I had the chance of “opening” up the SVG file using Adobe Illustrator and it
came out as a mosaic of “color squares”.
/ Roger
From: argyllcms-bounce@xxxxxxxxxxxxx <argyllcms-bounce@xxxxxxxxxxxxx> On Behalf
Of Wire ~
Sent: Thursday, October 15, 2020 2:08 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Looking for colorimetrically correct CIE 1931 xy image
I want to convey what the CIE 1931 xy (horseshoe diagram) color actually looks
like when viewed on modern displays.
The goal is twofold:
1) in whatever projection (color space) from which the diagram rendered,
in-gamut color is distributed accurately.
2) for color excerding CIE RGB, that the relationships in these zones are
illustrative of the excess at some logic.
For example, at Wikipedia there's an SVG that looks like it might be pretty
good:
https://commons.wikimedia.org/wiki/File:CIE1931xy_blank.svg
This image is a PNG derived from a corresponding SVG that has had some care put
into it.
It appears that it's not merely a hue-wheel within the area of the CIE RGB
triangle, it's has a subtle distribution. It shows the green/blue bulge of
Guild & Wright negative primary matches as artistically "cyan", and the line of
purples as a purply gradient from bluish to reddish. And there's a little nexus
near the blue primary that seems not to be in dispute for sRGB/709, Adobe RGB,
and DCI
One hitch in my synopsis is the assumption of web-color for the graphic. Just
for fun I ran the SVG through PS. It comes out filled with blocks. So I blurred
it, and assigned CIE RGB — What else, no? :)
I'm working on a web poster of the horseshoe diagram after reading this
A Beginners Guide to CIE
https://medium.com/hipster-color-science/a-beginners-guide-to-colorimetry-401f1830b65a
Which helped me see how the mappings evolved from the original experiments thru
maths to the diagram we now take for granted. This connects to wondering why so
little of the online discussion of color systems performance includes actually
showing what the systems are doing
regards
/wire