

-Greg

···

From: "Randolph M. Fritz" <[email protected]>
Date: February 6, 2012 8:57:23 PM PST

But then, why did Chas say to specify the radius of a projected unit sphere in RwR? The radius of the sphere does not seem to matter, at least in the case of the photometry I am working with. Does it make a difference with other sorts of photometry? Or…?

What am I missing?

Randolph

On 2012-02-07 01:40:29 +0000, Greg Ward said:

Hi Randolph,
Whatever radius you give, ies2rad should adjust the brightness to keep the total lumen output the same.
Best,
-Greg

I think that is exactly right.

Thank you.

Randolph

···

On 2012-02-07 15:08:41 +0000, Greg Ward said:

I guess you would have to ask Chas that one. Maybe he wanted the RGB values to come out (close to) 1.0 for some reason.