Hi guys,
I am not an expert with the radiance and i do not know if i am
repeating any simple stuff here.
Colour is a function of both the reflectivity and incident light.
Reflectivity itself is an intrinsic function and in this case i think
we are assuming lambertain( equal reflectance in all direction)
materials.
The amount of light that is reflected off a surface would be dependent
on the spectral shape of the incident light and the reflectance.
So in order to talk about the color properties of different object we
need to measure its reflectance ( use of a spectrophotometer).
Perhaps we can also include the type of incident light sources (
tungsten / sunlight) if we want to talk about reflectivity
Cheers
cw
···
On Fri, Nov 7, 2008 at 10:52 PM, David Smith <[email protected]> wrote:
Iebele,
I'm about as far removed from being a color expert as one can get. I
got the method from a thread in the forums for AGi32, and the method
has worked reasonably well for me (some tweaking required). I'm not
sure if the method is just a hack, but Ian Ashdown was the person who
suggested it, as well as some further reading on color science. Here's
the link, free registration required:
http://www.agi32.com/forums/view_topic.php?id=585&forum_id=4
Also, Konica-Minolta makes a luminance meter, CS-100A, that will give
you the XYZE values of what it's looking at, but I'm sure it costs a
small fortune. I think it's the one that Axel had just mentioned.
--Dave
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
Hi Mark,
I do not think a correct material results from using straight rgb values from the easyrgb site. I believe the thing to do is use the XYZ values and convert them. One option for conversion is the "Import" feature at:
http://www.luxal.eu/resources/radiance/cp4r.shtml
I have to assume that one of the xyz cal conversion functions is being used to do this.
Here is how to convert XYZ data using radiance (this is as I understand it):
FILE xyz.cal:
Rr = R(Xx,Yy,Zz)/100;
Gg = G(Xx,Yy,Zz)/100;
Bb = B(Xx,Yy,Zz)/100;
tx = recno * 10;
FILE xyz.in:
\(i\) {Xx} \{Yy\} {Zz}
FILE plastic.out:
void plastic \(i\)
0
5 {Rr} \{Gg\} {Bb} 0 0
!genbox \(i\) box 10 10 10 | xform \-t {tx} 0 0
FILE test.data:
bm.heritage.red 14.019 7.478 1.413
bm.cliffside.gray 57.06 60.837 63.055
bm.charcoal.gray 12.395 13.111 14.839
bm.merlot.red 16.828 11.710 5.175
bm.hearthstone.brown 8.643 8.368 6.745
bm.peacock.blue 27.583 34.415 44.796
bm.rosemary.green 22.552 31.374 10.008
bm.lightening.white 82.022 87.720 75.128
DO at command line:
rcalc -f ray/src/cal/cal/xyz_rgb.cal -f xyz.cal -i xyz.in -o plastic.out test.data > test.rad
objview test.rad
OR
rcalc -f ray/src/cal/cal/xyz_srgb.cal -f xyz.cal -i xyz.in -o plastic.out test.data > test.2.rad
objview test.2.rad
Now the one thing that I am not clear on is which file: xyz_rgb.cal or xyz_srgb.cal is the correct one to use and under what circumstances. I believe that the main difference is in the CIE primaries.
Best,
-Jack
Mark Stock wrote:
···
Jack,
Forgive my ignorance on the topic, but how do you convert those values to something that looks right in Radiance? I've tried just using RGB values like the ones from the site, but everything is simply too bright. And scaling them down uniformly seems to change the hue.
Mark
On Thu, 6 Nov 2008, Jack de Valpine wrote:
Hi,
Another resource is:
www.easyrgb.com
The focus is on paint colors. But they supply all kinds of color values including XYZ.
-Jack
Axel Jacobs wrote:
Sorry about the messed-up subject line in my last post...
---------- Forwarded message ----------
From: Axel Jacobs <[email protected]>
Date: 2008/11/6
Subject: Re: Radiance-general Digest, Vol 57, Issue 4
To: [email protected]
Erwin,
Anyway, as a start it might be worth thinking about a central web based material depository where we all can share our own creations. The display format can be a simple list with mat, cal, pic files and a little preview.
Anyone interested in starting that?
There is a site in Japan doing something rather similar:
http://arch.xtr.jp/radiance/mat_database.htm
Just pipe it through an on-line translator. There are a handful of
different pages. It seems that all examples are from freely available
example models and mailing list posts.
Axel
_______________________________________________
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
_______________________________________________
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
Hi All,
It looks like luxal is using a different formula for converting XYZ to RGB. I am not sure what the difference is but the values are quite different than the technique outlined in my previous post.
-Jack
Jack de Valpine wrote:
···
Hi Mark,
I do not think a correct material results from using straight rgb values from the easyrgb site. I believe the thing to do is use the XYZ values and convert them. One option for conversion is the "Import" feature at:
http://www.luxal.eu/resources/radiance/cp4r.shtml
I have to assume that one of the xyz cal conversion functions is being used to do this.
Here is how to convert XYZ data using radiance (this is as I understand it):
FILE xyz.cal:
Rr = R(Xx,Yy,Zz)/100;
Gg = G(Xx,Yy,Zz)/100;
Bb = B(Xx,Yy,Zz)/100;
tx = recno * 10;
FILE xyz.in:
\(i\) {Xx} \{Yy\} {Zz}
FILE plastic.out:
void plastic \(i\)
0
5 {Rr} \{Gg\} {Bb} 0 0
!genbox \(i\) box 10 10 10 | xform \-t {tx} 0 0
FILE test.data:
bm.heritage.red 14.019 7.478 1.413
bm.cliffside.gray 57.06 60.837 63.055
bm.charcoal.gray 12.395 13.111 14.839
bm.merlot.red 16.828 11.710 5.175
bm.hearthstone.brown 8.643 8.368 6.745
bm.peacock.blue 27.583 34.415 44.796
bm.rosemary.green 22.552 31.374 10.008
bm.lightening.white 82.022 87.720 75.128
DO at command line:
rcalc -f ray/src/cal/cal/xyz_rgb.cal -f xyz.cal -i xyz.in -o plastic.out test.data > test.rad
objview test.rad
OR
rcalc -f ray/src/cal/cal/xyz_srgb.cal -f xyz.cal -i xyz.in -o plastic.out test.data > test.2.rad
objview test.2.rad
Now the one thing that I am not clear on is which file: xyz_rgb.cal or xyz_srgb.cal is the correct one to use and under what circumstances. I believe that the main difference is in the CIE primaries.
Best,
-Jack
Mark Stock wrote:
Jack,
Forgive my ignorance on the topic, but how do you convert those values to something that looks right in Radiance? I've tried just using RGB values like the ones from the site, but everything is simply too bright. And scaling them down uniformly seems to change the hue.
Mark
On Thu, 6 Nov 2008, Jack de Valpine wrote:
Hi,
Another resource is:
www.easyrgb.com
The focus is on paint colors. But they supply all kinds of color values including XYZ.
-Jack
Axel Jacobs wrote:
Sorry about the messed-up subject line in my last post...
---------- Forwarded message ----------
From: Axel Jacobs <[email protected]>
Date: 2008/11/6
Subject: Re: Radiance-general Digest, Vol 57, Issue 4
To: [email protected]
Erwin,
Anyway, as a start it might be worth thinking about a central web based material depository where we all can share our own creations. The display format can be a simple list with mat, cal, pic files and a little preview.
Anyone interested in starting that?
There is a site in Japan doing something rather similar:
http://arch.xtr.jp/radiance/mat_database.htm
Just pipe it through an on-line translator. There are a handful of
different pages. It seems that all examples are from freely available
example models and mailing list posts.
Axel
_______________________________________________
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
_______________________________________________
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
_______________________________________________
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
Hi,
for converting spectral or XYZ data to RGB see Appendix in:
Ruppertsberg, A, and Bloj, M (2006) Rendering complex scenes for psychophysics
using RADIANCE: How accurate can you get?, J. Opt. Soc. Am. A, 23, 759-768
download pdf at:
http://www.brad.ac.uk/acad/lifesci/optometry/uploads/Publications/2006xJOPTSOCAMAx23x4x759.pdf
In that paper we also used spectral data from a data base (ref 47) of Munsell
chips and natural materials:
47. See http://spectral.joensuu.fi/databases/index.html\.
What I suspect you are after are surface reflectance descriptor RGB values. The
XYZ values provided by eg the macbeth color checker are colour signal data (ie
the product of surface reflectance with illumination). The illumination is most
probably D65, which is white, so it's not far off to correspond to the surface
reflectance. But a lot of people forget the distinction between colour signal
and surface reflectance.
Also remember that there are no generically valid RGB values. RGB values are
always only valid for a particular display device. In that paper up there we
have used the transformation matrix based on RADIANCE's 'built-in' primaries.
It is true that you can obtain RGB values beyond the [0 1] range when you
convert XYZ to RGB and that simply tells you that the gamut of your RGB-space
does not include that colour. Then, you have to do something, either clip it or
tone-map it....
happy colouring!
alexa
Quoting "Lars O. Grobe" <[email protected]>:
···
While I agree that extending the material file in the radiance
distribution would be useful for viz, I'd like to encourage those who
did not try yet to use their digital cameras and a calibration tool such
as macbethcal and the not too expensive chart as a measurement device.
Today as almost everyone has the equipment for digital photography, this
should be not too difficult. While one can certainly discuss the
accuracy, obtaining an "inaccurate" sample of the material that is
actually to be simulated should still be better then an accurate sample
of a generic material.
That said, the most important feature of a new material database must be
detailed documentation of the samples, measurement equipment, conditions.
CU Lars.
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
--
Dr. Alexa Ruppertsberg
Bradford School of Optometry and Vision Science
University of Bradford
Bradford, BD7 1DP
UK
Tel. ++44 1274 235378
------------------------------------------------------------
This mail sent through IMP: http://webmail.brad.ac.uk
To report misuse from this email address forward the message
and full headers to [email protected]
------------------------------------------------------------
Hi Axel,
Thank you so much for posting this file. I think both your RAL file as the Kodak grayscale chart measurements I send in my earlier mail are very usefull for Radiance.
In my previous mail I mistakenly wrote that we used a Kodak photospectrometer for the grayscale chart from Kodak's Q-14 color separation guide. That chart was also measured with a Minolta device, I don't remember exactly wich model it was - but it was a pretty advanced one that we hired for two weeks, so the file wich the greyscale measurements should be accurate also (which are YXZ values converted to Radiance RGB with a Radiance .cal file).
A list of all RAL colors in English (and other languages) can be found at:
http://www.ral.de/en/ral_farben/anwender/farbnamen_classic.php
I will be happy to translate it and send it back.
Maybe we can discuss a little bit more about the issue if such accurate measurements could be part of the Radiance distribution? I like to hear more about the pros and conts. My idea is that it would help users a lot.
-Iebele
Axel Jacobs wrote:
···
Jan Diepens of University of Technologie Eindhoven sent me a complete
list of RAL colours some time ago. I have to admit that they are still
not in the ColourPicker. I've put them up for your perusal:
http://luminance.londonmet.ac.uk/pickup/ral_mat.tgz
If anybody is going to translate the names to English, please bounce
the changes back to me for inclusion in the ColourPicker. The values
were measured with a Minolta chroma meter from samples. YXZ was then
converted to Radiance RGB with a Radiance .cal file, so the values
should be accurate.
Axel
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general
Radiance uses slightly different color primaries and a different (neutral) white point than the CCIR709 set that is the basis for sRGB (and other RGB) standards. This is historical, but not without reason, which is why you should stick with the xyz_rgb.cal file for back-and-forth conversions.
Best,
-Greg
···
From: Jack de Valpine <[email protected]>
Date: November 7, 2008 10:41:23 AM PST
Hi All,
It looks like luxal is using a different formula for converting XYZ to RGB. I am not sure what the difference is but the values are quite different than the technique outlined in my previous post.
-Jack
Hi Iebele,
Thank you so much for posting this file.
It's Jan Diepens who deserves our thank you.
A list of all RAL colors in English (and other languages) can be found at:
http://www.ral.de/en/ral_farben/anwender/farbnamen_classic.php
I will be happy to translate it and send it back.
It's already translated and integrated in the development version of
the Colour Picker. The link is in the main body of:
http://www.luxal.eu/resources/radiance/index.shtml
(don't use the navigation panel)
Cheers
Axel
Hi Axel,
The updates to the Colour Picker are great! I see that you updated the conversion of XYZ to Radiance rgb. This is a big help. I also like the addition of the instant reflectance value when using the sliders.
Thanks for all the work on this and all the other tools and information on the site.
Best,
-Jack
···
It's already translated and integrated in the development version of
the Colour Picker. The link is in the main body of:
http://www.luxal.eu/resources/radiance/index.shtml
(don't use the navigation panel)
--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction