Theoretical limit to human vision.

For the discussion of the sciences. Physics problems, chemistry equations, biology weirdness, it all goes here.

Moderators: gmalivuk, Moderators General, Prelates

blademan9999
Posts: 40
Joined: Sun May 01, 2011 5:18 am UTC

Theoretical limit to human vision.

Postby blademan9999 » Sun Sep 18, 2016 4:57 am UTC

What's the theoretical list of vision for a human size eye?
Say people start replacing their eyes with bionic eyes what's the theoretical limit for how good they can get.
20/8, 20/5, 20/2, better?
http://officeofstrategicinfluence.com/spam/
That link kills spam[/size][/b][/u]

User avatar
Copper Bezel
Posts: 2416
Joined: Wed Oct 12, 2011 6:35 am UTC
Location: Web exclusive!

Re: Theoretical limit to human vision.

Postby Copper Bezel » Sun Sep 18, 2016 9:22 am UTC

Better, with some caveats. I don't know where to start to approach this quantitatively, but it's worth noting that "20/20 vision" and the like are a relative scale compared to typical vision, on the one hand; on the other, the scale really does only account for angular resolution, how small of a detail (in angular terms, so independent of distance effects) you can resolve, and there might be other aspects of vision you're concerned about.

The way the eye is "designed", the fovea, the section of the retina under the center of the projected image, has a much higher resolution of receptors than outlying areas. (I'm trying to avoid the word "focus" when I don't mean it in an optical context here, but the fovea corresponds to the area of the image that your eye is actively targeting.) So the angular resolution across your visual space isn't at all uniform. Meanwhile, our "grayscale" perception is both higher in resolution and much more sensitive than our perception of color, and a lot of color information is interpolated or even simply assumed. You can see this playing with a bit of text in a desktop publishing app: put some text on a field with the same value, say 50%, but different hues and saturations, like a bright red on a bright green, and it'll become unreadable at scales where black-on-white or even contrasting grays will still seem clean and crisp. This is, again, different from how we think of a typical camera; if you do nighttime photography with a DSLR, you'll notice that there's actually color at night, we just can't see it. = . (In street lighting, that usually means a mix of bright blues and yellows that look like some kind of dance party.)

Meanwhile, animal eyes also vary in terms of the angle of the cone they actually take in, and a more "telescopic" eye (with a longer focal length) is naturally going to see a smaller section of the world at a higher resolution.

Since we're talking "theoretically", which means we can propose damn near any level of technological wizardry, the ultimate limits are going to be biological ones, how much information the optic nerve is able to process in what way. So in that context, you can have optically stabilized superzoom eyeballs, but the overall resolution across your entire field of view is never going to be more than a couple of times better than it is (a difference owing to optical imperfections in the eyes that originally came with your head.)
So much depends upon a red wheel barrow (>= XXII) but it is not going to be installed.

she / her / her

User avatar
pogrmman
Posts: 326
Joined: Wed Jun 29, 2016 10:53 pm UTC
Location: Probably outside

Re: Theoretical limit to human vision.

Postby pogrmman » Sun Sep 18, 2016 8:38 pm UTC

blademan9999 wrote:What's the theoretical list of vision for a human size eye?
Say people start replacing their eyes with bionic eyes what's the theoretical limit for how good they can get.
20/8, 20/5, 20/2, better?


20/20 is average vision by definition.

The average eye's angular resolution is around 1 arcminute.
Assuming these bionic eyes didn't have pupils larger than our own 7 mm pupil and were perfectly diffraction limited, they'd have an angular resolution of ~0.000087 radians, which is around .3 arcminutes.

On a 20/20 scale, that would be around 20/7 -- assuming Wikipedia is accurate in that the average eye (which should have 20/20 vision) has a resolution of 1 arcminutes.

If you could let the entrance pupil of the bionic eye be larger, you'd get better resolution. For instance, if you let the eye have a pupil of 8 inches (204 mm), you'd have a resolution of .0084 arcminutes, which would correspond to 20/0.16 vision!

User avatar
doogly
Dr. The Juggernaut of Touching Himself
Posts: 5212
Joined: Mon Oct 23, 2006 2:31 am UTC
Location: Somerville, MA
Contact:

Re: Theoretical limit to human vision.

Postby doogly » Mon Sep 19, 2016 12:29 am UTC

I thought 20/20 is healthy, and there are way more people with worse than better, no?
LE4dGOLEM: What's a Doug?
Noc: A larval Doogly. They grow the tail and stinger upon reaching adulthood.

Keep waggling your butt brows Brothers.
Or; Is that your eye butthairs?

User avatar
Angua
Don't call her Delphine.
Posts: 5652
Joined: Tue Sep 16, 2008 12:42 pm UTC
Location: UK/[St. Kitts and] Nevis Occasionally, I migrate to the US for a bit

Re: Theoretical limit to human vision.

Postby Angua » Mon Sep 19, 2016 7:40 am UTC

It is based on an average 'normal person', however, not that many people are better than average, whereas many are worse.

I don't know how did the statistical studies and how to define 6/6 / 20/20 vision.
'Look, sir, I know Angua. She's not the useless type. She doesn't stand there and scream helplessly. She makes other people do that.'
GNU Terry Pratchett

User avatar
Eebster the Great
Posts: 2752
Joined: Mon Nov 10, 2008 12:58 am UTC

Re: Theoretical limit to human vision.

Postby Eebster the Great » Mon Sep 19, 2016 11:09 am UTC

If you define "normal" people as people without diagnosed vision problems, the median and mean will both be significantly better than 20/20. That acuity standard is not an average in any sense and certainly is not perfect. It's just a convenient but fairly arbitrary standard for comparison that corresponds to the ability to discriminate gaps of one arcminute with high contrast from some distance away.

There are people with 20/10 vision, so that is at least possible. The maximum size of pupils does vary somewhat between individuals, but of course wide pupils introduce their own problems like chromatic aberration. This can be corrected by the retina to an extent, but some information is lost. Typical human vision is diffraction-limited with a pupil size of around 2.5 mm. A technologically superior eye could improve acuity by dilating the pupil further, whereas human eyes do not improve acuity by dilating the pupil because (among other things) the density of foveal cone cells is not high enough. Eagles have denser foveas and so have better acuity with larger pupils, though some claims about their vision are exaggerated.

If we suppose a person with 2.5 mm diffraction-limited pupils has 20/16 vision, which is probably somewhere in the right ballpark, then a perfect eye with a 10 mm pupil should have 20/4 vision, and that pupil is small enough to fit comfortably in a human eyeball (which frequently accomodates pupils up to 9 mm in diameter). If you want to get too much bigger than that, you might have to consider using a larger eye.


Note however that in ideal conditions it may be able to distinguish between points that do not meet the Rayleigh criterion. Although two points that just barely or almost meet the Rayleigh criterion will be slightly "blended together," with sharp enough contrast and accurate enough image processing, it is still possible to determine that they are distinct. The actual optimal human vision appears to be slightly better than the Rayleigh criterion when the pupil is small, so our vision in that sense is "better than perfect." A more accurate limit for human eyes is Dawes' limit, a credit to the computational power of our retinas. Allowing for even better image processing, this limit can be pushed back even further, and I'm not sure there is any theoretical limit to what you can do with bright enough light, long enough exposure, a distant enough source, and a powerful enough computer.

jewish_scientist
Posts: 660
Joined: Fri Feb 07, 2014 3:15 pm UTC

Re: Theoretical limit to human vision.

Postby jewish_scientist » Mon Sep 19, 2016 12:54 pm UTC

I think that in terms of optics, I think that the limit is the size of the pupil. The smaller the pupil is, the sharper the image; the drawback is that less light will travel through the pupil, which means less light for the retina to detect. The same principles are behind pinhole cameras.

Actually, I just did some quick calculations and I think that the human pupil is already pretty close to the optimal size, which is pretty cool when you think about it.
http://www.ncbi.nlm.nih.gov/books/NBK381/
http://www.huffingtonpost.com/2013/12/2 ... 41884.html
http://www.wolframalpha.com/input/?i=1. ... 0055+mm%29^%281%2F2%29

User avatar
Eebster the Great
Posts: 2752
Joined: Mon Nov 10, 2008 12:58 am UTC

Re: Theoretical limit to human vision.

Postby Eebster the Great » Mon Sep 19, 2016 1:50 pm UTC

A smaller pupil has a greater depth of field, but that doesn't matter much when your lens and ciliary muscles are working properly, because you can simply shift your focus by deforming the lens. What matters for acuity is the size of the Airy disk, which increases as the size of the pupil decreases. This is why everyone has been saying that larger pupils can theoretically provide better vision and why eagles do in fact have (slightly) larger pupils.

Something not mentioned yet is that the size of the Airy disk also depends on wavelength, which means diffraction-limited vision ought to be better for violet light than red light, by almost twofold in extreme cases. I'm not sure this is actually realized though.

ijuin
Posts: 799
Joined: Fri Jan 09, 2009 6:02 pm UTC

Re: Theoretical limit to human vision.

Postby ijuin » Mon Sep 19, 2016 4:00 pm UTC

The wavelength limitation also increases the Eagle's advantage over mammals, since most large predatory birds can see into the UVA band.



Return to “Science”

Who is online

Users browsing this forum: No registered users and 13 guests