purpose of Granger rainbow chart?
purpose of Granger rainbow chart?
- Subject: purpose of Granger rainbow chart?
- From: Jacob Rus <email@hidden>
- Date: Sun, 7 Mar 2010 00:49:55 -0800
Hi everyone,
On a few usenet groups, a discussion of a “Granger chart” came up [1],
but I was left puzzled about its purpose. It seems to be also called a
"Granger rainbow", and there's an explanation of how to make one at
the luminous landscape site [2]. Quite aside from the inaccuracy of
the text on that site (“it contains colours which no display device,
from screens to printers, can completely display properly” is just
false), I couldn't figure out what the point of the Granger chart is.
Basically it shows the maximally chromatic color for each combination
of HSL/HSV hue, and Rec. 601 Luma (i.e. Y' = .30R' + .59G' + .11B'),
for whatever RGB working space is in use (Photoshop uses the same
calculation for "luminosity" regardless of RGB space [3]). A few
places around the web showed this being used with Photoshop's "gamut
warning" to compare color spaces, but this strikes me as pretty
misleading: push in the RGB primaries by a couple percent and suddenly
the whole chart is going to go gray, even though the overall gamut
remains pretty similar. I suppose it could be used to show the
boundaries at which a CMYK or RGB gamut is larger, as just a binary
test for each hue/luma. It's not especially good for choosing colors
(e.g. for design purposes), because (a) at high chroma, the three
components hit different parts of the gamma curve, and luma
calculations go pretty wrong, and (b) the hue spacing of whatever RGB
space is in use has pretty little to do with perceptual hue spacing.
Anyway, I'd be glad to hear what this thing is for, or if it should
just be ignored as the nonsense it initially appears to be.
Thanks,
Jacob Rus
P.S. the second test chart on that luminous landscape page, with L* =
100, and a* and b* varying, seems even much more pointless, since only
one point is even well defined.
[1] the context was a successful troll attempt (leading to much
backbiting and recrimination), and is pretty irrelevant; I'll skip the
link.
[2] http://www.luminous-landscape.com/essays/test-charts.shtml
[3] http://lists.apple.com/archives/colorsync-users/2001/Sep/msg00488.html
_______________________________________________
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