Bigfoot Is Blurry
July 19, 2021 6:34 PM   Subscribe

Why we're blind to the color blue. I'm always in the market for surprising facts. One of my favorites is that the color blue is always out of focus for the human eye. It's hard to believe since it appears that we see blue clearly, but it's astonishing when shown an example.
posted by Literaryhero (63 comments total) 55 users marked this as a favorite
 
Shouldn't this mean that a pure blue shape, with no red or green component for the brain get sharpness information from, should look perceptibly blurry no matter what?
posted by nebulawindphone at 6:39 PM on July 19 [3 favorites]


Like, how are we able to read #0000FF text on a #000000 background?
posted by nebulawindphone at 6:43 PM on July 19 [17 favorites]


The human eye is able to reshape itself to adjust focus. If we have only blue light available to us, we can focus in blue, so it's wrong to say that blue is "always out of focus." Rather, that with multiple wavelengths of light available, we'll focus on the colors closer to green and red, and the blue will remain out of focus.

A pure blue object right next to a pure green and pure red object would in theory look blurrier.
posted by explosion at 6:54 PM on July 19 [20 favorites]


This is the same reason why strip mall stores with lighted deep blue-only signage are very, very hard to read at night.
posted by tclark at 6:58 PM on July 19 [14 favorites]


A while back there was a tablet application that tried to do focus training, to help people exercise focusing muscles. I don't know if it was fully scientifically vetted out, but the point about adjusting focus for wavelengths presents an interesting question: Does shifting the color of a block of text from red to blue and back in a smooth gradient-cycle cause extra work for the eye's focusing muscles, and eventually eyestrain?

Also, I suffer from contrast shimmer in some circumstances and I've never thought to pay attention to *which* colors I'm getting with the color fringing, but now I'm mildly curious. Not enough to seek out contrast shimmer, but I'm sure I'll see it soon enough, I suppose. Thanks for the explanation, explosion!
posted by Callisto Prime at 7:01 PM on July 19 [1 favorite]


At this very moment I am looking at the display lights on an electronic device across the room. The green light looks like a sharp little dot. Ditto the red. But the blue is haloed and blurry. If you weren't aware of this phenomenon before, and you think about it, you may recall finding blue faery lights appearing different from other colors--more shimmery, or diffuse-looking, or mysterious. My grandmother told us when we were little that this was because blue faery lights have more magic! It's really brain magic, but hey, that's the best kind anyway.
posted by DrMew at 7:02 PM on July 19 [7 favorites]


I read it on The BlueBlur.
posted by Thorzdad at 7:03 PM on July 19 [52 favorites]


If the blue shape is on a white background, can't the eye get a sharp view of the hole in the background, and just "fill in" the blue?
posted by joeyh at 7:10 PM on July 19 [1 favorite]


This explains why I have such a hard time making out the (blue LED) display on our dishwasher, while the (red LED) display on the stove is clear.

Thanks, literaryhero! You saved me from wasting an AskMe.
posted by Spathe Cadet at 7:10 PM on July 19 [12 favorites]


I call baloney, at least with the article's attempted demo. When I display the original image and the "blurred blue" image in separate tabs at full size, the blurriness isn't massive but it's easy to see a difference between the two images.
posted by Greg_Ace at 7:12 PM on July 19 [2 favorites]


You should be able to see a difference between the images. One of them's been altered to be blurrier.
posted by Spathe Cadet at 7:29 PM on July 19 [1 favorite]


This correlates to a thing I remember about the old Hi8 analog video tape format. Its limited chrominance (color) recording bandwidth was not divided evenly between red, green, and blue light frequencies. Rather, more bandwidth was allocated for capturing red and green frequencies because the human eye was more sensitive to these than to blue frequencies.

I know this because I used the format for underwater video and it did a terrible job capturing anything swimming in open water, e.g. sharks, turtles, etc.
posted by Insert Clever Name Here at 7:32 PM on July 19 [5 favorites]


You should be able to see a difference between the images. One of them's been altered to be blurrier.

Per the article:

blurred blue channel
Similar to before, I have extracted the blue channel from the image and blurred it. On the right, the original red and green channels are recombined with the blurred blue channel. Shockingly, the image appears perfectly sharp!


Which, as I said, it (un-shockingly) does not.
posted by Greg_Ace at 7:33 PM on July 19 [1 favorite]


"Blue skies smiling at me
Nothing but....blue skies
Do I see Bluebirds.....Singing a song
Nothing but blue...."
posted by clavdivs at 7:34 PM on July 19 [1 favorite]


Like, how are we able to read #0000FF text on a #000000 background?
It’s much more difficult to perceive blue-on-black than other combinations. Same for e.g. #0ff/#0f0, #fff/#ff0, or #f0f/#f00. Another factor is that human cornea slightly yellows with age and these combinations become harder to perceive for older people.
posted by migurski at 7:38 PM on July 19 [4 favorites]


Shockingly, the image appears perfectly sharp!

Yes. It appears shockingly sharp relative to blurred-green version which immediately precedes it, even though in both cases a third of the color information in the photo has been altered. When you blur the blue channel, it doesn't change our perception that much, because blues are normally blurry to us anyway.
posted by Spathe Cadet at 7:41 PM on July 19 [2 favorites]


One of my favorites is that the color blue is always out of focus for the human eye.

So this is why people don't read the posts. They can't!
posted by They sucked his brains out! at 7:44 PM on July 19 [26 favorites]


Homer, (iirc), mentioned "wine color seas". I remember one theory was that Greeks didn't have a word for blue, with the suggestion being that humans subsequently evolved to see blue.
posted by SecretAgentSockpuppet at 7:45 PM on July 19 [1 favorite]


Why we're blind to the color blue | Hacker News had thoughts on this a couple of days ago.

There's one specific shade of blue that to me is like neon orange/yellow reflective... it just stands out and catches my attention for no apparent reason. Hrm.
posted by zengargoyle at 8:01 PM on July 19 [2 favorites]




Yes. It appears shockingly sharp relative to blurred-green version which immediately precedes it,

Please re-read my comment, so that if nothing else we're at least disagreeing on the same issue. I said "When I display the original image and the "blurred blue" image in separate tabs at full size, ... it's easy to see a difference". i.e. a direct comparison between the unedited image and the blurred-blue image.
posted by Greg_Ace at 8:03 PM on July 19


counterpoint: only about 2% of cone cells in the retina are "blue-sensitive" and those are not concentrated in the fovea. It's a low-resolution channel, so blurring it is not removing much of the information anyway.
posted by logicpunk at 8:10 PM on July 19 [6 favorites]


One of my favorite "out there" controversies concerns how ancient Greek's perceived the color blue: Wine Dark Sea

Edit: Beaten damn it!
posted by eagles123 at 8:11 PM on July 19 [1 favorite]


"When I display the original image and the "blurred blue" image in separate tabs at full size, ... it's easy to see a difference". i.e. a direct comparison between the unedited image and the blurred-blue image

Direct links to original and blurred blue. When I tab between them I'd say the blurred one looks tinted but not notably blurry to me.

If you look closely at the NOAA and NASA logos you can see that they get a faint blue halo, NASA's because it's blue and NOAA's because of the blue component of the white circle. But I can't see the halos unless I zoom in.
posted by jedicus at 8:12 PM on July 19 [1 favorite]


The original image and the blurred blue image should look different. One of them has had a third of its color information blurred, while the other one has not. When you say that the original image and the blurred blue image, in separate tabs, look different, that is not surprising: they are not the same image. One of them is blurrier. The author tells you that. It would be disappointing if it were not true.

You then switch to saying that the blurred-blue image isn't perfectly sharp, implying that the author is saying that it should be. In context, they are referring to the difference between the blurred-green and blurred-blue images, not to the difference between the blurred-blue and original images. When comparing the blurred green image to the blurred blue image, one of them is notably clearer-looking than the other.

I can only guess that you must be putting a lot of emphasis on "perfectly," such that any detectable difference doesn't meet your understanding of what "perfectly sharp" should mean. This is an odd expectation to have, given that the author never says or implies that you should have it, but if it makes you happy, sure. It's not the point at all, but the original image is sharper than the blurred image, as is normal for image pairs in which one of the images has been made less sharp.
posted by Spathe Cadet at 8:15 PM on July 19 [6 favorites]


Also re: the wine-dark sea thing, it could be a sign of Berlin/Kay’s prediction of blue’s late emergence in action?
Berlin and Kay also found that, in languages with fewer than the maximum eleven color categories, the colors followed a specific evolutionary pattern. This pattern is as follows:

1) All languages contain terms for black and white.
2) If a language contains three terms, then it contains a term for red.
3) If a language contains four terms, then it contains a term for either green or yellow (but not both).
4) If a language contains five terms, then it contains terms for both green and yellow.
5) If a language contains six terms, then it contains a term for blue.
6) If a language contains seven terms, then it contains a term for brown.
7) If a language contains eight or more terms, then it contains terms for purple, pink, orange or gray.
posted by migurski at 8:37 PM on July 19 [17 favorites]


I thought it was just me! When they rotate the dominate color lighting the stage at, say, a concert, all of the performer’s features blur out when they get to blue. And there is one house in the neighborhood that is outlined in blue Christmas lights every year (yard, windows, fence, everything - blue!) and I hate looking at it because it’s a big blob and I have trouble refocusing after I drive past it.
posted by double bubble at 8:50 PM on July 19 [4 favorites]


[Hi8's] limited chrominance (color) recording bandwidth was not divided evenly between red, green, and blue light frequencies. Rather, more bandwidth was allocated for capturing red and green frequencies because the human eye was more sensitive to these than to blue frequencies.

For similar reasons, nearly all digital cameras have an uneven distribution of red, green, and blue pixels on their sensors. The typical arrangement, which dates back to the very early days of digital photography in the 1970s at Eastman Kodak, is the Bayer filter arrangement, consisting of a repeating 2x2 matrix with two green sensors, one red sensor, and one blue sensor—twice as many greens, which hit the middle of the visible-light spectrum, than red or blue, which are at either end. The camera (or RAW image processing software like Lightroom) then cheats a bit, creating an interpolated image with an R,G,B value for each pixel via demosaicing.

It carries with it a whole bunch of tradeoffs and weird edge-case issues.
posted by Kadin2048 at 8:55 PM on July 19 [4 favorites]


Also, I have been participating in a Zoom painting class since covid hit and the instructor gets the most frustrated with how the blues render on the video.
posted by double bubble at 9:01 PM on July 19 [2 favorites]


Is this something most people know about? Because I’ve mentioned my “blue-blindness” to friends several times and they looked at me like I had two heads.
posted by double bubble at 9:05 PM on July 19 [5 favorites]


If there's anyone who didn't read the article and who still believes that words mean things - no, we aren't blind to the color blue.
posted by thatwhichfalls at 9:08 PM on July 19 [4 favorites]


I can only guess that you must be putting a lot of emphasis on "perfectly,"

The author literally said about the blurred-blue image, "Shockingly, the image appears perfectly sharp!". I disputed that, since at full size it isn't hard to detect a difference between the two images. Your argument bolsters my own. It's the author you're arguing against rather than me, so I'm going to stop commenting now.
posted by Greg_Ace at 9:19 PM on July 19


“Shouldn't this mean that a pure blue shape, with no red or green component for the brain get sharpness information from, should look perceptibly blurry no matter what?”
No, because pure blue is not detected only by the “blue” cones in your eyes, but also by the “green” and “red” cones. Pure blue is in the area of the spectrum where the frequency responses overlap for all three cell types.

This means that the blue channel does stimulate all the cone cells in the retina, so our perception of blue alone does work at full resolution. But when a red and green channel are also present, they provide much more signal to the majority of receptors. This is why we can subtract or blur the blue channel without too much damage to our eyes’ ability to resolve detail in the image. Detail in the blue channel gets lost in the noise when combined with the red and green channels.
“Like, how are we able to read #0000FF text on a #000000 background?”
Quite poorly, if your eyes and your display are anything like mine.
posted by mbrubeck at 9:25 PM on July 19 [10 favorites]


That damn dress was Blue
ok, using the blue door
posted by clavdivs at 9:54 PM on July 19


So maybe TMI but I have this monitor with a dim blue blinking LED that I sometimes look at when I can't sleep. When I look at it fairly directly, but with my head propped up by a pillow, it quickly becomes invisible. When I look to the side, I can see it fully. If I stand up and look directly at it, I also see it clearly. It's eerie. This elusive blue light is a companion to my early morning anxious thoughts. Something about blue light. I'm not ready to write it off.
posted by treepour at 10:38 PM on July 19 [3 favorites]


The author literally said about the blurred-blue image, "Shockingly, the image appears perfectly sharp!". I disputed that, since at full size it isn't hard to detect a difference between the two images.

The word "perfectly" may be unfortunate here. I think it's fair, because if I saw that image--by itself--where the blue channel was blurred, I probably wouldn't have thought there was anything remarkable. If you asked me to really scrutinize the image, I maybe would have said it was a little blurry, but even then I'm not sure. In this sense, I think "perfectly sharp" is a fine description.

You're right: if you compare the blue-blurred image with the original, it is pretty easy to say that the blue-blurred image is a bit blurred. In this sense, "perfectly sharp" seems wrong.

To re-emphasize what Spathe Cadet said, I think the blue-blurred image looks very sharp compared to the green-blurred image. This is despite the fact that our naive expectation is that the same quantity of manipulation has occurred to these two images. This is the real point of the exercise, and I think the concern about "perfectly sharp" is a distraction.
posted by polecat at 11:24 PM on July 19 [3 favorites]


> When I look at it fairly directly [...] it quickly becomes invisible.
> When I look to the side, I can see it fully.

That's probably just because it's dim. The light receptors in the central part of your retina are mostly cones, which detect colour bur are are less sensitive. The edges are mostly rods, which can't detect colour differences but are more sensitive.

> If I stand up and look directly at it, I also see it clearly
I'd guess that's because it's emitting a beam which is stronger in the center so when you stand up your eye gets into the brighter part and it's bright enough for your cones to detect.

In summary, not because it's blue.
posted by merlynkline at 11:25 PM on July 19 [4 favorites]


I just read all the comments from the 2013 post about the wine-dark sea. About 90% of the comments were about what "wine-dark" does or doesn't mean, which is very edifying, but also I think is a huge derail. The point is that (so I am told) Homer never described the sea--or anything else--as being blue.

Also, I haven't heard that episode of Radio Lab in years, but as I remember, in typical fashion, they didn't conclude anything about whether the ancient Greeks could see blue. What they did was let whatever expert they were interviewing who thought the Greeks couldn't see blue have the last word. This greatly pissed me off at the time.

I have no doubts that the Greeks of Homer's time could see the color difference between a gray cloudy sky and a clear blue sky. However, as was pointed out on Radio Lab, there are very few things in the natural world that are blue, other than the sea and sky. Ancient people could have gotten along just fine without having a word for blue or thinking about "blue" as a thing. They didn't have a box of crayolas. "Blue" just didn't matter.

I also really like the thing that some interviewee in the Radio Lab story said about what color the ancients would have used to describe the sky. The answer was that they would have said it was the color of emptiness or nothingness. I suppose we can't ever know if that was true, but I think it could be.

Now, I wonder if there's any connection to the fact that blue is de-emphasized in our eyes to the fact that (maybe) Homer didn't have a word that translates to "blue" as we know it. It might be a coincidence or there might be some meandering relationship, and I probably can die happy without knowing the answer.
posted by polecat at 12:18 AM on July 20 [3 favorites]


One effect thatis not accounted for is that, regardless of chromatic aberration, the standard eye is much more sensitive to red/green and to yellow specifically, than to blue. See the graph here

So those two blur operations are not really equivalent at all, and don't really prove his point.

Random related link, Why objects in blue are closer than they appear
posted by haemanu at 2:45 AM on July 20 [7 favorites]


It seems that the figure/background content of the image also plays a role in how we would focus on the colors differentially. "Natural" images like these may usually have figure elements with more green and red components, while background elements are bluer. like the ocean/sky. A better control would be to switch (eg: circ shift) the channels so that all the blue information is in the red channel, and so on.. so that the effect of the color can be disentangled from the features in that channel.
posted by ssri at 3:24 AM on July 20 [2 favorites]


Will read link when I get back, but I am dyyyying to see if this will explain why I can never focus on blue LEDs, they always seem out of focus no matter how I squint.
posted by Iteki at 3:43 AM on July 20 [1 favorite]


The author literally said about the blurred-blue image, "Shockingly, the image appears perfectly sharp!". I disputed that, since at full size it isn't hard to detect a difference between the two images.

Anglophones use "perfectly" to mean both "without flaw or defect" and merely "adequately."

"Eh, it's perfectly fine" doesn't mean that it has some supreme virtue or that it's impossible to detect any flaw in it. It just means that it's not unacceptable and you should stop your bellyaching. Likewise, "it's perfectly cromulent," "that explanation was perfectly clear," and so on.
posted by GCU Sweet and Full of Grace at 4:27 AM on July 20 [9 favorites]


This is the same reason why strip mall stores with lighted deep blue-only signage are very, very hard to read at night.

I thought the same as soon as I saw the premise of this. I think it's PetSmart that has an intense blue sign color?

When I was a kid in the Florida West Coast Youth Symphony, one of the clarinet players had a highly saturated, dark blue clarinet that seriously confused the eyes. Same thing, I guess.
posted by Foosnark at 4:35 AM on July 20


Anglophones use "perfectly" to mean both "without flaw or defect" and merely "adequately."

I'd posit that it's context-dependent. I don't read the 'perfectly' in "perfectly fine" and "perfectly executed" in the same way, for example. "Perfectly sharp" is one of those where I absolutely read it in a more literal way on first blush, and then context clued me in to the fact that the author is just using it as a generic intensifier, like "very." I certainly don't think it's unreasonable to suggest that the phrase "perfectly sharp" usually implies something is in absolute focus, not just good enough.
posted by Dysk at 4:53 AM on July 20 [2 favorites]


But the blue is haloed and blurry

I have shitty eyes, and I especially hate blue LEDs.
posted by mikelieman at 5:34 AM on July 20 [1 favorite]


The idea that ancient Greeks couldn't see the color blue, or didn't have a word for it, is complete nonsense.
posted by Pyrogenesis at 6:00 AM on July 20 [15 favorites]


I'm with haemanu, these images don't really show chromatic aberration in the eye very well. Partly like they said: your eye is less sensitive to blue so you simply see less contrast. But also it's the source image: the ocean is blue, the land is not-blue. But there's very little sharp detail in the blue ocean to see and what there is is white, a lot of not-blue.

I believe the theoretical effect is real but then again if it is surely there's some better articles about it. Something like this or this. I'd like to see some better images, too. Synthetic test images with something like a fine grid in different colors. Also don't do this in a computer's RGB color space, like WTF dude? Use a proper perceptual color space tuned specifically to eye responses. (Although to be fair RGB is kinda close to that; part of the design of RGB is to roughly match human perception.)
posted by Nelson at 6:29 AM on July 20 [7 favorites]


There's one specific shade of blue that to me is like neon orange/yellow reflective... it just stands out and catches my attention for no apparent reason. Hrm.

Once in high school I could not stop staring at my classmate's scrunchie. It was a deep, vibrating lapis blue without shadow or contrast or three-dimensional form, and it felt like a glitch in the matrix. When Vantablack came out, I felt the same thing. Just slithering under the sharpness and edge-detection layers of my brain and delivering deep. blue. void.
posted by fountainofdoubt at 6:57 AM on July 20 [2 favorites]


Well, now I know that my phone camera is inherently frustratingly biased against certain colors and ranges, and I can curse it appropriately. Thanks to everyone, above, for the explanations!
posted by winesong at 9:32 AM on July 20


I'd guess that's because it's emitting a beam which is stronger in the center so when you stand up your eye gets into the brighter part and it's bright enough for your cones to detect.

In summary, not because it's blue.


That's totally reasonable and thank you, but... it's not the first blue light I've experienced this with, and I don't experience it with any other color (though I will have to pay more attention now to see if that's really true). The effect is that the blue "wears out" -- I see it at first, looking directly at it, then it fades to nearly undetectable.

Maybe the blue receptors in the center of my vision are kind of ineffective...? Perhaps I'm discovering I'm a bit colorblind?

At any rate, I'm not ready to completely write off that people are writing about the elusiveness of blue and blue seems to be the color that has this effect on me. Anecdata, single data point, I know.
posted by treepour at 11:42 AM on July 20


one theory was that Greeks didn't have a word for blue, with the suggestion being that humans subsequently evolved to see blue

Homer lived less than 3000 years ago. I'm no evolutionary scientist, but that seems way WAY too short a period of time for humans to evolve the ability to see a new color.
posted by Saxon Kane at 12:18 PM on July 20 [1 favorite]


(Yeah, no, the idea isn't that we evolved the ability to see blue since then.

It's that every language has plenty of useful distinctions that it doesn't bake into its dictionary. English doesn't have a single word that means "your mother's sister." Other languages do, and think it's bizarre that we don't. Ancient Greek might not have had a single word that meant "blue," though other ancient languages likely did, and Modern Greek definitely does now.)
posted by nebulawindphone at 12:33 PM on July 20 [1 favorite]


The idea that ancient Greeks couldn't see the color blue, or didn't have a word for it, is complete nonsense

Thanks! Excellent article--and worth reading the whole thing. It seems like the situation is almost the reverse of what Gladstone claimed. Whereas a modern English speaker would be happy to describe a huge range of the spectrum as "blue", the ancient Greeks had words for smaller sections of the blue range that were not understood to be overlapping.
posted by polecat at 1:32 PM on July 20


the idea isn't that we evolved the ability to see blue since then

Well, I don't think that's what happened, but apparently there are some people who do.
posted by Saxon Kane at 1:53 PM on July 20


I love the color blue (including blue leds and christmas lights) and the differences between the original and the blurred blue were noticeable and very minor but after a second or so of looking at the blurred blue, they both became the same.

Oh well.
posted by The_Vegetables at 2:17 PM on July 20




Blue incandescent lights look blue to me; blue LED lights have a depth and hard-to-articulate quality that I like quite a bit. I seem to recall a lively thread about this in the past, which also talked about LED lights feeling 'sharp'.
posted by theora55 at 3:08 PM on July 20


To visualize just how sparse blue cones are in the eye, check out Fig. 3 of Roorda & Williams (1998).

To an extent, saying this is a sampling effect and not a chromatic aberration effect is misleading. The cones are spaced to achieve approximately the Nyquist Sampling Rate on the transmitted optical field. The optics of the eye make blue blurrier, so it is sampled less. The optics of the eye also make the light outside the central region of the eye blurrier, so in those regions the cones are much larger and thus spaced farther apart.

In my ophthalmic imaging lab in grad school, we would sometimes speculate on how hard-coded these spacings are, or if the cones develop in response to high contrast stimuli. When you are born, the fovea (high resolution central region) does not exist--it takes about 18 months to develop. Thus we wondered: if you gave a baby super-glasses that gave it sharp optics at every angle, would it develop a whole-eye fovea?
posted by Maecenas at 5:26 PM on July 20 [5 favorites]


the ancient Greeks had words for smaller sections of the blue range that were not understood to be overlapping.

And that kinda makes sense, when you consider that colorfast blue dyes were one of the most recent to be invented, and the number of blue things in nature is pretty limited.

It seems possible that the Ancient Greeks potentially just never saw objects (aside from, perhaps a rainbow, or other phenomenon that cause refraction of sunlight) that covered all the possible shades of blue discernible by the eye.
posted by Kadin2048 at 6:14 PM on July 20


Just wanted to mention that our short-wavelength cones (which contribute most directly to our sensation of blue) do not contribute much to our sensation of luminance (the medium "green" cones contribute most, followed by the long-wavelength "red" cones). I suspect that our sense of "sharpness" is more directly related to the spatial resolution of luminance than of either our red/green distinction or our blue/yellow distinction.
posted by Jpfed at 10:34 AM on July 24


Ok after this thread (and apparently in complete contrast with Jpfed above which is really only a coincidence but I’m glad someone else is thinking in the same one as I am), I have been paying more attention to colors and I have noticed that there is a lot more of our visual aids/toys/what-have-you that increase sharpness are blue:

-almost all the new headlights on cars are blue. You can see the difference between the old style yellow bulbs
-lots of sunglasses that are supposed to make the world clearer have a blue tint
-some of the screen/glare protection has a blue tint

I’m sure there are other examples. But it seems like we are overlaying blue onto our visual environment when we want more sharpness. I wonder if the lack of ability to see blue, makes things pop better when they have extra blue…
posted by LizBoBiz at 6:28 AM on July 25


Or what if it’s not extra blue but the reduction of the other colors that we see better and clog up our vision?
posted by LizBoBiz at 9:28 AM on July 25


cool
posted by BeeLIC at 7:09 AM on July 26


« Older I Have No Mouth And I Must Scrum   |   little canada Newer »


You are not currently logged in. Log in or create a new account to post comments.