Is our accepted color space wrong?

[etype2]

An article from Sound&Video Contractor. The industry may need to change if this new information is correct.

A new study identifies an important error in the 3D mathematical color space developed by the Nobel Prize-winning physicist Erwin Schrödinger and others, For more 100 years we used this mathematical space to describe how the eye distinguishes one color from another, and to build electronic color reproduction accordingly. Now it seems the underlying model may be faulty. On the bright side, the research has the potential to boost scientific data visualizations, recalibrate the textile and paint industries, and improve (or completely disrupt) how displays are made. If–science can figure out what the color space is, now that we know it’s not what we thought.

“The assumed shape of color space requires a paradigm shift,” said Roxana Bujack, a computer scientist with a background in mathematics who creates scientific visualizations at Los Alamos National Laboratory. Bujack is lead author of the paper by a Los Alamos team in the Proceedings of the National Academy of Sciences on the mathematics of color perception.

“Our research shows that the current mathematical model of how the eye perceives color differences is incorrect. That model was suggested by Bernhard Riemann and developed by Hermann von Helmholtz and Erwin Schrödinger—all giants in mathematics and physics—and proving one of them wrong is pretty much the dream of a scientist,” said Bujack.

“Our original idea was to develop algorithms to automatically improve color maps for data visualization, to make them easier to understand and interpret,” Bujack said. So the team was surprised when they discovered they were the first to determine that the longstanding application of Riemannian geometry, which allows generalizing straight lines to curved surfaces, didn’t work.


To create the industry standards that video professionals have depended on for everything, a precise mathematical model of perceived color space is needed. If you recall, Riemannian geometry plots red, green and blue in the 3D space. Those are the colors registered most strongly by light-detecting cones on our retinas–the all important RGB spectrum.

In the new study published the online journal PNAS; phys.org describes it as blending psychology, biology and mathematics, Bujack and her colleagues discovered that Riemannian geometry overestimates our perception of large color differences. People perceive a big difference in color to be less than the sum of the incremental differences that lie between two widely separated shades–a principle of diminishing returns. Color-driven painters like Kandisky understood this subjectivity and the relativity of perception. Not surprisingly that didn’t make it into the science that drives digital video; Riemannian geometry cannot account for it

“We didn’t expect this, and we don’t know the exact geometry of this new color space yet,” Bujack said. “We might be able to think of it normally but with an added dampening or weighing function that pulls long distances in, making them shorter. But we can’t prove it yet.”

No matter how subtle the necessary adjustments will need to be, in the world of color pixels the effect of a changed color space will be anything but subtle.

More information: Roxana Bujack et al, The non-Riemannian nature of perceptual color space, Proceedings of the National Academy of Sciences (2022). DOI: 10.1073/pnas.2119753119
Journal information: Proceedings of the National Academy of Sciences

It’s worth reading the paper even with all the math. https://www.pnas.org/doi/full/10.1073/pnas.2119753119

[old_tv_nut]

As the text says, this has to do with the perceptual difference between different colors. Important if you are trying to maximize differentiation of colors in a pie chart or signal lights,but totally irrelvant in choice of primary colors, which depends on the much more primitive level of matching the cone cell stimulations. Note that matching implies equality so no estimation of a difference is relevant.

I have also read speculation that no consistent color space is possible because color difference integration may be path dependent.

[old_tv_nut]

Color difference perception models generally do not have a strong effect on display design or choice of physical primaries, but can and do have effects on digital coding of color, for example, the optimum coding for high dynamic range images to prevent visible quantization.

[etype2]

Originally Posted by old_tv_nut

Color difference perception models generally do not have astrong effect on display design or coice of physical primaries, but can and do have effects on digital coding of color, for example, the optimum coding for high dynamic range images to prevent visible quantization.

Good to know. Thought the article was interesting. My wife and I sometimes argue on color selections and then I read that women have an additional cone that men don’t have or something to that effect. :-)

[ppppenguin]

The way we use 3 primaries is a basically a trick, though one that works very well in most cases. For example there's the essentially monochrormatic yellow/orange light from a sodium lamp. Or you can mix RGB primaries to match it. But many objects illuminated by one will look different when illuminated by the other. This is an extreme illustration of metamerism.

https://en.wikipedia.org/wiki/Metamerism_(color)

In your bathroom your white ceramic basin and resin bath may look the same in daylight but different in electric light. And different people may see the effect differently.

All our colour reproduction processes, whether additive (RGB on screens) or subtractive (CYMK in print) are fooling the eye to a greater or lesser extent. But it all works amazingly well most of the time.

[nasadowsk]

Photo film sees things differently based on the light hitting it. Hence the “daylight” and “tungsten” balanced films of years past…