Let me break this down into a really simple (though not quite accurate with the real world) example.
You can have a single frequency color of green.
You can also have a combination of blue and yellow to make green (to our eyes).
To our normally functional eyes, they look the same.
A person with blue-green colorblindness would not be able to see the green or the blue, so to them the first one looks gray and the second one looks yellow.
So the colorblind person sees a difference where the normally sighted person doesn’t.
This is kind of a bogus example because I arbitrarily picked primary colors that were easy to explain instead of the actual color ranges that our eyes use, but it should illustrate the point.
Except that the reason we see blue and yellow the same as green is because we don’t see green in the first place. If a mix of two colors looks like a single color, that’s because the single color triggers the same receptors that the two mixed colors do.
Some of that can be from further processing in the brain (and/or optic nerves). Everything we see is subject to the perception formed in post processing.
I think you are pointing out that blue and yellow mixes that look like green are actually reflecting green light to the eye (it has to be cyan, not blue anyway). An additional factor is the overlap in sensitivity in the cones across the spectrum. This may lead to ways to fool the eye since the overlap of blue colors in the green receptors are indistinquishable from green light.
Also, the color-blind camouflage theory should be easily testable by having people wear colored lenses. I have some form of Tritanopia, and I doubt I could pick out hidden images based on better color sensitivity, more the opposite effect from lack of distraction in a limited palette. Of course if you wear dark blue in front of a black background, I might not see you at all.
Post-processing can destroy information but not create it! If you look at the cone-response curves at the link I posted, you can deduce that cone output from yellow is indistinguishable from various red-green combinations. Indeed the CIE color space diagram was designed to have this linear superposition property (although pigment variations mean the precise such diagram would vary from human to human). (BTW, I don’t understand your comment about blue colors.)
Doubtful, especially if we disallow “fluorescent filters.”
Tritanopia is rare, I guess. Of the many types of color blindness, which are useful for camouflage detection?
No. That is a bogus example because it does not reflect how color vision, or color blindness, works at all. It is full of misconceptions.
For a start, there is no such thing as “blue-green” color blindness, and even if we change your example to use a form that does exist, such as “red-green” color blindness, you are still getting it wrong. “Red-green” color blindness does not mean that neither the “red” nor the “green” cones are working (as, in you example, you apparently thought “blue-green” color blindness would require that neither “blue” nor “green” cones were working), it means that either (but not both) of the “red” or “green” cones are not working, so that reds and greens (and yellows) cannot be distinguished.
For example, one of the types of so called “red-green” color blindness (deuteranopia) occurs when the so called “green” cones are not functional, although the other cone types are fine. This does not lead sufferers to see greens (monochromatic or otherwise) as greys, with reds still appearing as red. What it actually means is that hues in the red-yellow-green part of the spectrum all look the same, although they can still be distinguished from bluer hues.
As I said earlier, color blindness is about not having the ability to make certain hue discriminations that normally sighted people can make, and it does not confer the ability to make distinctions that normals cannot. If the color-blind are sometimes less likely to be fooled by camouflage, this is not because they can make color distinctions that the rest of us cannot, but because they spot other cues as to the presence of camouflaged objects that tend to be missed by us normals, because we are distracted by color differences that are very salient and attention grabbing to us, but which the color-blind do not see. (See the two peer reviewed scientific articles that I linked to before.)
The nerve signals sent from the eye to the brain do not represent the levels of stimulation of each of the three cone types, as you seem to be assuming, but rather differences between those levels. The levels of stimulation of the three types of cone cell are processed within the retina into two opponent channels, which (simplifying somewhat) may be called the red-green and the blue-yellow channels. The former encodes, essentially, the difference in response between the “red” and “green” cones, and the latter encodes, roughly speaking, the difference in response between the “blue” cones and the combined signal from the “red” and “green” ones. Thus, for each point in the visual field at which we see a color, the brain receives two signals, one indicating to what extent that point is more blue than yellow, and the other indicating to what extent it is more red than green. The color vision system works with the differences between the responses of the different cone types, and knocking out one of the cone types can only reduce these differences, never increase them.
Also, as Quercus pointed out. Even if color vision did work the way you think, that would not provide color-blind people with any advantage in looking at reconnaissance photographs. Let’s say, the camouflage paint reflects a monochomatic green light, whereas the foliage around it reflects some blue and yellow wavelengths that are perceived in combination by the normal eye as the same green. Despite those differences, the color film is going to render both those greens into the same green pigment (if it did not, normally sighted people would be able to see the difference too - and, of course, the picture would look all wrong) so any difference that may have existed originally is gone long before the stimulus reaches the color-blind person’s eyes.
As I said before, any advantage that color-blind people may have in viewing reconnaissance photographs arises not from their being able to see color distinctions that the rest of can’t, but from their not being distracted from other cues by color distinctions that tend to distract the rest of us. (Printing versions of the photos in black-and-white might have worked as well, or better, but perhaps it was cheaper and easier to find a color blind soldier to view the color photos.)
Post processing cannot create information?! Have you ever seen an optical illusion? Any that involve perceived movement are definitely due to post-processing. I’m not up on the explanations of all color phenomena, but some are definitely post processing.
Yellow is green and red light. You don’t have yellow receptors, and that is all post- processing.
Why wouldn’t color filters work? It would change the perception of color, which as njtt points out is probably the only effect in the color blind - camo case.
Perhaps there’s a confusion about the word “information” which has a specific meaning in, well, information theory.
Perhaps it would have been clearer to write: “Post processing cannot create new information.” When you submit an image to a Jpeg compressor, you get a new representation, but you don’t get any new information; instead you actually lose information. Optical illusions can give you strange sensations, but the only external input to the system is through the rods and cones of retina.
To understand that color filters cannot in general emulate a color blindness, it may be good enough to look at the cone-response diagrams, to remember that each pixel in a real-world image is itself an arbitrary function of the visible spectrum (i.e. a point in an infinite-dimensioned Hilbert space) and then try to imagine what specific filter could possibly achieve the effect you seek for every such real-world color (*). If the different cone type’s spectral responses were disjoint (i.e. if the blue cones never respond when green cones do), then blocking blue light would emulate the loss of blue cones, but blue and green cones’ responses do overlap.
(* - where color here refers to the arbitrary function of spectrum, not a tristimulus system in which, e.g. yellow = red + blue for poople with normal vision.)
I used ‘post processing’ to describe brain functions, which you have taken to literally. You are also misapplying info theory. Nothing stops the creation of new information, its just not based solely on the input. As for filters, I’m just talking about the effect of removing some information. It would change the image received. That’s all.
An interesting aside you might have info on: A couple of ‘red-green’ color blind people I’ve met swore by different colored contact lenses that gave them the ability to better distinquish between red and green. I’ve never heard of this anywhere else.
My own color blindness is mild, and possibly genetic. I only recently found out about it, and then learned my father had a worse problem distinquishing colors. Unfortunately he’s no longer around to get more information. I still need to go to a specialist with the big book of test images to get a more specific diagnosis, but I’ve always noticed that blue-green colors that others describe as blue have appeared distinctly green to me. I can’t tell navy blue from black except in the brightest sunlight, and blue displayed on CRTs looks extremely low in intensity compared to red and green. Hasn;t been much of problem though, much less overall effect than other forms of colorblindness.
I think this is only part of the answer. My officemate and I can look at the same graph with, say, 10 different colors of dots, and thousands of dots of each color on the same page. Because of bleeding and a poor color printer, I (with normal vision) can have trouble discerning purple from red. He has no trouble, because to him purple is blue, and red is only visible because of a small response from the other receptors. So, his natural filtering heightens a contrast. I can easily believe that the same would hold true looking at photographs: filtered vision heightens some contrasts, which makes artificial shapes and texture difference more obvious, and not just because he would cue on it more strongly.
Impossible to tell without testing you, but the mild forms are anomalous trichromacy, likely protanomaly and deutanolmaly. These people have three cones, but two are extremely similar in sensitivity, so that they have difficulty with some colors, but don’t confuse reds and greens to the extent of colorblindness. It may also be something completely unrelated to your visual system.
Incidentally, blue on CRTs seems dimmer because it is. Green is the brightest.
To him, purple is blue, but then, to you also, purple is blue. It’s just that, for you, it’s red as well. If there’s any connection at all between his being able to tell the difference and being color-blind, it’s just that he has more practice in compensating for hard-to-distinguish colors. But it doesn’t give him any fundamental capabilities that you lack.
Yes, I’m familiar with that. But good CRTs can adjust the level for each color. I have the effect until blue is at the max (without excessive bleed, more allowed if its only me looking), and red and green brought down below standard matched levels. Then everyone brings up blue phosphors and blue receptors and I just ignored it. Recently I saw a test image online and couldn’t read the numbers. I figured it was just an artifact of the plasma screen, but then recalled all the ways blue colors seemed ‘different’ for me and tried a photographic image, and still couldn’t pick out the numbers. A specialist is supposed to have a whole range of similar test images for a better diagnosis. It’s not like I’m can’t distinquish blue from green altogether, so it hasn’t had much affect on my life.
Brown is a more interesting color. Its all in the post processing. Very interesting because almost everybody distinquishes brown easily, but it requires all the cones and maybe the rods as well to provide input. njtt mentioned you can’t perceive brown without contrasting samples. If this processing is occuring in the eye or optic nerve, that supports the statement made by someone that the eyes are part of the brain and not a seperate organ (paraphrasing).
Well, yes, if you are missing your “red” cones, reds can look very dark, besides being indistinguishable in hue from greens and yellows. Indeed, such people can sometimes use this difference in experienced shade to guess that what they are looking at is red rather than green, even though they do not seem to differ in hue. (But, of course, they could be confused by very light reds and dark greens, where the later would still seem the darker.) You are right that this could also sharpen the apparent contrast between a red and a purple of similar shades. In such a case, they would still see a hue difference between red and purple, but they would also see quite a large difference in shade that might not be there at all for the normally sighted.
So yes, in some cases colorblindness can sharpen differences between different colors, by trading hue differences for relatively large shade differences. What I do not think it will do, however, despite what ECG claims, is create a contrast between metameric colors (i.e., colors that look the same to normals, but may be composed of different mixtures of wavelengths), such as the green of foliage and the green of camouflage paint (which, presumably, will be matched as closely as possible by the camouflagers). I suppose if someone tried to camouflage something on a red background by painting it purple then a protanope (someone without functional “red” cones) might be able to spot the contrast more easily, but why would the camouflager do it that way?
A deuteranope (someone lacking functional green cones) will presumably, for similar reasons, experience some darkening of greens relative to cyans, although this effect is not nearly as marked because the “red” cones are, in fact, still responsive to light far into the green, and even the blue regions of the spectrum. Still, this might give them a slightly enhanced shade contrast between greens and cyans. But why should a camouflager paint his tank cyan, when green paint is readily available?
The first page I found with a search shows respective RGB luminances of 13.5, 43.1, and 5.22 cd/m^2. You can lower G down to B, then it would be at about 12%, which may limit the gamut. I’m not sure how the adjustment you describe works (brightness matching?).
It would be cool to figure out what you have, but then it doesn’t affect you much and probably costs more than it’s worth. They would run a full battery of tests.
Completely unsubstantiated speculation: Lenses generally have chromatic aberration and, if the myopic correction differs between eyes, so would the chromatic aberration, and this might lead to detectable difference. My own eyeglasses result in blue objects appearing slighly closer than red objects, though I only notice this in special cases. Note that this effect affects hues differently, even hues the color-blind person would consider identical. As a wild probably-wrong guess, it may be possible, at least in theory, to reconstruct color information by comparing the outputs of two eyes, affected by lenses with different chromatic aberrations.
By the way, I have my own relatively rare vision disorder it took me some time to get diagnosed. I have “monocular diplopia”, caused by map dot fingerprint dystrophy, which in turn may have caused by staring at computer screens for too long without blinking! :smack:
