Does/has anyone used the CIECAM02 model in post?

[xpatUSA]

It's quite complicated:

https://en.wikipedia.org/wiki/CIECAM02

A good few sliders involved with some of them labeled quite similarly.

But if you were commissioned to produce an image to be displayed on a TV in a dark space it can apparently account for that - and more.

I had a quick play without much idea of what does what:



Did it in RawTherapee.

Comments from cognoscenti would be most welcome ...

[Manfred M]

Interesting academically in terms of trying to describe colours and human perception, but too complex to be of any practical use in working an image. Trying to work with six variables at the same time; brightness (luminance), lightness, colorfulness, chroma, saturation, and hue is too much for my simple mind.

I have enough issues when working in the CMYK colour space with its four variables and a C component that needs to be pushed harder than the M and Y channels for an equivalent effect.

Based on your demo and its overly heavy application of magenta, I don't see how that works. My understanding is that work for TV output is generally done in the sRGB colour space and even there, most flat screen televisions can't even handle that.

I just finished setting up a new OLED TV (our 15 year old unit died a couple of weeks ago) and given the output on most settings, colour fidelity is not high on the mind of most people watching TV.

[xpatUSA]

Interesting academically in terms of trying to describe colours and human perception, but too complex to be of any practical use in working an image. Trying to work with six variables at the same time; brightness (luminance), lightness, colorfulness, chroma, saturation, and hue is too much for my simple mind.

My thought was to put some effort into learning more about the variables with a view to working up a sidecar file or two for certain kinds of images.

I have enough issues when working in the CMYK colour space with its four variables and a C component that needs to be pushed harder than the M and Y channels for an equivalent effect.

que pobrecito. Can you imagine trying to cope with today's discovery -an artist's color wheel?

https://www.handprint.com/HP/WCL/vwheel.html

if you were commissioned to produce an image to be displayed on a TV in a dark space it can apparently account for that.

Based on your demo and its overly heavy application of magenta, I don't see how that works. My understanding is that work for TV output is generally done in the sRGB colour space and even there, most flat screen televisions can't even handle that.

I just finished setting up a new OLED TV (our 15 year old unit died a couple of weeks ago) and given the output on most settings, colour fidelity is not high on the mind of most people watching TV.

My choice of a TV was poor example and was not aimed at the experience of "people watching TV". I should have made that more clear. Possible more relevant is one of those TV's or monitors that hang on a wall and cycle though pictures. After reading more, the surrounding levels of darkness can also apply inside the image itself. For example the good old flower on a light or dark background.

Your somewhat negative response has made me wonder what the purpose of CIECAM is and why some editors include it, though. I need to learn more for my own interest

[Manfred M]

Your somewhat negative response has made me wonder what the purpose of CIECAM is and why some editors include it, though. I need to learn more for my own interest

That was really the gist of my comment. I have experience working in both RGB and CMY colours; the latter in chromogenic printing; the K in CYMK only applies in the printing process because of the properties of "real" ink on "real" paper.

I've done quite a bit of work using the L*a*b* colour space (mostly to work in the L* channel as the a* and b* channels are not intuitive). I use HSL in a lot of my work, when dealing with saturation issues in highly localized areas.

I not quite sure what CIECAM02 does that other colour models cover, outside of the research / academic side of things.

[xpatUSA]

That was really the gist of my comment. I have experience working in both RGB and CMY colours; the latter in chromogenic printing; the K in CYMK only applies in the printing process because of the properties of "real" ink on "real" paper.

I've done quite a bit of work using the L*a*b* colour space (mostly to work in the L* channel as the a* and b* channels are not intuitive). I use HSL in a lot of my work, when dealing with saturation issues in highly localized areas.

I not quite sure what CIECAM02 does that other colour models [do not] cover, outside of the research / academic side of things.

As to what other color models do not cover, I read this today:

Because they are limited to the trichromatic outputs from the L, M and S cones, both CIELUV and CIELAB only model the color information generated by the photoreceptor response to small and isolated color samples. They cannot model many of the contextual effects of color perception, including luminance adaptation (surround illuminance levels), chromatic contrast or chromatic assimilation that arise in real world viewing conditions with heterogeneous, strongly contrasted, or three dimensional color stimuli.

source: https://www.handprint.com/HP/WCL/color7.html (scroll about 2/3 down)

I'm leaning toward the view that CIECAM allows image adjustment to match intended viewing conditions and media of many, if not all kinds.

From that POV, I could develop a sidecar file to match my own known conditions but which will not specifically match anyone else's. Possibly CIECAM would still good for a print in a bricks-and-mortar gallery or images in someone's fixed-location digital picture-frame.

About the only other use might be in keeping perfect rendition of a smallish image on a widely different lightness or hue digital canvas since the surroundings and the background do affect color appearance.