8 ms·

New study overturns 100-year-old understanding of color perception

226 points by beefman 4y ago

user- 4y ago
403
EarlKing 4y ago
Well this is a first for me. I don't think I've ever seen a paper published to PNAS that wasn't open access.
- tingletech 4y ago
  embargo should expire October 29, 2022: "PNAS is a delayed open access journal, with an embargo period of six months that can be bypassed for an author fee (hybrid open access). " https://en.wikipedia.org/wiki/Proceedings_of_the_National_Academy_of_Sciences_of_the_United_States_of_America https://en.wikipedia.org/wiki/Proceedings_of_the_National_Ac...
  - fnordpiglet 4y ago
    LOL I came to the comments to find the paper
  - EarlKing 4y ago
    Huh. Well today I learned. I always thought it was just plain open access.
- collegeburner 4y ago
  hopefully somebody put it on scihub soon
ggm 4y ago
I've always found the printed book approximations of how people see colour fascinating. The amount of work a printer has to go through to achieve colour fidelity printing a continuously varying hue field in some specified geometry.. Now imagine doing this before PDF/Postscript.
AdamH12113 4y ago
When they talk about Riemannian geometry, is that related to the Riemann curvature tensor from general relativity? What’s the significance of being non-Riemannian? Is it just that (as the abstract says) “the distance between two colors is [not] the length of the shortest path that connects them.”?
- adrian_b 4y ago
  A point in the color space is determined by 3 numbers. Given 2 colors, i.e. 2 points in the color space, one can define a distance function from them to a number corresponding to the perceived color difference. One can try various formulas for the color distance and compare them in experiments with humans who must assess which colors are more similar or more different. Such experiments are extremely time consuming and require many test subjects, to average over individual variability. The problem is that a formula for a very good approximation of the color differences over the entire color space has not been found yet, even if there are many formulas that give acceptable results in certain cases, usually when the color differences are small. The Euclidean distance does not give a good approximation, so the color space is not Euclidean. The distances corresponding to various Riemannian spaces (i.e. "curved" spaces) give better approximations, but which are still not good enough. In fact the color distance is not even a distance in the mathematical sense, i.e. it does not satisfy the triangle axiom, so the color space is not a metric space. Perhaps there exists a non-linear transformation from the color space to a metric space.
  - hilbert42 4y ago
    "Riemannian spaces (i.e. "curved" spaces) give better approximations, but which are still not good enough." Right, as I said above this work shows we've nonlinear effects at work, whilst this was obvious in the past we should have taken more notice of the fact. Whilst Riemann provides say an ideal mathematical model (the ideal case) as we obserce Nature doesn't always comply, it insists on adding 'distortions' of its own. For the same reason these 'transfer' curves have nonlinear characteristics it seems other significant interactions are at work, namely cross-coupling between color channels. We see this in color film emulsions where color bleeds from one layer to another and this effect is also nonlinear (if the bleed were linear across the transfer curve then we could simply bias it out but it's not thus in practice it's nigh on impossible to correct). In effect this cross-coupling is essentially equivalent to cross-modulation distortion in electronic circuits and the mixing can be modeled mathematically in a similar way. No doubt it remains to be seen whether this new work has any significant bearing on this or not.
canjobear 4y ago
I thought it was already known that perceived large differences in colors were not geometrically consistent with small differences. For example CIECAM02-UCS is based on a compromise between fitting the small-difference data and large-difference data.
- adrian_b 4y ago
  Yes, from the abstract of the article it is impossible to understand what is new in this study. The "non-Riemannian nature of perceptual color space" has been known for decades. For example the CIE color difference formulas from 1994 and from 2000 (CIEDE2000) are older attempts to model the color space, taking this into account, i.e. that the color space is not even a metric space (the triangle axiom is not true). There are various newer attempts to better model the color space. Without being able to read the paper, I assume that it might refer to one such better model, but as I have said, neither the title nor the abstract give any clue about what is novel in it. They say "Rethinking them outside of a Riemannian setting could provide a path to extending them to large differences". From the "could provide a path" I understand that they have not found yet any formula valid for large color differences, which would have been something novel.
  - OscarCunningham 4y ago
    Can you give an example of three colours where the triangle inequality is false? Would be interesting to look at.
    - adrian_b 4y ago
      One paper where some such experiments are discussed and which includes some color diagrams is: https://psyarxiv.com/vtzrq/download?format=pdf https://psyarxiv.com/vtzrq/download?format=pdf The violations of the triangle inequality are caused, as others have also mentioned, both by the nonlinearity of the dependence between the perceived intensity of a color and its corresponding radiant flux and by the interactions between the 3 color channels. Therefore, while the 3 RGB or XYZ values are enough for being able to recognize if any pair of colors are the same or different, to be able to recognize whether in a set of 3 colors the 3rd is closer to the 1st or to the 2nd, one needs to apply some non-linear function of 3 arguments, providing 3 output numbers, in order to map the RGB/XYZ space to a space where a simple distance formula, e.g. the Euclidean distance or the Manhattan distance, can be used. Alternatively the non-linear 3 to 3 transform can be combined with the distance formula in the transformed space into a non-linear complicated distance formula that can be used directly in some applications, on RGB/XYZ colors, but separating the space mapping transform can be useful in some other applications, e.g. in color interpolation (though color interpolation can also be done based on only a distance formula, by solving an equation for each intermediate point, which might be not slower than mapping the extreme points, interpolating in the uniform space and then reverse mapping the interpolated points).
crazygringo 4y ago
> ...could yield more vibrant computer displays, TVs, printed materials, textiles and more I'm sorry, how exactly? From the paper's abstract: > Consequences of this apply to color metrics that are currently used in image and video processing, color mapping, and the paint and textile industries. At best, I could imagine how computer-selected colors might be chosen differently, i.e. adding to the many ways in which gradients between two colors are computed. But I can't for the life of me imagine how computer displays might become more vibrant -- this isn't going to affect the spectrum of backlights or LCD color filters or OLED phosphors or anything as far as I can tell? E.g. in the way that Macs shifted from sRGB screens to P3 screens for their color gamuts? Similarly with paints and textiles -- this isn't going to lead to new paints or dyes. At the end of the day, colors -- whether in a movie scene or a paint chip -- are ultimately chosen by eye, not by math. So seems like the headline is making claims that can't be substantiated? Would love to know if I'm missing something here.
- robbrown451 4y ago
  Agree that they haven't offered much in terms of how it could make displays more vibrant. > At the end of the day, colors -- whether in a movie scene or a paint chip -- are ultimately chosen by eye, not by math Well.... math is involved if you want to do even simple stuff like blending two colors in a paint program. There is subjective stuff, sure, but you aren't going to get very far in computer graphics, display manufacture or camera chip manufacture, if you just conclude that it is 100% subjective and there is no mathematical logic behind it.
  - refulgentis 4y ago
    It inherently doesn't make any sense because the limits, the "bounds of space" of displays is physical. Changing your indexing system doesn't sprout new colors
    - orbital-decay 4y ago
      It very much does, as color is not a physical phenomenon. It's just perception - essentially a bunch of weights in your neural network that makes some internal conceptual sense. You can trivially go outside of those "physical limits" e.g. by exploiting temporal perception quirks to get impossible colors (not represented by any spectral combination in typical conditions). The same can be said about all human perception in general, not just colors. Artists exploit this for centuries, and are trying to formalize this. Newer color appearance models try to precisely quantify perception effects to make better color appearance possible.
meltyness 4y ago
So principally that the perceived luminance would be dependent on hue so that each hue in the gamut, and each interrelationship has such a relationship perceived intensity as a function of powered intensity. I can only assume that's what they mean by second-order. Now, is this in equilibrium viewing? What's really gonna cook these guys noodle is the effect of pupillary action, e: and afterimages for that matter
- hilbert42 4y ago
  "Now, is this in equilibrium viewing? What's really gonna cook these guys noodle is the effect of pupillary action." Right - and more I reckon. In hindsight things seem so easy or fall easily into place and this is no exception. The obvious thing with respect to this new 'color space' is its nonlinearity/'curvature'. We should have gleaned this (or have taken more cognizance from existing overvations) from the way the eye works in different circumstances - and for that matter from the fact that different eyes perceive colors in different ways - as Ishihara tests bear out. It also should have been obvious from the 'dynamical' way the eye perceives color. Anyone who has done color balancing/grading or tried to match colors knows how easily it is to be fooled unless they've lots of calibration equipment around them, as the eye automatically adjusts color balance and such depending on lighting and circumstance (remember those shadow images that fool everyone). There are other clues too: why do a small percentage of women have superior color resolution in the green spectrum (as if they had multiple green channels) and the fact that the degree of this capability also depends on conditions and circumstance (a seemingly nonlinear effect)? That's right, the signs were there and it's good that we've now picked them up and are now running with them. That said, we cannot be too critical of that earlier work, in fact what was done was pretty monumental. After all, the 1931 CIE work was and still is quite remarkable and it's still serving us well today - 91 years on. Like Newtonian mechanics and Relatively, every development inches us closer to the truth and or provides us with better accuracy.
  - ff317 4y ago
    > There are other clues too: why do a small percentage of women have superior color resolution in the green spectrum (as if they had multiple green channels) Apparently a small number of women have Tetrachromacy, where they have a 4th type of cone in their eye and can see a much broader range of colors, thanks to their genes. The gene isn't super-rare (I think one article claimed ~12% of women have the genetic condition for it), but the functional ability to perceive with all 4 types of cones is very rare. https://www.bbc.com/future/article/20140905-the-women-with-super-human-vision https://www.bbc.com/future/article/20140905-the-women-with-s... has a story about one confirmed case.
mrleinad 4y ago
Alternative link (looks like the original is down): https://phys.org/news/2022-08-math-error-overturns-year-old-perception.html https://phys.org/news/2022-08-math-error-overturns-year-old-...
Cadwhisker 4y ago
I have always found the XKCD color survey of 2010 to be a most excellent reference in color perception: https://blog.xkcd.com/2010/05/03/color-survey-results/ https://blog.xkcd.com/2010/05/03/color-survey-results/ And still, nobody can spell fuchsia.
- gtsop 4y ago
  Think of the word Nokia. Now say foo - ksee - a. On color perception, it is worth noting that it is culture-depedent. I've read some indegenous people can name you tens of different shades of green, while grouping together everything we would call red-orange-etc into a single color
- BrandoElFollito 4y ago
  This was amazing to read as a colorblind person. The graph showing simple names was a feast as I was trying to find out where my colour-blindness fls in the green/yellow region. xkcd really rules.
- JoeAltmaier 4y ago
  I didn't see any acknowledgment of the effect computer monitors could have had in muddying the data. Some monitors represent colors differently. In fact if you don't use a gamma-corrected monitor then you have no chance at all.
- Karawebnetwork 4y ago
  It's easier when you remember that is is named after Leonhard Fuchs.
light_hue_1 4y ago
So Los Alamos is just going to rip off their own researchers now? Are reviewers at PNAS asleep? Zeyen M, Post T, Hagen H, Ahrens J, Rogers D, Bujack R. Color interpolation for non-Euclidean color spaces. In 2018 IEEE Scientific Visualization Conference (SciVis) > "Furthermore, human color perception is also non-Riemannian, due to the principle called diminishing returns [9], Figure 1" These people in 2018 literally pointed out the same thing! Also from Los Alamos. I'm starting to think HN should be renamed to "another sad day in science".
achow 4y ago
..Bernhard Riemann, 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. Nice :-)
tgv 4y ago
Another day, another hyperbolical title. 100 years ago, knowledge of perception was in its infancy. We still don't know much, but we do know that all attempts at a simple, mathematical description of perceptual processes {*} have failed. They're interesting, can serve as an introduction to a problem, may help you to build a better web app, etc., but come nowhere near the complexity of an actual brain. It takes a pretty large neural net to decypher phonemes even in clear speech. Viewing color perception as an independent function mapping one space onto another is already flawed, as the chessboard illusion already shows (https://en.wikipedia.org/wiki/Checker_shadow_illusion https://en.wikipedia.org/wiki/Checker_shadow_illusion). {*} This is not limited to perception. Attempts to described even small parts of reasoning, language and memory with simple models failed too, but that seems less surprising, since perception is closer to the physical reality.
- orbital-decay 4y ago
  Certain models do try to account for spatial localization and relativity too (iCAM), although they are also simple as those effects remain largely unexplored. Honestly, ML seems to be better suited to this task than complex analytical solutions, as long as you have the data to train on.
- sandgiant 4y ago
  While the title does sound a bit sensationalist the actual paper refers to a standard in the International Commission for Weights and Measures which is indeed (according to the paper) based on 100-year old paradigm introduced by Riemann and furthered by Helmholtz and Schrodinger. From the abstract: The scientific community generally agrees on the theory, introduced by Riemann and furthered by Helmholtz and Schrödinger, that perceived color space is not Euclidean but rather, a three-dimensional Riemannian space. We show that the principle of diminishing returns applies to human color perception. This means that large color differences cannot be derived by adding a series of small steps, and therefore, perceptual color space cannot be described by a Riemannian geometry. https://www.pnas.org/doi/10.1073/pnas.2119753119 https://www.pnas.org/doi/10.1073/pnas.2119753119
  - hilbert42 4y ago
ximm 4y ago
Can someone explain what impact this has on "perceptual" color spaces like Oklab? Is this just something they live with?
eterevsky 4y ago
It sounds like all this paper says is that the metric of the color spaces is non-linear, i.e. the perceived distance between two colors doesn't really agree with the Euclidean distance between 3D points representing them. I think this is more or less well-known.
- OscarCunningham 4y ago
  They're saying it's not just non-Euclidean but also non-Riemannian. The distance between two points is shorter than the length of the shortest path between them. For a simple example of such a space, imagine a circle where we define the distance between two points to be measured 'as the crow flies', but we demand that all paths remain on the circle rather than passing through its interior.
  - quickthrower2 4y ago
    That example would be similar to using as the crow flies for distances when driving on roads.
BrandoElFollito 4y ago
I am an ex-physicist and was always annoyed by the color ←→ wavelength mapping in school books. Then in biology class they come up with the retina cones that are supposed to detect only a specific wavelength (or "color"). First belly dancing exercise to break out of the mapping. Unfortunately the cones are explained as e.g. "red color cone" - and this leads top the second belly dancing exercise and handweaving about other colors (the ones not covered by the cones), using paint mixing as examples. Colors teaching at school is a complete shitshow. Fortunately this is not something that is important until you are in a relationship and have to paint a wall. Or that you have this idiot eye doctor I had as a 10 yo kid, who told my father in front of me that I was colorblind and would never do science. Fuck you idiot who said words I remember till this day (as someone who loved science) - I would be so glad to make a copy of my engineering diploma and of my PhD in physics (and all the papers I wrote) and shove it into your ass. I really resent such people who do not realize how easy it is to crush a kid's dream, similar to the doll vs truck.
throwaway290 4y ago
I hope the industry catches on quickly. New color spaces, new displays.
JoeAltmaier 4y ago
I'm confused somebody even uses a mathematical model of color. Human perception of color is certainly a matter of biology. I'd have expected this paper to be a study of retinal cell response saturation or some such. Which is whatever it is, never mind famous mathematician's opinions.
- waqf 4y ago
  > Human perception of color is certainly a matter of biology. Of course it is. The mathematics is a model of human biology (in terms of the effects of interactions with different colours of light). Mathematical models of the underlying electromagnetic phenomenon known as light already existed in the 19th century, but that's very different from the problem at hand.
- LocalH 4y ago
  Mathematical models of color can serve several useful purposes, even if they're not entirely accurate. As long as one keeps in mind that at best, they are descriptive, there can be valid use cases.
DoneWithAllThat 4y ago
Once again, we should never take a single study as proof of anything. Whether or not anything new has been learned will be demonstrated by repeatability, application of the new research, and time.
turnsout 4y ago
I love that they fell down this rabbit hole while trying to choose colors for data visualizations. It supports my belief that 80% of all color science research is the result of extreme bikeshedding!
EricE 4y ago
"...and proving one of them wrong is pretty much the dream of a scientist." What an odd and somewhat alarming statement!
- dj_mc_merlin 4y ago
  Why do you think so? People are famous throughout history because they leave behind (or are associated with) big achievements which form a "node" in the history of science. If you prove them wrong, the node you leave behind is also important. Scientists (and most people in general) want do important work. There's also the scientific benefit that overturning a generally held idea usually precipitates a large amount of new and fruitful study afterwards.
todd8 4y ago
One hundred years ago, it would have been appealing to have a mathematical model of color perception that scientists could work from. The problem, of course, is that the systems of perception within humans and other mammals are very complex. There are large differences in color perception between individuals and, I suspect, even between genetic groups. The visual system has many layers and a lot of neural processing goes on before we "see" what we are looking at. This system is not built out of a handful of components easily modeled by differential equations like an electronic circuit; it is built out of non-linear, stateful components (neurons) with at least millions of interconnects. A mathematical model that is intended to capture every nuance is bound to fail. At best, in my uninformed opinion, we could simply do subjective surveys of perceptual color response and encode the averages and differences between individuals within standardized tables. It would then be simple for applications to make use of these tables for making and measuring systems intended for human use. An illustration of the complexity of the visual system can be found in the wide range of visual illusions we are subject to, see [1] and [2]. Most of these are due to the visual system developing to interpret the visual world with very limited equipment, our human eyes. A purely mathematical model wouldn't be complete unless it too encompassed the strange optical illusions we all have. (I should point out that individuals have different responses to optical illusions.) [1] https://www.sciencedirect.com/topics/neuroscience/visual-illusion https://www.sciencedirect.com/topics/neuroscience/visual-ill... [2] https://en.wikipedia.org/wiki/Optical_illusion https://en.wikipedia.org/wiki/Optical_illusion

226 points by beefman 4y ago

user- 4y ago

403

EarlKing 4y ago

Well this is a first for me. I don't think I've ever seen a paper published to PNAS that wasn't open access.

tingletech 4y ago
embargo should expire October 29, 2022: "PNAS is a delayed open access journal, with an embargo period of six months that can be bypassed for an author fee (hybrid open access). " https://en.wikipedia.org/wiki/Proceedings_of_the_National_Academy_of_Sciences_of_the_United_States_of_America https://en.wikipedia.org/wiki/Proceedings_of_the_National_Ac...
- fnordpiglet 4y ago
  LOL I came to the comments to find the paper
- EarlKing 4y ago
  Huh. Well today I learned. I always thought it was just plain open access.
collegeburner 4y ago
hopefully somebody put it on scihub soon

ggm 4y ago

I've always found the printed book approximations of how people see colour fascinating. The amount of work a printer has to go through to achieve colour fidelity printing a continuously varying hue field in some specified geometry.. Now imagine doing this before PDF/Postscript.

AdamH12113 4y ago

When they talk about Riemannian geometry, is that related to the Riemann curvature tensor from general relativity? What’s the significance of being non-Riemannian? Is it just that (as the abstract says) “the distance between two colors is [not] the length of the shortest path that connects them.”?

adrian_b 4y ago
A point in the color space is determined by 3 numbers. Given 2 colors, i.e. 2 points in the color space, one can define a distance function from them to a number corresponding to the perceived color difference. One can try various formulas for the color distance and compare them in experiments with humans who must assess which colors are more similar or more different. Such experiments are extremely time consuming and require many test subjects, to average over individual variability. The problem is that a formula for a very good approximation of the color differences over the entire color space has not been found yet, even if there are many formulas that give acceptable results in certain cases, usually when the color differences are small. The Euclidean distance does not give a good approximation, so the color space is not Euclidean. The distances corresponding to various Riemannian spaces (i.e. "curved" spaces) give better approximations, but which are still not good enough. In fact the color distance is not even a distance in the mathematical sense, i.e. it does not satisfy the triangle axiom, so the color space is not a metric space. Perhaps there exists a non-linear transformation from the color space to a metric space.
- hilbert42 4y ago
  "Riemannian spaces (i.e. "curved" spaces) give better approximations, but which are still not good enough." Right, as I said above this work shows we've nonlinear effects at work, whilst this was obvious in the past we should have taken more notice of the fact. Whilst Riemann provides say an ideal mathematical model (the ideal case) as we obserce Nature doesn't always comply, it insists on adding 'distortions' of its own. For the same reason these 'transfer' curves have nonlinear characteristics it seems other significant interactions are at work, namely cross-coupling between color channels. We see this in color film emulsions where color bleeds from one layer to another and this effect is also nonlinear (if the bleed were linear across the transfer curve then we could simply bias it out but it's not thus in practice it's nigh on impossible to correct). In effect this cross-coupling is essentially equivalent to cross-modulation distortion in electronic circuits and the mixing can be modeled mathematically in a similar way. No doubt it remains to be seen whether this new work has any significant bearing on this or not.

canjobear 4y ago

I thought it was already known that perceived large differences in colors were not geometrically consistent with small differences. For example CIECAM02-UCS is based on a compromise between fitting the small-difference data and large-difference data.

adrian_b 4y ago
Yes, from the abstract of the article it is impossible to understand what is new in this study. The "non-Riemannian nature of perceptual color space" has been known for decades. For example the CIE color difference formulas from 1994 and from 2000 (CIEDE2000) are older attempts to model the color space, taking this into account, i.e. that the color space is not even a metric space (the triangle axiom is not true). There are various newer attempts to better model the color space. Without being able to read the paper, I assume that it might refer to one such better model, but as I have said, neither the title nor the abstract give any clue about what is novel in it. They say "Rethinking them outside of a Riemannian setting could provide a path to extending them to large differences". From the "could provide a path" I understand that they have not found yet any formula valid for large color differences, which would have been something novel.
- OscarCunningham 4y ago
  Can you give an example of three colours where the triangle inequality is false? Would be interesting to look at.
  - adrian_b 4y ago
    One paper where some such experiments are discussed and which includes some color diagrams is: https://psyarxiv.com/vtzrq/download?format=pdf https://psyarxiv.com/vtzrq/download?format=pdf The violations of the triangle inequality are caused, as others have also mentioned, both by the nonlinearity of the dependence between the perceived intensity of a color and its corresponding radiant flux and by the interactions between the 3 color channels. Therefore, while the 3 RGB or XYZ values are enough for being able to recognize if any pair of colors are the same or different, to be able to recognize whether in a set of 3 colors the 3rd is closer to the 1st or to the 2nd, one needs to apply some non-linear function of 3 arguments, providing 3 output numbers, in order to map the RGB/XYZ space to a space where a simple distance formula, e.g. the Euclidean distance or the Manhattan distance, can be used. Alternatively the non-linear 3 to 3 transform can be combined with the distance formula in the transformed space into a non-linear complicated distance formula that can be used directly in some applications, on RGB/XYZ colors, but separating the space mapping transform can be useful in some other applications, e.g. in color interpolation (though color interpolation can also be done based on only a distance formula, by solving an equation for each intermediate point, which might be not slower than mapping the extreme points, interpolating in the uniform space and then reverse mapping the interpolated points).

crazygringo 4y ago

> ...could yield more vibrant computer displays, TVs, printed materials, textiles and more I'm sorry, how exactly? From the paper's abstract: > Consequences of this apply to color metrics that are currently used in image and video processing, color mapping, and the paint and textile industries. At best, I could imagine how computer-selected colors might be chosen differently, i.e. adding to the many ways in which gradients between two colors are computed. But I can't for the life of me imagine how computer displays might become more vibrant -- this isn't going to affect the spectrum of backlights or LCD color filters or OLED phosphors or anything as far as I can tell? E.g. in the way that Macs shifted from sRGB screens to P3 screens for their color gamuts? Similarly with paints and textiles -- this isn't going to lead to new paints or dyes. At the end of the day, colors -- whether in a movie scene or a paint chip -- are ultimately chosen by eye, not by math. So seems like the headline is making claims that can't be substantiated? Would love to know if I'm missing something here.

robbrown451 4y ago
Agree that they haven't offered much in terms of how it could make displays more vibrant. > At the end of the day, colors -- whether in a movie scene or a paint chip -- are ultimately chosen by eye, not by math Well.... math is involved if you want to do even simple stuff like blending two colors in a paint program. There is subjective stuff, sure, but you aren't going to get very far in computer graphics, display manufacture or camera chip manufacture, if you just conclude that it is 100% subjective and there is no mathematical logic behind it.
- refulgentis 4y ago
  It inherently doesn't make any sense because the limits, the "bounds of space" of displays is physical. Changing your indexing system doesn't sprout new colors
  - orbital-decay 4y ago
    It very much does, as color is not a physical phenomenon. It's just perception - essentially a bunch of weights in your neural network that makes some internal conceptual sense. You can trivially go outside of those "physical limits" e.g. by exploiting temporal perception quirks to get impossible colors (not represented by any spectral combination in typical conditions). The same can be said about all human perception in general, not just colors. Artists exploit this for centuries, and are trying to formalize this. Newer color appearance models try to precisely quantify perception effects to make better color appearance possible.

meltyness 4y ago

So principally that the perceived luminance would be dependent on hue so that each hue in the gamut, and each interrelationship has such a relationship perceived intensity as a function of powered intensity. I can only assume that's what they mean by second-order. Now, is this in equilibrium viewing? What's really gonna cook these guys noodle is the effect of pupillary action, e: and afterimages for that matter

hilbert42 4y ago
"Now, is this in equilibrium viewing? What's really gonna cook these guys noodle is the effect of pupillary action." Right - and more I reckon. In hindsight things seem so easy or fall easily into place and this is no exception. The obvious thing with respect to this new 'color space' is its nonlinearity/'curvature'. We should have gleaned this (or have taken more cognizance from existing overvations) from the way the eye works in different circumstances - and for that matter from the fact that different eyes perceive colors in different ways - as Ishihara tests bear out. It also should have been obvious from the 'dynamical' way the eye perceives color. Anyone who has done color balancing/grading or tried to match colors knows how easily it is to be fooled unless they've lots of calibration equipment around them, as the eye automatically adjusts color balance and such depending on lighting and circumstance (remember those shadow images that fool everyone). There are other clues too: why do a small percentage of women have superior color resolution in the green spectrum (as if they had multiple green channels) and the fact that the degree of this capability also depends on conditions and circumstance (a seemingly nonlinear effect)? That's right, the signs were there and it's good that we've now picked them up and are now running with them. That said, we cannot be too critical of that earlier work, in fact what was done was pretty monumental. After all, the 1931 CIE work was and still is quite remarkable and it's still serving us well today - 91 years on. Like Newtonian mechanics and Relatively, every development inches us closer to the truth and or provides us with better accuracy.
- ff317 4y ago
  > There are other clues too: why do a small percentage of women have superior color resolution in the green spectrum (as if they had multiple green channels) Apparently a small number of women have Tetrachromacy, where they have a 4th type of cone in their eye and can see a much broader range of colors, thanks to their genes. The gene isn't super-rare (I think one article claimed ~12% of women have the genetic condition for it), but the functional ability to perceive with all 4 types of cones is very rare. https://www.bbc.com/future/article/20140905-the-women-with-super-human-vision https://www.bbc.com/future/article/20140905-the-women-with-s... has a story about one confirmed case.

mrleinad 4y ago

Alternative link (looks like the original is down): https://phys.org/news/2022-08-math-error-overturns-year-old-perception.html https://phys.org/news/2022-08-math-error-overturns-year-old-...

Cadwhisker 4y ago

I have always found the XKCD color survey of 2010 to be a most excellent reference in color perception: https://blog.xkcd.com/2010/05/03/color-survey-results/ https://blog.xkcd.com/2010/05/03/color-survey-results/ And still, nobody can spell fuchsia.

gtsop 4y ago
Think of the word Nokia. Now say foo - ksee - a. On color perception, it is worth noting that it is culture-depedent. I've read some indegenous people can name you tens of different shades of green, while grouping together everything we would call red-orange-etc into a single color
BrandoElFollito 4y ago
This was amazing to read as a colorblind person. The graph showing simple names was a feast as I was trying to find out where my colour-blindness fls in the green/yellow region. xkcd really rules.
JoeAltmaier 4y ago
I didn't see any acknowledgment of the effect computer monitors could have had in muddying the data. Some monitors represent colors differently. In fact if you don't use a gamma-corrected monitor then you have no chance at all.
Karawebnetwork 4y ago
It's easier when you remember that is is named after Leonhard Fuchs.

light_hue_1 4y ago

So Los Alamos is just going to rip off their own researchers now? Are reviewers at PNAS asleep? Zeyen M, Post T, Hagen H, Ahrens J, Rogers D, Bujack R. Color interpolation for non-Euclidean color spaces. In 2018 IEEE Scientific Visualization Conference (SciVis) > "Furthermore, human color perception is also non-Riemannian, due to the principle called diminishing returns [9], Figure 1" These people in 2018 literally pointed out the same thing! Also from Los Alamos. I'm starting to think HN should be renamed to "another sad day in science".

achow 4y ago

..Bernhard Riemann, 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. Nice :-)

tgv 4y ago

Another day, another hyperbolical title. 100 years ago, knowledge of perception was in its infancy. We still don't know much, but we do know that all attempts at a simple, mathematical description of perceptual processes {*} have failed. They're interesting, can serve as an introduction to a problem, may help you to build a better web app, etc., but come nowhere near the complexity of an actual brain. It takes a pretty large neural net to decypher phonemes even in clear speech. Viewing color perception as an independent function mapping one space onto another is already flawed, as the chessboard illusion already shows (https://en.wikipedia.org/wiki/Checker_shadow_illusion https://en.wikipedia.org/wiki/Checker_shadow_illusion). {*} This is not limited to perception. Attempts to described even small parts of reasoning, language and memory with simple models failed too, but that seems less surprising, since perception is closer to the physical reality.

orbital-decay 4y ago
Certain models do try to account for spatial localization and relativity too (iCAM), although they are also simple as those effects remain largely unexplored. Honestly, ML seems to be better suited to this task than complex analytical solutions, as long as you have the data to train on.
sandgiant 4y ago
While the title does sound a bit sensationalist the actual paper refers to a standard in the International Commission for Weights and Measures which is indeed (according to the paper) based on 100-year old paradigm introduced by Riemann and furthered by Helmholtz and Schrodinger. From the abstract: The scientific community generally agrees on the theory, introduced by Riemann and furthered by Helmholtz and Schrödinger, that perceived color space is not Euclidean but rather, a three-dimensional Riemannian space. We show that the principle of diminishing returns applies to human color perception. This means that large color differences cannot be derived by adding a series of small steps, and therefore, perceptual color space cannot be described by a Riemannian geometry. https://www.pnas.org/doi/10.1073/pnas.2119753119 https://www.pnas.org/doi/10.1073/pnas.2119753119
- hilbert42 4y ago

ximm 4y ago

Can someone explain what impact this has on "perceptual" color spaces like Oklab? Is this just something they live with?

eterevsky 4y ago

It sounds like all this paper says is that the metric of the color spaces is non-linear, i.e. the perceived distance between two colors doesn't really agree with the Euclidean distance between 3D points representing them. I think this is more or less well-known.

OscarCunningham 4y ago
They're saying it's not just non-Euclidean but also non-Riemannian. The distance between two points is shorter than the length of the shortest path between them. For a simple example of such a space, imagine a circle where we define the distance between two points to be measured 'as the crow flies', but we demand that all paths remain on the circle rather than passing through its interior.
- quickthrower2 4y ago
  That example would be similar to using as the crow flies for distances when driving on roads.

BrandoElFollito 4y ago

I am an ex-physicist and was always annoyed by the color ←→ wavelength mapping in school books. Then in biology class they come up with the retina cones that are supposed to detect only a specific wavelength (or "color"). First belly dancing exercise to break out of the mapping. Unfortunately the cones are explained as e.g. "red color cone" - and this leads top the second belly dancing exercise and handweaving about other colors (the ones not covered by the cones), using paint mixing as examples. Colors teaching at school is a complete shitshow. Fortunately this is not something that is important until you are in a relationship and have to paint a wall. Or that you have this idiot eye doctor I had as a 10 yo kid, who told my father in front of me that I was colorblind and would never do science. Fuck you idiot who said words I remember till this day (as someone who loved science) - I would be so glad to make a copy of my engineering diploma and of my PhD in physics (and all the papers I wrote) and shove it into your ass. I really resent such people who do not realize how easy it is to crush a kid's dream, similar to the doll vs truck.

throwaway290 4y ago

I hope the industry catches on quickly. New color spaces, new displays.

JoeAltmaier 4y ago

I'm confused somebody even uses a mathematical model of color. Human perception of color is certainly a matter of biology. I'd have expected this paper to be a study of retinal cell response saturation or some such. Which is whatever it is, never mind famous mathematician's opinions.

waqf 4y ago
> Human perception of color is certainly a matter of biology. Of course it is. The mathematics is a model of human biology (in terms of the effects of interactions with different colours of light). Mathematical models of the underlying electromagnetic phenomenon known as light already existed in the 19th century, but that's very different from the problem at hand.
LocalH 4y ago
Mathematical models of color can serve several useful purposes, even if they're not entirely accurate. As long as one keeps in mind that at best, they are descriptive, there can be valid use cases.

DoneWithAllThat 4y ago

Once again, we should never take a single study as proof of anything. Whether or not anything new has been learned will be demonstrated by repeatability, application of the new research, and time.

turnsout 4y ago

I love that they fell down this rabbit hole while trying to choose colors for data visualizations. It supports my belief that 80% of all color science research is the result of extreme bikeshedding!

EricE 4y ago

"...and proving one of them wrong is pretty much the dream of a scientist." What an odd and somewhat alarming statement!

dj_mc_merlin 4y ago
Why do you think so? People are famous throughout history because they leave behind (or are associated with) big achievements which form a "node" in the history of science. If you prove them wrong, the node you leave behind is also important. Scientists (and most people in general) want do important work. There's also the scientific benefit that overturning a generally held idea usually precipitates a large amount of new and fruitful study afterwards.

todd8 4y ago

One hundred years ago, it would have been appealing to have a mathematical model of color perception that scientists could work from. The problem, of course, is that the systems of perception within humans and other mammals are very complex. There are large differences in color perception between individuals and, I suspect, even between genetic groups. The visual system has many layers and a lot of neural processing goes on before we "see" what we are looking at. This system is not built out of a handful of components easily modeled by differential equations like an electronic circuit; it is built out of non-linear, stateful components (neurons) with at least millions of interconnects. A mathematical model that is intended to capture every nuance is bound to fail. At best, in my uninformed opinion, we could simply do subjective surveys of perceptual color response and encode the averages and differences between individuals within standardized tables. It would then be simple for applications to make use of these tables for making and measuring systems intended for human use. An illustration of the complexity of the visual system can be found in the wide range of visual illusions we are subject to, see [1] and [2]. Most of these are due to the visual system developing to interpret the visual world with very limited equipment, our human eyes. A purely mathematical model wouldn't be complete unless it too encompassed the strange optical illusions we all have. (I should point out that individuals have different responses to optical illusions.) [1] https://www.sciencedirect.com/topics/neuroscience/visual-illusion https://www.sciencedirect.com/topics/neuroscience/visual-ill... [2] https://en.wikipedia.org/wiki/Optical_illusion https://en.wikipedia.org/wiki/Optical_illusion