Luminance nomenclature, STILL wanting to see constant L*
Luminance nomenclature, STILL wanting to see constant L*
- Subject: Luminance nomenclature, STILL wanting to see constant L*
- From: John <email@hidden>
- Date: Thu, 2 Nov 2006 00:37:23 -0800
Thank you for the reply,
"Luminance" apparently means different things in different contexts.
In a really great technical book, "A Field Guide to Digital
Color" (2004, pg 245), Maureen Stone says:
"In many graphics texts and tutorials, an equation derived from the
NTSC definition of luma (the Y of IYQ), is routinely recommended to
compute luminance for any arbitrary RGB triples. This is wrong. Not
only is this based on an obsolete set of phosphors, it confuses the
non-linear luma with luminance. (See chapter 6)."
In chapter 6, among other things, Maureen says:
It is important to understand that the brightness channel, Y in a non-
linear difference encoding is +not luminance+, even though the
notation is the same. It is luma, which is computed from non-linear
R, G, and B values. Some references, especially in computer graphics,
suggest using the luma equations from video encodings to compute
"luminance." While this generates a plausible grayscale image, true
luminance can only be computed from linear values of RGB of known
chromaticity."
Maureen goes on record saying that true luminance, based on the
physiological luminance efficiency curve, is precisely the L* of
CIELAB, and in the context of RGB, can only be computed from linear
values of RGB of known chromaticity.
Bruce Lindbloom's site (mentioned below), uses the word "luminance"
for Y (luma). His calculators work fine, as long as that nomenclature
is accounted for.
Regards,
John Sabin
P.S.
I STILL have not found a way to show all iMac-calibrated-display-
gamut colors of a given luminance (L*). Thanks to a hint from Bruce
Fraser, Photoshop color-picker ALMOST does it in LAB mode, by
choosing my calibrated monitor profile in photoshop "View->Proof
Setup", and then turning on Gamut warning in the color-picker with
"View->Gamut Warning". ALMOST, but instead of constraining the LAB
colors to those of my display gamut, the color-picker resets my Proof
Setup to "Working CMYK" (as can be seen in the "View->Proof Setup"),
so I'm still missing all those colors my monitor CAN show.
On Oct 31, 2006, at 8:09 PM, tom lianza wrote:
John,
You seem to be confusing the meaning of L* and luminance. They
are related, but certainly not equivalent. For instance an L* of 48
would correspond to a white normalized L of about 16.8 in a linear
system scaled 0 to 100. The example that you gave indicates that
the the Digital Color Meter is returning the value for a gamut
mapped color. The display probably cannot display a=127, b=-127
that you asked for so it has to be mapped into a displayable
color . It's clear that there are a number of color management
steps that are going on here.
With regard to your goal, there is a solution, but it involves a
number of numerical translation. Basically you need to map LAB->XYZ-
>RGB. Take a look a bruce lindblooms site: http://
www.brucelindbloom.com That would be a good place for you to play...
Tom Lianza
Video and Motion Picture Technology
X-rite corporation.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden