Take the 2-minute tour ×
Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. It's 100% free, no registration required.

I am not a physiologist, but whatever little I know about human eyes always makes me wonder by its details of optical subtleties. A question always comes to mind. Are human eyes the best possible optical instrument evolved by natural selection? Is there any possible room for further major improvement as per the laws of physics in principle? Can there be any camera prepared by human beings which can be superior to human eyes?

share|improve this question
Owls? Hawks, etc.? Also, detection by human eyes of infrared, ultraviolet, Terahertz, gamma radiation, etc. is terrible. –  Peter Morgan Apr 14 '11 at 15:34
Not a physics question. Only a vague allusion to physical limits in the question, and nothing at all in the answers. At a minimum the optic nevre attaches in the wrong way. Squid eyes are similar in a lot of ways, but don't have a blind spot because the optic never attachment is better. –  dmckee Apr 14 '11 at 16:39
With the best possible camera, you should be able to see radio waves too, thus seeing music –  user1708 Apr 15 '11 at 1:00
It is a good idea that you are recognizing the power of natural selection, but I feel human-centrism, ego-centrism in your question. Why didn't you come up with an idea that the eyes of other animals might better? –  sawa Apr 15 '11 at 1:01
Answer-in-an-image: infoaddict.com/fileadmin/Images/Television/Star_Trek/… –  Sklivvz Apr 15 '11 at 13:50
show 1 more comment

6 Answers

up vote 34 down vote accepted

This question is sort of difficult to answer in an objective way, because it depends very strongly on your definition of "best." Natural selection favors traits which provide a reproductive advantage; no more, no less.

Could our eyes be better by the standards of modern optical design, in terms of precision and features? Sure. I could easily design a camera with a larger aperture, better resolution, less abberation, a broader wavelength sensitivity, etc. There are even some precedents for this in nature. Some animals can see ultraviolet or infrared; some can even detect polarization (mostly birds I think?). I believe there are some fish that can swivel their eyes around through nearly a full circle in any direction. Cats and dogs have higher sensitivity in low light due to the reflective layer behind their retina, and squid have eyes with truly huge apertures.

It is worth pointing out though that the human eye, despite what I've said above, is still a pretty good camera. It can re-focus pretty quickly, it's got a fantastic image processing computer attached to it that can do incredible amounts of pattern recognition, noise reduction, and image stabilization. While its wavefront aberration is not perfect, it is pretty good for a biological system. They have very low chromatic aberration, and are quite compact. Additionally, our eyes have a curved focal plane (our retina) which is a lens designers dream -- it eliminates the problem of field curvature, which is an inherent aberration in any optical system that is nearly impossible to eliminate.

share|improve this answer
Thanks for your answer which in my opinion is the most balanced. +1 and acceptance. –  user1355 Apr 14 '11 at 15:55
Eyes are a terrible optical design, but like most crappy bits of engineering you just fix it all up in software –  Martin Beckett Apr 14 '11 at 16:05
@Martin: Lol, that's definitely true. The characterization and calibration process for human eyes takes years. Far longer than even the most complicated space telescopes. :) –  Colin K Apr 14 '11 at 16:11
BTW, insects are the polarization kings. They can almost all detect polarization in some form or another. –  PearsonArtPhoto Apr 14 '11 at 18:06
@Pearsonartphoto - Mantis shrimps have a linear polarizer band across their eye but can rotate the entire eye so the have a complete polarimeter. –  Martin Beckett Apr 15 '11 at 0:11
show 4 more comments
  1. This is a common misconception -- evolution has not stopped a million years ago leaving all the creatures in the "best possible state"; it is a continuous pursuit of adapting to the current environmental conditions, with only aim in reproductive success. Moreover most of the population stays in even more suboptimal surroundings due to random mutations (the advantage of that is to achieve faster adaptation).

  2. Yes; even a huge one, see Georg's answer. Squids have eyes without blind spot and with no risk of retina detach. Bees see light polarization and some UV. Cat's eye has a mirror that doubles the signal strength. Etc, etc.

  3. Human eye has about 2Mpixel resolution -- my phone's camera is better. Of course this is not a fair comparison, since the most usability of a sight stays in the brain, but if we talk about the eye itself...

share|improve this answer
@mbq: Thanks for your answer. I've not stated that evolution has stopped. I just asked if any major change can be thought of in principle. And yes your answer is valuable. So +1. –  user1355 Apr 14 '11 at 15:41
Squid (or at least some of them, such as octopus and cuttlefish) can also see light polarization). +1 for pointing out the misconception re evolution –  Lagerbaer Apr 14 '11 at 15:42
@mbq: Just one nitpick: I'd be really careful assigning "purpose" to anything evolution related, even with the quotes. I know this isn't biology.SE, but I think we should do our fellow scientists a favor and avoid confusing the public about a topic which is already drowned in disinformation. Nothing about evolution has any goal or purpose. Mutations happen simply as a consequence of the world we live in, and natural selection allows beneficial mutations to propagate. There is no goal, nor any guarantee of optimality. –  Colin K Apr 14 '11 at 15:50
@Colin that's why the quotes... but I'll change it. –  mbq Apr 14 '11 at 15:59
@mbq: Yeah, I figured we were on the same page already. Like I said it was just a nitpick. Very good answer overall. –  Colin K Apr 14 '11 at 16:08
show 6 more comments

[Updated to correct a couple of mistakes pointed out in comments. Thanks!]

At my age, it's clear that there's room for at least one major improvement: more accommodation. Accommodation, in this context, means the ability of the eye to focus at different distances. This is accomplished by changing the shape, and hence the focal length, of the lens. The lens material gets stiffer over time, so that most people lose the ability to focus on nearby objects as they age. In addition, of course, many people, including me, lack the ability to focus on faraway objects.

If I were in charge of natural selection, I'd make those muscles strong enough to let me focus on nearby objects. Presumably, back in the savannah, the decreasing ability of middle-aged people (or, as they were known back then, very old people!) to focus at objects less than a meter away had little effect on the number of descendants those people produced.

Another natural figure of merit to use to assess any optical system is how close it is to the diffraction limit -- that is, how close it comes to attaining the best possible resolution for a system of its size. I'm not an expert, it looks like the eye is within a factor of a few of the diffraction limit (example), so there's a bit of room for improvement there, but not much.

Another thing to look at is the quantum efficiency, effectively the fraction of all incident photons that an optical system uses. Again, things depend a lot on conditions, but in general the eye's very good on this score, reaching levels of about 10%. (That may not sound so great, but I think it's hard to imagine ways of improving it dramatically.)

Another way to improve the eye would be to boost the range of frequencies it could see. Going into the ultraviolet probably wouldn't do much good, but if you added some infrared receptors on your retina, your night vision would be much better.

share|improve this answer
I could be wrong but I believe that the loss of accommodation in middle age is due mostly to the stiffening of the lens tissue, not a loss of muscle strength. I'm not certain of this, and it's an insignificant detail either way. Good answer. –  Colin K Apr 14 '11 at 15:46
That sounds very plausible to me, and various Google hits seem to back you up. I've misstated this many times over the years, then! Thanks for letting me know. –  Ted Bunn Apr 14 '11 at 16:31
"the decreasing ability of middle-aged people (or, as they were known back then, very old people!) to focus at objects less than a meter away had little effect on the number of descendants those people produced." Hmm. There is also the advantage to the tribe of less non productive mouths to feed,when the bear or the lion gets the elder due to his/her bad vision. –  anna v Apr 14 '11 at 16:32
I disagree with the 10x worse than diffraction limit. You can do the experiment yourself: draw two closely spaced lines on a paper and see how far away you can get before you can't tell they're two lines anymore. I got (while wearing my glasses) an answer quite close to the diffraction limit. Heck, even the link you give claims that for people with good eyes it's within a factor of 2, not a factor of 10. –  Anonymous Coward Apr 14 '11 at 18:04
I think you're right. I was referring specifically to this statment: "So 20/20 is about 12x the Rayleigh criterion." But actually, now that I think about it, that's not a fair comparison. The eye test requires you to "resolve" the letter in the sense of telling it apart from other letters, which is a stronger criterion than just telling two lines from one line (which is, as you say, what the Rayleigh criterion is all about). I'll edit my answer accordingly. Thanks! –  Ted Bunn Apr 14 '11 at 18:41
show 1 more comment

This question can’t be answered exactly properly. The eyes we humans have evolved to be “about” the best available to us hominids. There are other species of life with eyes better adapted for their econiches. Eagles and hawks have eyes capable of far better resolution at great distances. They can soar 500 meters above the ground and spot a small animal for prey, where upon they dive from that height to snatch it. Our eyes are nowhere near that sharp. On the other hand their eyes may not be as adapted for fine resolution of very close vision.

Human eyes have some funny aspects to them which illustrate how evolution selects for adaptive feature built from previously existing structures. For instance, the optical neuron innervates the eye from the front, which is quite the opposite of what one might expect. It would be as if one built a CCD with the pixel connections connected in the optical path.

The best answer which can be given is no, but the eye is evolved to work within the limits of what is biologically adaptable.

share|improve this answer
add comment

The human eye is an indifferent camera. Compared to a hawk's, it is laughable. A hawk can sit on top of the Empire State Building and make out a dime on the sidewalk.

What makes the human eye any good at all is the software behind it: what our brains provide. You do realize that what really focuses on our retinas is an upside-down, tunnel-vision image that our brains sort into the images that we think we see, right? And that in experiments, people who were made to wear glasses that made everything look upside-down adapted in a few days so that upside-down looked right-side up?

We recognize patterns, colors, shapes where cameras only record differences in color and tone. It is not the eye that is remarkable. It is the brain.

share|improve this answer
I think that the eye of "the hawk" is a question of data processing too. –  Georg Apr 15 '11 at 9:21
@Georg: To some extent, surely. But all the data processing in the world can't improve the basic resolution of each lens and retina combination. The hawk's optics are simply better suited to seeing small things at a distance. Similarly, a cat's eyes see better in low-light situations because of physical, not neural, characteristics. We can prove this by taking various forms of human color-blindness into account. If software were the issue there, the brain would compensate. But it hasn't, because the cause is physical. –  Robusto Apr 15 '11 at 13:14
add comment

a few different answers:

  • no, it's easy to imagine small improvements: more regular lenses, wider spectrum sensibility, more than three color sensors...

  • not only not the best possible, but also have some fortuitous bad choices, the most commented is the retina orientation that makes it easy to detach with a big impact. the very similar but independently evolved squid eye has the retina backwards, making it a lot more resistant. unfortunately there seems no way for evolution to flip it without many blind generations in between. obviously not a good evolution path...

  • best for what? for producing images that will be seen by whom? IOW: if the end observer is a pair of human eyes, then it will always be better to observe directly than to use an intermediary, no matter how high the quality/resolution/contrast/etc of such an intermediate reproduction. (unless you want to see things not visible to the naked eye)

share|improve this answer
add comment

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.