3 ms·
So far, so good, but it seems to me like many others would be able to manufacture such calibration devices. I mean, the CIE colour space is known to everybody.
by Bromskloss 10y ago
So far, so good, but it seems to me like many others would be able to manufacture such calibration devices. I mean, the CIE colour space is known to everybody.
- cptskippy 10y agoThe problem with CIE is that there are no constraints to the colorspace at all. It's gamut range is every color imaginable by a person. All devices have different ranges of color they can reproduce and nothing can produce the full CIE colorspace. Even if we possessed a device with that capability, we lack the ability to describe all possible colors to that device. AdobeRGB and sRGB are subsets of CIE and computer monitor makers struggle to make professional grade monitors that can cover 90%-100% of those gamut ranges. The Pantone system is constrained further by just being a subset of colors exist within a limited gamut range. Pantone colors are only ones that can be produced by the mixing of, I think, 14 base colors. Their system sets clear limits on the scope of the problem and provides clear and consistent reference points.
- makomk 10y agoThe other problem is that thanks to metamerism, you can only use CIE for colour matching under a single fixed choice of illuminant - samples that appear to be the same CIE colour under (say) tungsten lighting may not be under florescent lighting or daylight.
- cptskippy 10y agoAnd through the rabbit hole of color management we climb...