# Luminance to RGB

Hello

I need to make the conversion of Luminance values (cd/m2) to the RGB colors to write-in Radiance...

Can anyone give me a hint of this basic question?

Steph

Hi, Steph,

Not sure if it's helpful but please refer to Axel Jacob's tutorial in page
40 (
).

According to Axel, the illuminance value (lux) of a point is calculated by
using the following formula:

I'm not sure if a Luminance value can be converted and split back into RGB

So, maybe the experts here can give you more clearer advices.

- Cheers, Ji

···

On Thu, Jul 7, 2011 at 11:57 PM, Art Cygne <artcygne@yahoo.com> wrote:

Hello

I need to make the conversion of Luminance values (cd/m2) to the RGB colors

Can anyone give me a hint of this basic question?

Steph

_______________________________________________

Yes, you basically invert this formula. If you have no color information, you can get a grayscale result by dividing your luminance by 179 and using the result for all three (RGB) channels. The pvalue program can produce a Radiance picture from raw input values, and the -b option can be used together with -r to take in a single radiance value for each pixel.

-Greg

···

From: Ji Zhang <hope.zh@gmail.com>
Date: July 7, 2011 9:36:56 AM PDT

Hi, Steph,

According to Axel, the illuminance value (lux) of a point is calculated by using the following formula:

I'm not sure if a Luminance value can be converted and split back into RGB radiance values.

So, maybe the experts here can give you more clearer advices.

- Cheers, Ji

On Thu, Jul 7, 2011 at 11:57 PM, Art Cygne <artcygne@yahoo.com> wrote:
Hello

I need to make the conversion of Luminance values (cd/m2) to the RGB colors to write-in Radiance...

Can anyone give me a hint of this basic question?

Steph

I hope I'm not interrupting the sense of the discussion, but I kind of have
the same question: when is about the properties
of materials (RGB plus roughness and specularity). If I want to simulate an
existing material in Radiance I need to
decompose it in the 3 primary colours.

As I understand Radiance uses Hemispherical Reflectance (ratio of flux
leaving the surface to the incident flux). I have this
data from the site...but how to get it to RGB in the case that I dont have
an spectrophotometer to measure. One way I know
could be using a grey scale chart, but then I also need to know under which
light source were those reflectances calculated
in order to compare them to my luminances. The objective of all this is to
obtain illuminance calculations.

Sorry again for the interruption...hope anybody has an answer.

Min.

Hi Chantal,

it would be great of we could point you to the book "Rendering with Radiance", which has a chapter on material data aquisition. Unfortunately it has been unavailable for a while, and you need a lot of luck to find an old copy in some library. Maybe you still try.

You question was not 100% clear to me as what data you have:

of materials (RGB plus roughness and specularity). If I want to simulate
an existing material in Radiance I need to
decompose it in the 3 primary colours.

You need that ONLY if you are interested in color in your simulation results, too. Photometric units are typically not aware of color (luminance, illuminance) as they are the weighted integral over the perceived wavelength range of light.

As I understand Radiance uses Hemispherical Reflectance (ratio of flux
leaving the surface to the incident flux). I have this
data from the site...but how to get it to RGB in the case that I dont
have an spectrophotometer to measure. One way I know
could be using a grey scale chart, but then I also need to know under
which light source were those reflectances calculated
in order to compare them to my luminances.

You should compare your grey chart with the sample under daylight or incadescent lighting condition. As long as you have a continous source spectrum, the grey chart should(!) perform rather stable, and unless you use a green LED to lid your sample you should get acceptable results. Keep in mind that the idea is not to get reflectance data in high accuracy ranges, but to find out whether your wall is bouncing back light more like 30 or more like 60% here...

You should take a grey chart with you. In many cases, there is measured data available by manufacturers, so it is always a good idea to take your notebook and write down any material/product information on site. One other way is to try using calibrated photographs (there is a tool called macbethcal coming with Radiance) if you have a color chart. This is useful if you need color and not just integrated reflectance. Still take the chart and write down a rough guess of the reflectance and compare to what you get from your calibration efforts.

The objective of all this is

to obtain illuminance calculations.

Again - for this, you typically need no color information, just reflectance - all surfaces will be grey, all sources white.

Cheers, Lars.

Just a head's up -- we are close to finally having print copies of the 1998(!) book available again, hopefully in time for people to order in advance of this year's workshop. Randolph Fritz has been working tirelessly on getting the PDF in shape for printing again, and has most of the short-run printing details worked out. We're just sorting a few details before soliciting orders.

Of course, we've promised this sort of thing before, but this time it looks like it's really going to happen.

Best,
-Greg

P.S. Colors may interact to give you different results than if you had used equivalent grays everywhere, but if you don't have color measurements or at least photographs, it's difficult to come up with good estimates of what the color should be. See the link:

and skip down to the "Tools for Materials" section.

···

From: "Lars O. Grobe" <grobe@gmx.net>
Date: July 11, 2011 7:46:37 AM PDT

Hi Chantal,

it would be great of we could point you to the book "Rendering with Radiance", which has a chapter on material data aquisition. Unfortunately it has been unavailable for a while, and you need a lot of luck to find an old copy in some library. Maybe you still try.

You question was not 100% clear to me as what data you have:

of materials (RGB plus roughness and specularity). If I want to simulate
an existing material in Radiance I need to
decompose it in the 3 primary colours.

You need that ONLY if you are interested in color in your simulation results, too. Photometric units are typically not aware of color (luminance, illuminance) as they are the weighted integral over the perceived wavelength range of light.

As I understand Radiance uses Hemispherical Reflectance (ratio of flux
leaving the surface to the incident flux). I have this
data from the site...but how to get it to RGB in the case that I dont
have an spectrophotometer to measure. One way I know
could be using a grey scale chart, but then I also need to know under
which light source were those reflectances calculated
in order to compare them to my luminances.

You should compare your grey chart with the sample under daylight or incadescent lighting condition. As long as you have a continous source spectrum, the grey chart should(!) perform rather stable, and unless you use a green LED to lid your sample you should get acceptable results. Keep in mind that the idea is not to get reflectance data in high accuracy ranges, but to find out whether your wall is bouncing back light more like 30 or more like 60% here...

You should take a grey chart with you. In many cases, there is measured data available by manufacturers, so it is always a good idea to take your notebook and write down any material/product information on site. One other way is to try using calibrated photographs (there is a tool called macbethcal coming with Radiance) if you have a color chart. This is useful if you need color and not just integrated reflectance. Still take the chart and write down a rough guess of the reflectance and compare to what you get from your calibration efforts.

The objective of all this is

to obtain illuminance calculations.

Again - for this, you typically need no color information, just reflectance - all surfaces will be grey, all sources white.

Cheers, Lars.

That's exciting news! The other day someone borrowed my sed/awk book without telling me, and while that was no big deal, if someone borrowed my RwR book without telling me I would have mounted a massive search, and once discovered, the borrower would never borrow from me again. I still have the Amazon receipt from my original purchase of this text. It's one of a very few books in my library that I consider irreplaceable, so It's nice to hear that print copies will be available once again!

- Rob

···

On Jul 12, 2011, at 10:05 AM, Greg Ward wrote:

Just a head's up -- we are close to finally having print copies of the 1998(!) book available again, hopefully in time for people to order in advance of this year's workshop. Randolph Fritz has been working tirelessly on getting the PDF in shape for printing again, and has most of the short-run printing details worked out. We're just sorting a few details before soliciting orders.

Of course, we've promised this sort of thing before, but this time it looks like it's really going to happen.

Best,
-Greg

P.S. Colors may interact to give you different results than if you had used equivalent grays everywhere, but if you don't have color measurements or at least photographs, it's difficult to come up with good estimates of what the color should be. See the link:

and skip down to the "Tools for Materials" section.

From: "Lars O. Grobe" <grobe@gmx.net>
Date: July 11, 2011 7:46:37 AM PDT

Hi Chantal,

it would be great of we could point you to the book "Rendering with Radiance", which has a chapter on material data aquisition. Unfortunately it has been unavailable for a while, and you need a lot of luck to find an old copy in some library. Maybe you still try.

You question was not 100% clear to me as what data you have:

of materials (RGB plus roughness and specularity). If I want to simulate
an existing material in Radiance I need to
decompose it in the 3 primary colours.

You need that ONLY if you are interested in color in your simulation results, too. Photometric units are typically not aware of color (luminance, illuminance) as they are the weighted integral over the perceived wavelength range of light.

As I understand Radiance uses Hemispherical Reflectance (ratio of flux
leaving the surface to the incident flux). I have this
data from the site...but how to get it to RGB in the case that I dont
have an spectrophotometer to measure. One way I know
could be using a grey scale chart, but then I also need to know under
which light source were those reflectances calculated
in order to compare them to my luminances.

You should compare your grey chart with the sample under daylight or incadescent lighting condition. As long as you have a continous source spectrum, the grey chart should(!) perform rather stable, and unless you use a green LED to lid your sample you should get acceptable results. Keep in mind that the idea is not to get reflectance data in high accuracy ranges, but to find out whether your wall is bouncing back light more like 30 or more like 60% here...

You should take a grey chart with you. In many cases, there is measured data available by manufacturers, so it is always a good idea to take your notebook and write down any material/product information on site. One other way is to try using calibrated photographs (there is a tool called macbethcal coming with Radiance) if you have a color chart. This is useful if you need color and not just integrated reflectance. Still take the chart and write down a rough guess of the reflectance and compare to what you get from your calibration efforts.

The objective of all this is

to obtain illuminance calculations.

Again - for this, you typically need no color information, just reflectance - all surfaces will be grey, all sources white.

Cheers, Lars.

_______________________________________________

is really helping me to put things into context.

The data I have is only luminance (cd/m2), from walls, ceiling and floor of
a real room. The intention is to simulate
that room in Radiance, and one of those walls in reality is orangish so in
this case color might matters as the reflectance
of the material will affect the lighting environment in the room. The aim is
obtaining illuminance levels in simulation
closer to reality (measurements I already took).

As far as I know, in order to describe the materials in a virtual model,
Radiance takes 3 values which correspond to RGB. This
is my question...how to reproduce real material properties in Radiance
(closest to reality).

I'll have a look to the method using the grey scale chart and using the
Macbeth ColorChecker Chart as you mentioned.
I highly appreciate that you pointed me in this direction...

But I'm pretty sure that I'll be here around with more questions in the near
future.

Thanks again!

Best Regards.

Hi Chantal!

The data I have is only luminance (cd/m2), from walls, ceiling and floor
of a real room. The intention is to simulate