Resolution of the Human Eye

[Forever Young]

I am wondering about the complaints of people who say that 22 MP is not enough? What is the resolution of the human eye? To see on a photograph what I see with my eyes, do I need more resolution?

What is the resolution of the human eye?

What is the resolution of the current printers? I guess much less than 22 MP.

What is the resolution of the monitors? I guess much less than 22 MP at A4 or at most A3 format usual at home?

--
Thanks, Imre

 

[Sante Patate]

You are right - but the concept of "resolution" is different for the eye.

There are over 100 million photoreceptors in the eye, but only about 5 million are the "cones" (as opposed to "rods") which are concentrated in the fovea are account for what you think of as visual resolution - what you see when you look directly at something. What you see in the dark, or as peripheral vision, is completely different. It is even more complicated because a lot of what you see consciously is actually post-processing in the brain - which is why we are so prone to optical illusions.

Plus, because of the way human colour vision works there is no single value for resolution in relation to coloured real-world scenes.

 

[hunk]

Let's say the eyes render 16MP, the thing is, you can point these 16MP to a tiny piece of your photograph. The eyes zoom in.

The resolution of printers? Well, mostly 300-360 dpi. But I print files that end on trains and busses. Huge billboards that end up in waitingrooms etc. And yes, the picture on a bus is looked at from 1 meter distance as people tend to stand next to a bus. So yeah, I'm happy with more pixels.

Another thing. More MP are much easier for retouche / postproduction. It's easier to make masks etc. on files with more pixels.

I am wondering about the complaints of people who say that 22 MP is not enough? What is the resolution of the human eye? To see on a photograph what I see with my eyes, do I need more resolution?

What is the resolution of the human eye?

What is the resolution of the current printers? I guess much less than 22 MP.

What is the resolution of the monitors? I guess much less than 22 MP at A4 or at most A3 format usual at home?

--
Thanks, Imre

--
fun; http://www.x32.nl
work; http://www.ministerievanbeeld.nl

 

[Darrell Spreen]

I am wondering about the complaints of people who say that 22 MP is not enough? What is the resolution of the human eye? To see on a photograph what I see with my eyes, do I need more resolution?

What is the resolution of the human eye?

The optical resolution of the eye is, at best, about 15 LP/mm. However, the eye is always scanning very rapidly and the brain constantly integrates the information, resulting in much higher apparent visual acuity.

It's just not realistic to expect a photograph to match what you see with your eyes. Whether 22MP is enough is more a matter of what level of detail is needed for the intended display and viewing conditions for the photograph. Too many variables.

--
Darrell

 

[GURL]

What is the resolution of the human eye?

For a recorded image this depends on the distance between your eyes and said image.

This results in various discussions where the distance from witch you could or should look at prints hanging on a wall is the matter. For an image on a computer screen, provided the distance is such that you can't see the screen colored dots, the discussion is just the same.

As I'm quite short-sighted and like wide angle images I use my own rules but would not suggest to others that they should follow them...
--
Georges Urbain René Lagarde

 

[GuyMcKie]

Look here

http://clarkvision.com/imagedetail/eye-resolution.html

Gr.

 

[Vlad S]

I am wondering about the complaints of people who say that 22 MP is not enough? What is the resolution of the human eye? To see on a photograph what I see with my eyes, do I need more resolution?

In the end, extra camera resolution simply allows one to print larger (assuming we compose to avoid cropping).

I think the most practical rule is that for hand held prints we do not distinguish additional detail past 300 dpi.

I think that if you take that as the criterion for the largest size print that you can produce, this rule will allow to determine the largest size of the print, that would not only be viewable from some distance away , but that would also allow to approach to a close distance if you want to see more detail.

So decide how large you want to print, then calculate necessary camera resolution at 300 pixels per inch of print. That's how many pixels you need to see in the photograph what you see with your eyes.

Of course, if you only display on the monitor, then you can take the monitor resolution as the criterion, but that's not future proof, as 300 dpi monitors are already available in mobile devices.

Vlad

 

[Cy Cheze]

The human eye has an area of focal vision, a whole lot of peripheral vision, and a blind spot were the optic nerve connects. You don't notice that blind spot because it is blind and because the second eye can usually pick up the gap.

Anyway, a camera sensor tends to capture equal resolution across an entire field of view. The camera depth of focus may be limited, in a way that mimicks the eye, but a narrow aperture can also allow a camera to have very deep focus.

An eye is an organ of a creature geared to perceive what it is conditioned to consider important, or what the viewer prefers to select from objects in a field of view. The rest is forgotten or blurred. The camera, meanwhile, has no such preference and will record lots of stuff that challenges human memory or which cannot be transmitted from one human to another by mere description.

All visual images are, arguably, abstractions or analogies of their subject. Blind people get by with no eyesight or photographs at all. In a world of darkness, all organisms would presumably confront the same. Some creatures see or hear things we cannnot. There are radio or supra-visible radiation frequencies we cannot perceive at all, except with instruments. Photographs capture light reflected from subjects that are no longer where they were when the picture was taken, or which may no longer exist at all. On the other hand, radio telescopes capture radiation now from events that occurred billions of years ago, though what we "see" is rather vague.

So which images are real and which are not?

Anyway, bottom line, most people see very little or only what they want to see. Some folks can manage with their eyes shut most of the time.

 

[becasabot]

I think not only human vision but al living vision is quite different than cameras. Eyes may see all in its field of view however brain "sees" selectively, it does not use entire range of vision simultaneously. It does not see uniformly, detect mostly motion at the edges (where only B&W vision helps with contrast) and detail in the center. That is due to aversion to danger, and focusing on prey/enemy There are saccades, which is the fast movement of both eyes, when the vision interpretation at the brain is inhibited (this is why you can never see your eyes' movement in the mirror say when focusing in the mirror image of your left eye and then shifting your focus to the right eye, even though they both are always nearly in the center of your field of view). This also helps for sensing a scene with greater resolution, and due to it there is no motion blur in human vision due to eyes' movement (motion blur in the opposite case, when the object moves and the eye is fixed)

There are many interesting phenomena about vision of living beings. Also related to consciousness, memory, psychology, NLP. It's not just sensing the light intensity and color similar to a camera but interpretation of environment using info from optic nerve, memory, preconceptions, emotions, priorities, mood etc. Shortly, brain greatly alters data from our pixel, the cones so our sight is only an interpretation of what cones record. Interesting topic, eyesight.

 

[BobYIL]

" The approximate field of view of a human eye is 95° out, 75° down, 60° in, 60° up. About 12–15° temporal and 1.5° below the horizontal is the optic nerve or blind spot which is roughly 7.5° high and 5.5° wide." meaning if we take out the blind spot out of the temporal portion then we get something like an 7-8° angle of "attention with high resolution".

Clarkvision estimates 324MP for an assumed total angle of view of 90°x90°. Calculating for the 8° angle for high resolution, the real resolving power of the human eye should be around 2.5MP within 8° angle of attention; i.e. a 2.5MP picture seen from a distance to consitute 8° diagonally in our sight should have the same resolution as our eye.

(Worth to try.. )

--
BobYIL

 

[Victor Engel]

The optical resolution of the eye is, at best, about 15 LP/mm.

How do you figure? Where I'm sitting right now, about 60cm from my screen, I can clearly see the lines making up the screen. The monitor displays 1280 lines over 375mm, so the lines have a density of about 3.4 line sets per millimeter (three colors make up each pixel). I don't perceive the distinct colors at this distance, though.

The focal length of the human eye is about 22mm. So the spacing on my retina must be denser than 600/44*3.4 = 46 LP/mm. And my vision is not what it used to be.

Where did the 44 come from? Well, I was lazy. I know the screen is somewhere between 1:1 (too close for my eyes to focus) and infinity. At 1:1, the screen would be 44mm from the lens. At infinity, it would be 22mm from the lens. So I was conservative and used the 44mm figure. It really is somewhere between 22 and 44, but like I said, I'm too lazy to figure it out. It's sufficient, though, to show that the resolution is at LEAST 46 LP/mm.
--

 

[Joofa]

What is the resolution of the human eye?

in the Fovea.

--
Dj Joofa

 

[Mentor_1]

Look here

http://clarkvision.com/imagedetail/eye-resolution.html

Gr.

I think if I can ever figure out just what the heck it all means it has the possibility of being very interesting!
--
May the light be with you
Stop global whining
Stupid should hurt

Gear: My Lab Lara, cameras, lens, tripod, monopod, cars, MH, motorcycles (yah bought another), helmet, yacht, etc. too much stuff.

 

[gloaming]

I can't answer your question directly, but I can tell you that the human eye's ability to split two tiny items very close together, which is an effective test of resolving power of any optical system, of which the eye is an example, is determined chiefly by the main aperture allowing light to fall incident on the detector. Eyes that are fully dark adapted, and young, and female, will have resolution better than those of their male peers, on average, and certainly of those of their elders much more than 20 years advanced in age. The reason is that youngsters, say under 10, and especially females, are likely to be able to relax their pupils sufficiently when dark-adapted that they provide an effective aperture of greater diameter than those in the eyes of their parents. There is plenty of anecdotal evidence from astronomers well into their fifties being assured by their six year old grand daughters that they can indeed detect the Galilean moons around Jupiter on clear and still nights. Sceptical grandpa asks her how many she sees, and if they are spread around the planet. Youngster dutifully tells grandpa there are two on the left and only one on the right. He verifiies using binoculars and is amazed to see that she is correct. Grandpa's aged pupils may dilate near 5 mm, whereas hers will dilate well over 6 mm.

After the aperture, resolution is then determined by the number of photocrecptors, their sensitivity, their abilty to work in low signal to noise ratios, and their proximity. How a typical eye compares to a CCD or CMOs I have no idea.

 

[Victor Engel]

Also, the rings of Saturn. Alas, my vision isn't there anymore. It also helps to have good skies. When I was a teenager in the highlands of Guatemala, I saw my shadow by the light of Venus. That's not an acuity thing, of course. It's just a memory that was triggered by your Jupiter comment.
--

 

[Darrell Spreen]

The optical resolution of the eye is, at best, about 15 LP/mm.

How do you figure? Where I'm sitting right now, about 60cm from my screen, I can clearly see the lines making up the screen.

......It really is somewhere between 22 and 44, but like I said, I'm too lazy to figure it out. It's sufficient, though, to show that the resolution is at LEAST 46 LP/mm.

I tried to make the distinction between the optical resolution of the eye (as an optical element) vs. the visual acuity of the eye which results from the microscopic eye motion and the integration of the information by the brain. I would say that is what your experiment is addressing and possibly what the OP was most interested in.

In fact, I'm probably too optimistic saying 15 LP/mm -- more like 5 according to the following reference (page 94).

John C. Russ (2006). The Image Processing Handbook. CRC Press

--
Darrell

 

[Victor Engel]

I tried to make the distinction between the optical resolution of the eye (as an optical element) vs. the visual acuity of the eye which results from the microscopic eye motion and the integration of the information by the brain.

I realize the brain is a major, perhaps the major component of human vision, but I'm not ready to agree that such integration can make that big a leap in realized resolution.

I would say that is what your experiment is addressing and possibly what the OP was most interested in.

I still see the lines distinctly if I don't move my eyes. Of course, there can still be integration going on. What's needed is for the image to be flashed, but then focus would be an issue.

In fact, I'm probably too optimistic saying 15 LP/mm -- more like 5 according to the following reference (page 94).

John C. Russ (2006). The Image Processing Handbook. CRC Press

Well, since I don't have that book, that reference is not particularly useful. I see it's available for purchase, but it has a pretty hefty price. Perhaps you could post a relevant exerpt?
--

 

[Darrell Spreen]

I tried to make the distinction between the optical resolution of the eye (as an optical element) vs. the visual acuity of the eye which results from the microscopic eye motion and the integration of the information by the brain.

I realize the brain is a major, perhaps the major component of human vision, but I'm not ready to agree that such integration can make that big a leap in realized resolution.

I still see the lines distinctly if I don't move my eyes.

You have no control over the micro scanning that the eye is doing on a continuous basis. It's not about the eye motion that you are aware of.

Well, since I don't have that book, that reference is not particularly useful. I see it's available for purchase, but it has a pretty hefty price. Perhaps you could post a relevant exerpt?

It's available online to read -- the discussion actually runs for about 6 pages -- it translates the angular resolution of they eye (with 20/20 vision) to the equivalent linear grating variation (in lines/mm) for a 100% contrast test object at a viewing distance of 50cm.

--
Darrell

 

[Darrell Spreen]

Rudolph Kingslake, author of a number of texts on optics and Emeritus Professor at the University of Rochester, goes through the same type of analysis to evaluate the eye's ability to distinguish between a true point source and a blur circle and reports it to be about 1.5 minutes of arc which translates for black lines on a clear background to be 4.6 LP/mm at 50cm distance (under conditions he qualifies as ideal illumination that is "rarely encountered").

--
Darrell

 

[Victor Engel]

It's available online to read

Can you post a link?
--