Obtaining RGB output data

hi all,

   I'm starting to use Radiance and I want to work with it in order to
simulate both daylight and artificial light. I want to change both the
RGB propeties of the light source and the materials. After Radiance
analizes the scene one can obtain as the output data, the luminance of
any point I select in my scene. Besides knowing what is the luminance
in any desired point, I'm very interested in knowing what is the RGB
component also of any point of the scene (such as if I were measuring
the RGB spectral component in several points of a room with a
spectometer). Is that possible? And if yes, what do you recommend me to
do in order to obtain this as an output data?

   I would appreciate any help.

Than you,
-adriana

Hi Adriana,

if I'm not mistaken you want to look into the use of 'pvalue'.
Note, that RGB values will give you some idea about the colorimetric properties of any point in the scene, but if you are remotely interested in accurate data - your comparison with the spectroradiometer suggests this - then you might want to look into spectral rendering. You can also do that with RADIANCE. see e.g. Yang & Maloney, 2001, Vision Research, 41:2581-2600 (the appendix).
As far as I remember RADIANCE will return RGB values < 0 and > 1, and it's up to you how you interpret this data, i.e. outside of the RGB-gamut of the default phosphors (also somewhere defined in RADIANCE), which will definitely become an issue with day-light scenes.

Good luck!

alexa

[email protected] wrote:

···

   hi all,

   I'm starting to use Radiance and I want to work with it in order to
simulate both daylight and artificial light. I want to change both the
RGB propeties of the light source and the materials. After Radiance
analizes the scene one can obtain as the output data, the luminance of
any point I select in my scene. Besides knowing what is the luminance
in any desired point, I'm very interested in knowing what is the RGB
component also of any point of the scene (such as if I were measuring
the RGB spectral component in several points of a room with a
spectometer). Is that possible? And if yes, what do you recommend me to
do in order to obtain this as an output data?

   I would appreciate any help.

Than you,
-adriana

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

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Alexa I. Ruppertsberg
Department of Optometry
University of Bradford
Bradford
BD7 1DP
UK

email: [email protected]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Hi Alexa!

Thanks for responding my email. I've been `playing´ with the "pvalue"
command and I assume I obtain the x-y position of the pixel of any picture
and the R, G, and B value of that pixle. You told me in your e-mail that
is up to me how I interpret this RGB data. When I use the pvalue I obtain
data like this (for example):

   xposition y position R G B

      345 240 7.495e-02 7.495e-02 3.495e-02

However, which type of units are this? Are there no units at all? Is this
illuminance? Also, in most of my pixels, the R,G, and B data is the same,
i.e, 7.495e-02 in the three (RGB). Does this means that that pixel has a
neutral color?

Thanks for your help,

Adriana

···

Hi Adriana,

if I'm not mistaken you want to look into the use of 'pvalue'.
Note, that RGB values will give you some idea about the colorimetric
properties of any point in the scene, but if you are remotely interested
in accurate data - your comparison with the spectroradiometer suggests
this - then you might want to look into spectral rendering. You can also
do that with RADIANCE. see e.g. Yang & Maloney, 2001, Vision Research,
41:2581-2600 (the appendix).
As far as I remember RADIANCE will return RGB values < 0 and > 1, and
it's up to you how you interpret this data, i.e. outside of the
RGB-gamut of the default phosphors (also somewhere defined in RADIANCE),
which will definitely become an issue with day-light scenes.

Good luck!

alexa

[email protected] wrote:

   hi all,

   I'm starting to use Radiance and I want to work with it in order to
simulate both daylight and artificial light. I want to change both the
RGB propeties of the light source and the materials. After Radiance
analizes the scene one can obtain as the output data, the luminance of
any point I select in my scene. Besides knowing what is the luminance
in any desired point, I'm very interested in knowing what is the RGB
component also of any point of the scene (such as if I were measuring
the RGB spectral component in several points of a room with a
spectometer). Is that possible? And if yes, what do you recommend me to
do in order to obtain this as an output data?

   I would appreciate any help.

Than you,
-adriana

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

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Alexa I. Ruppertsberg
Department of Optometry
University of Bradford
Bradford
BD7 1DP
UK

email: [email protected]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

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

Hi Adriana,

Unless you use the -o option of pvalue, the output is not in any particular units. With the -o option, the reported RGB values are in radiance units, i.e., watts/sr/meter^2. However, it is probably better for you to work directly in photometric units, which you can easily do with the ra_xyze command, like so:

% ra_xyze rendered.pic | pvalue -o [other options]

The "RGB" values reported will then be in calibrated CIE XYZ coordinates (i.e., the Y channel will be luminance in cd/meter^2).

The only caveat to the above is if the pictures are rendered using the -i option of rpict, then the output will be in units of irradiance (watts/meter^2) in the first case, and illuminance (lux) in the case where you apply ra_xyze.

Best of luck,
-Greg

···

From: [email protected]
Date: September 8, 2005 3:19:48 PM PDT

Hi Alexa!

Thanks for responding my email. I've been `playing´ with the "pvalue"
command and I assume I obtain the x-y position of the pixel of any picture
and the R, G, and B value of that pixle. You told me in your e-mail that
is up to me how I interpret this RGB data. When I use the pvalue I obtain
data like this (for example):

   xposition y position R G B

      345 240 7.495e-02 7.495e-02 3.495e-02

However, which type of units are this? Are there no units at all? Is this
illuminance? Also, in most of my pixels, the R,G, and B data is the same,
i.e, 7.495e-02 in the three (RGB). Does this means that that pixel has a
neutral color?

Thanks for your help,

Adriana