Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I don't believe the blurred images at the end have anything to do with eye focus, as the author suggests.

After all, chromatic aberration is blurring of only a very, very small amount.

The demonstrated seemingly negligible perceptual effect of blurring blue to a huge degree in a multicolor image doesn't seem to have anything to do with that, but rather the fact that we perceive primary blue as a much darker color than primary red or green, and we perceive differences in lighter colors much more easily.

If the author were correct that we have big problems focusing on blue, then we'd see that blue text against a black background would be massively blurry -- but it's simply not. It's comparatively low-contrast (because blue is a dark color), but it's nearly indistinguishably as sharp as red and green.



Right, I picked the "blurred blue" and "blurred green" pictures, converted them to grayscale using luminance and "blurred blue" still looks sharp and "blurred green" still looks blurry.

If it really was the effect of blue light, the effect should have disappeared by converting to grayscale.

It is well known that luminance matters much more than color when it comes to perceived sharpness. Digital and analog video exploit that by encoding color at a lower resolution (chroma subsampling). And blue only accounts for less that 10% of luminance while green is around 70%. You may find different values because color spaces are a mess but that's the general idea.


That’s not an apples to apples test because of how greyscale is computed. Try swapping the blue and green channels rather than converting to greyscale.


I think for a proper comparison, you would need to swap them by perceived brightness (luminance), not just the RGB value. You can't do that here because the green channel in this image is out of the possible range of luminance you can achieve with blue (in any standard color space, probably).

Which really illustrates why TFA doesn't make any sense- our eyes are less sensitive to blue, so the contrast provided by max RGB value blue is going to be completely drowned out by red and green if the source of your contrast is white on black.

Here's what it looks like with all channels shifted to have the same luminance: https://i.imgur.com/AnKNdfX.jpg - note it is perhaps a little softer as the peak brightness is closer to black.

Now here's what each channel looks like blurred by 2.5px: https://i.imgur.com/XzVWeZg.jpg https://i.imgur.com/LB5JArJ.jpg https://i.imgur.com/jf4ntth.jpg

They're all clearly fuzzy in comparison to the un-blurred image. The answer to "why we're blind to the color blue" is not chromatic aberration (although it could be a contributing factor, maybe even why we have less blue receptors), it's that we're less sensitive to blue and therefore contrast is usually defined by red and green.


Perceived brightness is very tricky as our vision naturally adjusts to normalize perceived colors. https://en.wikipedia.org/wiki/The_dress

Personally I don’t see each image as equally blurred, but YMMV.


There's a difficulty with testing perceived color by using a computer display, which is that the "blue" is the spectrum of the blue pixels, whereas the "blue" receptors in your eye may have a different spectral response.

Where I've noticed weird things with blue are with blue sources that have fairly short wavelengths, such as some of the blue LEDs used in Xmas tree lights, and the old blue lights that were on police call boxes. Both of those are very hard for me to focus on.


>such as some of the blue LEDs used in Xmas tree lights, and the old blue lights that were on police call boxes. Both of those are very hard for me to focus on

This happens to me as well and I though I was becoming blind (to blue light) because of my heavy use of monitors/"white" light, etc...


This article is false. Tried this out Photoshop and not even close.

Here are my results, with a 22px Gaussian Blur on the Channels

https://imgur.com/a/AS0LAyk

IMO The only reason the blue appears a little less blurry is that most of that color is in the water & clouds.

Not in the land masses, with the sharp borders


See the very low coefficients for the blue channel when converting (gamma-compressed) RGB to luma. E.g. the common Rec. 709 standard assigns only 0.0722 weight to blue.

  Y' = 0.2126R' + 0.7152G' + 0.0722B'


Indeed, green contributes more. Example of the Y channel after blurring r/g/b: https://imgur.com/a/3p15Qe1

And colured versions: https://imgur.com/a/Knq2Ue3

(image source: https://en.wikipedia.org/wiki/Flower#/media/File:Flower_post...)


More, as in 10 times more for green, and 3 times more for red. So it's true that we're pretty blind to blue, just the "focusing" explanation is not correct...


Except that our perception of brightness, like loudness, is logarithmic, not linear.

So even if blue is 1/10th or 1/3rd as strong as another color, it's not really a big deal in terms of sensitivity.


I think you're imagining something like a logarithm for each channel, log(R/3) + log(G) + log(B/10). But that's not how it works.

Instead it's like log(R/3 + G + B/10). So when G and B are about the same size, the effect of B will be negligible. It's only when R and G are small that the logarithm will kick in and let you see detail using the blue.

So for seeing detail in a normal picture, blurring the green will have a much larger effect than blurring the blue. But entirely removing the green would let it still look sharp, because we could see the detail using the red (or if we removed that, then blue). If we just blur the green it overwhelms the blue and red to make the picture look blurry even if the blue and red were enough to provide sharp detail if the green wasn't there.


To be honest I need to rethink the arguments of the linked article - if we just use coloured filters in front of our eyes which exclude each channel, the image (I think) remains sharp (or at least that happens with red/cyan anaglyph glasses).


This is basically the same as the Y component used in JPEG, right? Could the phenomenon described in the article be caused by the fact that they used JPEG images? I.e. would we observe the same thing happen with raw/uncompressed images?


No, it's not an encoding artifact.


Did you mean to respond to another comment in this thread where they were talking about YUV? Your comment does not make much sense to me here but would make more sense to me there.


Luma is an approximation of perceived brightness. All the conversion formulae weigh blue substantially less than the other primaries. This supports crazygringo's assertion that "we perceive primary blue as a much darker color than primary red or green".


It makes sense to me here - GP discusses how we see perceive blue as a dark color, and parent comment corroborates that with a low luma coefficient for blue.


I dunno, but when I look at glowing blue signs and blue Christmas lights at night, they look significantly more fuzzy than other color lights.


Yes! I have difficulty with some blue signage and especially with certain intense blue LED lights as they seem to vibrate to my eyes. I also find it hard for me to read the text in blue on the small, scrolling LED signs common in storefronts.


> At long last, we can see why the human eye can't focus on blue light

At long last indeed. For decades I wondered why blue lights are disturbingly blurry at night, to the point I'd rather not look at them. I always thought it was just me since other people didn't seem to care as much.


I’ve found my people! I have similar issues and nobody I’ve ever spoken to about it seems to have the same experience.

Two similar issues I’ve had, and I’ve wondered if they’re related, are:

- at conferences with very large overhead projection sometimes the setup produces an effect where each time I blink it’s like it separates the RGB elements. It’s like a combination of being able to see the refresh rate (like when a CRT was filmed out of sync) and the colours being projected out of alignment. - a stadium near me as those digital display advertising signs around the sidelines. If I’m not looking directly at them they appear to flicker. Which actually makes watching a game at that venue not enjoyable as watching the action means I have a permanent shimmering right on the periphery.

Does anybody know what’s happening with either of these?


That's most certainly the display sequentially flashing red, green and blue sequentially in time to approximate shades and hues. When you stare right at it, your eye averages the photons with a time constant of 1/30th of a second or so, and so you don't perceive the flicker. When you move your eye or blink, though, individual portions of your eye are exposed to only relatively narrow time slices, and so re only exposed to the red, green or blue photons.

Cheaper DLP projectors use a single light source and mechanically spin a color wheel with alternating red, green and blue filters. They look great when staring at it, but if you move your head or wave your hand in front of it, you can easily see the three color channels.

Perhaps those stadium displays are DLP projectors based, or maybe they're RGB LED and are simply PWMed at a relatively slow rate. Most LEDs can be switched very fast, at say 10Khz, but maybe there's electrical limitations of building such a large high brightness display. If it's only PWMing at 100-200Hz, you'd see similar effects. In particular, each color channel will be on for different duty cycle durations, and LEDs are very fast to turn on and off. So, when you move your eyes or blink, you'll once again get separation of the channels in your vision.

You can do a similar trick with your smartphone camera. Record video, and point it at the display then wiggle the phone up and down and side to side. The phone most certainly has a "rolling shutter" which means it captures an image sequentially in lines either horizontally or vertically. It does it quickly, but slow enough that different lines should be able to pick up colors. You may not even need to shake the camera up and down to see a funky image. It's the same reason why CRT monitors and TVs look funky on video but not film.


Thanks for the explanation!

I can certainly perceive all those flickering effects when moving the eyes. Slow PWM is hell, but some stuff like colour separation can be cool too.

My favourite trick is making a digital clock's numbers "slide" over the clock's surface, in a sort of parallax way. I guess that's due to low refresh rates, so for a very brief moment there's a disconnect between the clock's physical position and the last know position of the digits.


I totally know that digital clock effect you're talking about. It's caused by something else though. It's actually a physiological hallucination due to high power ultrasonic waves coming out of those types of clocks. Rather than smoothly turning a small crank at 60rpm like in an analog clock, the clockwork gnomes inside digital clocks have to frantically push and pull a lever at 32.768KHz, and their stiff little pointy silk hats make perfect tweeter elements. It's also why those clocks get so warm, and why the snooze button is so unreliable; the gnome gets a well deserved nap, but is so exhausted he sleeps through his own alarm...


Oh... What? Hallucinations from a 32kHz RTC? Can't tell if your entire comment is a joke or just the last half... Couldn't find a thing after a quick search but now I'm curious.


None of it was serious. :-) Although I have indeed noticed that particular optical illusion of the clock digits moving relative to the clock.


I hated blue christmas lights because of this. Also, I couldn't read the time on the stove from any reasonable distance since it used blue LEDs.

After LASEK, though, I can see blue LEDs nearly as clearly as everything else. My eye surgery gave me nearly 20/10 vision and the greatest thing I got from it was the ability to read the stove clock from across the room. Lol.


I had a car with blue and red illumination in the dashboard. It always felt like I couldn’t see it properly - like it was blurred.

While I like the colour combination, my next car had orange displays, which always look sharp.


Also the blue LED on a PS3. Those always looked blurry to me. It’s especially obvious because of the adjacent green LED.


You are so right! I always assumed it was the LED's problem until I started reading all these comments. I think I remember the PS2 having the same thing.


Does this only affect people with Blue eyes ? As someone with utterly black eyes, I don't have a problem looking at glowing blue signs and don't find them fuzzy either.


Eye color doesn't matter in vision, it's not part of the optical path, its role is the same as an aperture on a camera lens.


Ok it seemed to me that that the stroma of blue eyes scatters back more light: https://medium.com/@ptvan/structural-eye-color-is-amazing-24...


Yes, scatters it back out — that light does not enter the pupil or contribute to what that person sees (save probably a vanishingly small amount that is re-scattered in the cornea, which is totally de minimus).


I have mostly brown eyes, so no?


It's true that we don't focus well on blue light sources, but it's also true that this web page does not demonstrate that fact in the slightest.


It also matters a bit that the blue channel is only 2% of all color-sensitive cones in the retina. That has a lot more to do with poor spatial resolution int the blue channel than the optics.

http://hyperphysics.phy-astr.gsu.edu/hbase/vision/rodcone.ht...


Chromatic aberration may be a contributing factor, but I am surprised the author didn’t mention that S cones (which we use to perceive blue) are only 2% of the cones in the retina [1]. Additionally S cones are distributed randomly when compared the regular lattice of M and L cones. The distribution of the different cone types alone may be sufficient to explain why our acuity for blues is impoverished relative to reds and greens.

[1] https://en.m.wikipedia.org/wiki/Cone_cell


This lower resolution of blue is pretty well known in recent image compression work (XYB space of JPEG-XL and guetzil), and number of S cones is the only explanation I have seen on that.


Also, S cones are mostly found outside the fovea.


I'm not sure you're right. At night both I and my wife have reported difficulty reading glowing blue signs compared to glowing red/green signs at the same font size, brightness, and distance.

I'm also not sure that the author is correct; the wrong-focal-distance explanation seems rather weak simply because our focal length is adjustable.


You have astigmatism. I have a similar issue with blue when not wearing my glasses. I have 20/20 vision, but my astigmatism makes it difficult to focus on certain things. A computer being a big one. Blue light blockers help, but with proper astigmatism correction I don’t need them.


20/10 vision in one eye, 20/13 in another, no astigmatism.

Blue LEDs in clocks when viewed at night look completely fucked up.


I’ve never met anyone without some astigmatism; especially, people with other vision impairment. Astigmatism is the number one cause of night blindness which is what you described.


HN is so funny. 'I have no astigmatism.' -- 'Sure you do!'


It's right up there with 'I've been clinically diagnosed with ADHD by a qualified psychiatrist and this medication has improved my life tremendously' -- 'There's no such thing, the medication doesn't do anything and you just need to mediate more, have you tried mindfulness?'


There were three people involved in that exchange, not two, assuming nobody's using multiple aliases...

crdrost described their vision problems, jbluepolarbear said "you have astigmatism", techrat popped in and said "[I have] no astigmatism" and jbluepolarbear made a more general statement that most people do.

I'd assume techrat doesn't know anything about crdrost's vision either.


Because nearly every person has some amount of astigmatism.


Sure, but presumably the parent has been checked by their eye doctor and indeed, not everyone has it.

Signed, a dude with very slight astigmatism :)


pro tip: don't tell other people about their own experiences unless they ask you, and you're a professional.


Pro tip: nope


You're precious. Why don't you tell me what else my eye doctor has never diagnosed me with?


9 in 10 people have some affliction of astigmatism. Astigmatism is just that your eye isn’t perfectly spherical. What does your eye prescription say in the cyl and axis fields, that’s your astigmatism correction.


20/10 and 20/13 means much better than average good vision. It would be bad if they were e.g. 20/30. Funny that you're LARPing an eye doctor without knowing this.


You can still have 20/10 vision with Astigmatism. I guess my years of research into astigmatism because of my own affliction and years of optics study visual effects means I’m larping. Never said I was a doctor, I’m just intimately familiar with how corneal distortion.


'eye prescription'?

Bruh. Read the comment of mine you replied to. 20/10 and 20/13 vision. I don't need prescriptions.


Aren't those numbers about your ability to focus on distance objects, though?

I could totally believe the someone might have "perfect" vision that doesn't require correction, but still have a slight astigmatism that impacts their vision under certain specific scenarios such as when viewing blue LEDs in low light.

I'm not necessarily saying that's what you have, but more just that to the extent your eyes have been evaluated, it was likely "yeah they look great as far as your ability to perceive the brightly lit eye chart, no need to do the more detailed analysis where we figure out the other parameters that will never be used because you're fine, bye."


That's like saying you can run a 3 minute mile so you don't have a shoe size.

This whole thread is silly.


Shoe size is an inherent byproduct of having feet of a certain dimension. A prescription has to be prescribed.

If we're doing dumb analogies, it's like you're trying to insist a homeless man must have an address.


...Actually, that's a real problem in the financial services sector. You can't satisfy KYC requirements without an address of record.

Off topic, but I'm jumping around a bit tonight.


i'm imagining a homeless man receiving his begging money via a bitcoin QR code on his phone.. i hope he has a bitcoin atm nearby.


I think more than one person has confused you with the person a couple levels up thread, i.e. https://news.ycombinator.com/item?id=27859324


> About one in three people has some degree of astigmatism.

https://my.clevelandclinic.org/health/diseases/8576-astigmat...


> about 9 in 10 people have some degree of Astigmatism

https://patient.info/eye-care/astigmatism


I wonder why there's such a big degree in difference between the two claims.


The 30-60% is most likely for those affected by their astigmatism. Astigmatism is when the eye isn’t a perfect sphere. Many people have very slight astigmatism and don’t require correction.


> In Europe and Asia, astigmatism affects between 30 and 60% of adults.

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


I can’t actually verify that number presented in that paper. I’ve looked and that stat is attributed to that paper, but there’s no evidence that the paper said that. Plus the paper is pay walled so unverified.


Considering the obvious inability for anyone to reliably know the astigmatism status of everyone they meet, not only is your anecdata unverifiable it's completely absurd.


I don't think it's night blindness. I have the same problem, with and without glasses (I have astigmatism).


Astigmatism is easy to self-detect, and it's quite different from the light focusing to the wrong plane. Astigmatism will noticeably warp the entire shape of the distant point object, rather than just make it "fuzzy" in an isotropic manner.


Source? Astigmatism can blur and distort.

https://patient.info/eye-care/astigmatism


for me, it's mainly the difference in looking up and down compared to looking left and right.


You're not alone. I normally have excellent night vision but seeing things in glowing blue, such as the clock on the coffee maker or the microwave, causes the digits to split like double vision and become blurry while everything else remains the same.


I have the exact same thing. It always makes me wonder why companies choose to have blue neon lights on their buildings because it's nearly impossible to read them when it's a thin font.


I mostly agree with you, but would add that blurring the blue is affecting the sharpness of the ocean, which has little detail in that image; blurring red or green affects details on the land, which are very noticeable. One might think the cloud-ocean edges would be blurred by the blurring of blue, but the clouds are so much brighter than the ocean (red & green channels), that you can barely notice any difference.


I agree. Further, almost all of the blue channel in this image is from the white clouds or nearly black water. There's no other major source of blue.


What would be the result if you flipped the colors, so that the sharp elements are blue and the ocean is red or green?


That's an interesting thought but unfortunately I'm travelling.

One thing is that the ocean is hardly blue. Not sure why my eyes register it as such, but it's mostly very close to black with nearly equal parts of red and green, at least the parts I sampled. I think a certain amount of this article's claim is predicated on the reader erroneously believing the ocean should become blurry.


I think that the title is a massive oversimplification because chromatic aberration by itself is not enough for us to be blind to the color blue. We do have cones that can detect blue for one.

We are however, less sensitive to it so maybe the eye doesn't focus based on that channel(?).


I wanna point out that LCA cannot be responsible for e.g. blue displays being basically impossible to read. Why? Because they are still impossible to read when they're the only thing that's around, and blue LEDs are very monochromatic. So the eye would have to focus on the blue light, which would make LCA go away.

I suspect that a plausible cause could be that there just aren't a lot of blue receptors in the retina, as the eye is pretty insensitive to blue overall.


I’d love to see the same comparison across a range of images.


Exactly. In the real life I find very difficult to focus on violet signs against the black night background, but it’s a completely different effect from the one in this article and it’s exacerbated by the bigger pupil diameter in the night that increases the aberration.


I agree, in that it was pretty easy to dismiss effects on the example image. Doing this with shapes and a variety of hues and luminances would be a better way to prove the point if it bears out.


Maybe the (evolutional) reason that we perceive blue as darker is exactly that it would otherwise reduce the sharpness of what we see.


It has a whole lot to do with our eyes having far fewer blue receptors.


I'm sure you could find an image where blurring blue ruins it, and blurring red and green have no impact. This feels like cherrypicking especially given how trivial it would be to just show a bunch of examples.


The author is correct. Lighting designers sometimes use the effect deliberately to create blue spotlights that viewers can't focus on.

That description doesn't do it justice - you have to experience it to appreciate it. It's very striking and slightly surreal.


They might be correct about chromatic aberration and the difficultly of focusing on pure blue, but the conclusion from their experiment is completely wrong.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: