Aperture of the human eye?

It's usually said it's around f/16 in 35mm format, so since Matt Crumin is using a 4/3 sensor size camera, so f/8.3 is about right, in term of DoF.
Where do you find "It's usually said it's aroundf/16 in 35mm format"?
As mentioned, it's simply said to me. I also mentioned, I'd like to know its empirical study as well, so I can't answer your question. I also would like to know the source of the study myself.
See the two last links here: http://www.dpreview.com/forums/post/53490894

The f/ range of the human eye does not change just because its body is carrying a different camera! :-D

DOF has little consequence since the rate of accommodation of the human lens is as fast as can be perceived. (roughly 1/3 sec or about 30 frames/second)
I don't know how fast exactly is our "human lens", but I do know it reacts faster than any camera we have at the moment (it's also something people simply said, I've yet to read any study on it). I think you misread my statement and perhaps due to the fact you don't know what I do know.
What is important is exposure which is varied by the f/ as the iris adapts its diameter to keep exposure levels comfortable for the retina. [Brightness is also affected by changing the physiology of the rods and cones of the retina especially when at low light levels.]

Focusing is determined by the lens of the eye as it accommodates and perceived brightness is controlled by the pupil (iris) size as it adapts to scene luminance. Important distinctions. Analogous to adjusting the focusing elements in a lens (accommodation) or the f/ in a lens (adaptation). Important not to confuse them.
None of these are what I was discussing in the original post. Focus is focus, DoF of a given aperture is another issue. You can get very shallow DoF even with very small aperture, due to subject-to-focal plane distance, and you can get out of focus area even with small aperture. They're all separate issues, so don't mix them in this discussion.

I have some limited medical training, though not a neurologist nor an optometrist. But I am a photographer and I do know my focus, aperture, DoF, perspectives, exposure and stuff. Please focus on one issue at a time, though they can be somewhat interrelated, but it's not what I was talking about. We are purely talking about a very specific subject, which is aperture of human eye, in term of DoF (not exposure), when eyes are focused toward the infinite end, during a typical daylight condition (assuming the eyes are healthy). If you dragged more variables into this discussion (i.e., not ceteris paribus) then the condition would be very different indeed, and probably too complex for the sake of this discussion.
I believe there is a reason why Sunny 16 rule uses f/16 as the example. It's the most natural aperture for the 35mm format.
Sunny f/16 rule comes from the Exposure Metering Equation which is quite independent of format or human eyeballs.

35 mm focal length, though is related to the field of view of humans, but that is a different discussion.
I am pretty sure I said 35mm format (as oppose to 4/3 format), not 35mm Focal Length. Please do not change the subject. And that's the reason I also discussed 43mm "focal length" in the following paragraph, hoping some readers would at least catch the mistake when they read on. And I do mean DoF, not field of view. Some people's field of view are closer to 35mm (in 35mm format), and some are closer to 50mm (in 35mm format).
That's why 50mm focal length is called the standard focal length, because it's very close to 43mm focal length, which has the most natural, naked eye, perspective (not necessarily your field of view, but in term of perspective, I find it to be true).

Yes, our eyes' "aperture" (called pupil, by the way) will vary in size due to light condition. But under sunlight, it should be f/16 or as Matt Crumin mentioned, about f/8.3 for 4/3 sensor size.
No. See above.
Do read the condition in which the argument is set. Any logical argument must be set in a prescribed premise before any conclusion could be reached, and I did set my premises very specifically (and some are implied, which I thought there was no need to elaborate). How can we come up with a logical conclusion if you keep change the rules of the game?

I also avoided talking too technical about 43mm focal length issue, precisely because I don't want to drag the perspective vs field of view argument into the equation.... Most people on DPR already have trouble understanding a given aperture's equivalent DoF and equivalent exposure between 35mm format and 4/3 format.... Perspective vs field of view can be even more confusing. Let's keep the argument simple: just human eyes' theoretical aperture. The Wiki article you mentioned also said something, which I also questioned in my original post, there can be too many variables, so a precise focal length and therefore aperture value (since it's calculation is depended on focal length) is very difficult to pinpoint. Wiki article mentioned the eye is liquid filled, and I did mention retina is curved and optical nerve (which translate image to our brain to process) is not in the exact center of the eye. With that said, you should already know I've been trying to say they're all obstacles to figure out the actual aperture of human eyes (and our retina isn't flat like a sensor, and our eyes aren't glass lenses, and our eyes' dynamic range is vastly superior than current sensor technology). Not to mention people's rods and cones arrangements are different, therefore, vision is an interpreted reality. You don't have to be a doctor to know this. As most modern artists know, reality is a subjective interpretation made by our brains.

The entire argument about what's human's natural aperture only needs an approximation, for our own reference when creating art. They don't need to be scientifically precise (though it's pretty mind-intriguing I must say, if we could find it out). And who said you must shoot at the same aperture and focal length that represent your eyes see? If that's so, why create wide angle lenses which give wide angle perspective or telephoto lenses that create compression perspective? And why all these DoF control tools with our cameras? If that's the case, shouldn't you be shooting with a Holga? (Holga is fixed to 47mm focal length with f/8)
 
I was looking at Matt Crumin's site which is relay excellent. (Using E-M5). I found the link through 4/3rumors. Matt has a little pull down in the corner that gives an interesting factoid:

"The f-number of the human eye varies from about f/8.3 in brightly lit conditions, to about f/2.1 in dark conditions."

"To work out the focal length of the human eye, you would need to take into account the light-reflecting properties on liquids in the eye."

Really interesting. What do you think? Is this for all human eyes? It would be very interesting to know what the aperture for other species eyes would be.

Cheers,

Seth
Reportedly 22/3.5 or so. Pretty slow MFT normal lens.

http://clarkvision.com/articles/eye-resolution.html
Good link!

Thanks, Anders.

Tom (probably stuck around f/9 due to lack of muscle tone in the iris.)

[Actually, I heard that Impressionist Artists painted their stuff that way as they got older, with cataracts etc. Maybe that is why I am starting to make more abstract stuff. :-) ]
Well, we've got great DoF and my myopia helps with close-ups. What more can we really expect? :-)
The colour sensitivity and DR are very good. One of the best eyes in the animal kingdom.
 
A now-retired colleague used to get out his "finger-glasses" when he needed to get a closer look at something. It's a great example of the sharper focus and greater DOF offered by a tighter f-stop. It works best in good light.

Pinch your thumb and pointer finger together, then bring your middle finger over to form the third wall of a 3-bladed aperture. The hole should be just big enough for the tip of a pencil to poke into. Keep this shape and bring your hand against your face, so you can look through this gap with one eye.

To test it, move your face and 'aperture' as close as you can to the screen… while still maintaining a sharp image. Stay in that position and remove your hand, to see how blurry things are without this device.

I really should patent this. LOL!
 
…if you are thinking of comparing what the eye sees with photography since the eye's exposure is continuous and scanning and post processing is enormous.

Cheers, geoff
 
I was looking at Matt Crumin's site which is relay excellent. (Using E-M5). I found the link through 4/3rumors. Matt has a little pull down in the corner that gives an interesting factoid:

"The f-number of the human eye varies from about f/8.3 in brightly lit conditions, to about f/2.1 in dark conditions."

"To work out the focal length of the human eye, you would need to take into account the light-reflecting properties on liquids in the eye."

Really interesting. What do you think? Is this for all human eyes? It would be very interesting to know what the aperture for other species eyes would be.

Cheers,

Seth
Reportedly 22/3.5 or so. Pretty slow MFT normal lens.

http://clarkvision.com/articles/eye-resolution.html
I prefer to consider a human eye as having 30 mm focal length.

The second Principal Plane of a human eye is located 1.6-2 mm behind the cornea, for an eye diameter of ~24 mm this indeed yields around 22 distance from the second PP to retina which we would call "Focal Length" under usual circumstances (the sensor/image in air).

However, in higher (than air) refractive index medium the optical path is "lengthened" by the factor of index of refraction itself. Relative size of two objects imaged on retina would correspond to such imaged by a thin 30 mm lens in air (though the absolute size would be that of 22 mm lens). The wavelengths of light shrinks in media by the same factor, and the diffraction limit is correspondingly affected. Therefore the amount of detail that could be resolved over the smaller size in medium is the same as over a larger size in air.

For 7 mm open pupil that would correspond to 30 mm f/4.3 lens. Our sharpest vision is achieved at pupil open to 3-4 mm so that's f/7.5-f/10.1

Angular resolution is independent on the way you treat the eye (whether or not lengthening the optical path by a factor of index of refraction) it is only function of the pupil opening. For a 4 mm pupil, the diffraction limit (Rayleigh) is ~1.22 lambda/D, or ~ 0.5 arc min for 500 nm wavelength (yellow). Our eyes can resolve beyond Rayleigh, closer to the Sparrow criterion (the factor would be 0.94 instead of 1.22) so make it 0.4 arc min.
 
LOL how confused can you get?
 
…if you are thinking of comparing what the eye sees with photography since the eye's exposure is continuous and scanning and post processing is enormous.

Cheers, geoff

--
Geoffrey Heard
Down and out in Rabaul in the South Pacific
http://pngtimetraveller.blogspot.com/2011/10/return-to-karai-komana_31.html
Exactly! I've read that the quality of our eye lenses fall off dramatically towards the periphery, so much that 70-90% of colour and contrast and 90% of resolution is lost. By this, our brain doesn't work and process images at a 'pixel' level, but rather takes points of interests and forms a database of sorts, recording contrasts/colours/textures etc. When we look at things, our brain actually uses these 'focus' points to form the super high res images we see based mostly on memory. As sensor-die processing is better and faster, a decent enough resolution can produce a great image and this trend is evident in m4/3 lens design where they trade off optical accuracy for size and weight. Think of how much processing goes into the billion dollar space telescopes!

Comparisons between eyes and lenses are entirely valid where optical quality and characteristics are concerned, but we know that for the most part - they're the same thing after all and the eye makes the crappier image. but we know that images are formed by processing the raw data provided by lenses and in that respect the brain is way out of the scope of comparison. It's not unreasonable that in the future, images could be created via a similar process to our brain - realtime scanning, recording, retrieving, processing - but it certainly won't make it into your m4/3. And then, imaging would no longer be an optical science, but a computer science.
 
Last edited:
I was looking at Matt Crumin's site which is relay excellent. (Using E-M5). I found the link through 4/3rumors. Matt has a little pull down in the corner that gives an interesting factoid:

"The f-number of the human eye varies from about f/8.3 in brightly lit conditions, to about f/2.1 in dark conditions."

"To work out the focal length of the human eye, you would need to take into account the light-reflecting properties on liquids in the eye."

Really interesting. What do you think? Is this for all human eyes? It would be very interesting to know what the aperture for other species eyes would be.

Cheers,

Seth

--

What if the hokey pokey really is what it's all about?
--
wallygoots.smugmug.com
wallygoots.blogspot.com
from the formula linking pupil size to f number and focal length, it is easy to derivate that the f number will just be given by the focal length divided by the diameter.

For the focal length, various sources give it at 17mm, other sources at 22mm. Just like anything on a living organism, it will obviously vary by individual.

For the pupil diameter, we also have a range. For a small pupil in bright light, various sources give a range of 3 to 5mm. For a large pupil in low light, various sources give a range of 4mm to 9mm.

So at one extreme, if you take a large eye with a 22m focal length, and if this eye can contract its pupil to just 3mm, then you get an f-number of 22/3 = 7.3.

And at the other extreme, if you take a small eye with a 17mm focal length, and this eye can open its pupil to 9mm in he dark, then you get an f-number of 17 / 9 = 1.9.

But perhaps the larger eye cannot contract to 3mm and stops at 5mm and thus its maximum f-number is just 22/5 = f4.4. And perhaps the smaller eye cannot open its pupil to 9mm.
 
I was looking at Matt Crumin's site which is relay excellent. (Using E-M5). I found the link through 4/3rumors. Matt has a little pull down in the corner that gives an interesting factoid:

"The f-number of the human eye varies from about f/8.3 in brightly lit conditions, to about f/2.1 in dark conditions."

"To work out the focal length of the human eye, you would need to take into account the light-reflecting properties on liquids in the eye."

Really interesting. What do you think? Is this for all human eyes? It would be very interesting to know what the aperture for other species eyes would be.

Cheers,

Seth
Reportedly 22/3.5 or so. Pretty slow MFT normal lens.

http://clarkvision.com/articles/eye-resolution.html
Good link!

Thanks, Anders.

Tom (probably stuck around f/9 due to lack of muscle tone in the iris.)

[Actually, I heard that Impressionist Artists painted their stuff that way as they got older, with cataracts etc. Maybe that is why I am starting to make more abstract stuff. :-) ]
Well, we've got great DoF and my myopia helps with close-ups. What more can we really expect? :-)
The colour sensitivity and DR are very good. One of the best eyes in the animal kingdom.
Hmm. Mine must be out of spec. Sample variation I guess. ;-)
 
I was looking at Matt Crumin's site which is relay excellent. (Using E-M5). I found the link through 4/3rumors. Matt has a little pull down in the corner that gives an interesting factoid:

"The f-number of the human eye varies from about f/8.3 in brightly lit conditions, to about f/2.1 in dark conditions."

"To work out the focal length of the human eye, you would need to take into account the light-reflecting properties on liquids in the eye."

Really interesting. What do you think? Is this for all human eyes? It would be very interesting to know what the aperture for other species eyes would be.

Cheers,

Seth
Reportedly 22/3.5 or so. Pretty slow MFT normal lens.

http://clarkvision.com/articles/eye-resolution.html
I prefer to consider a human eye as having 30 mm focal length.

The second Principal Plane of a human eye is located 1.6-2 mm behind the cornea, for an eye diameter of ~24 mm this indeed yields around 22 distance from the second PP to retina which we would call "Focal Length" under usual circumstances (the sensor/image in air).

However, in higher (than air) refractive index medium the optical path is "lengthened" by the factor of index of refraction itself. Relative size of two objects imaged on retina would correspond to such imaged by a thin 30 mm lens in air (though the absolute size would be that of 22 mm lens). The wavelengths of light shrinks in media by the same factor, and the diffraction limit is correspondingly affected. Therefore the amount of detail that could be resolved over the smaller size in medium is the same as over a larger size in air.

For 7 mm open pupil that would correspond to 30 mm f/4.3 lens. Our sharpest vision is achieved at pupil open to 3-4 mm so that's f/7.5-f/10.1

Angular resolution is independent on the way you treat the eye (whether or not lengthening the optical path by a factor of index of refraction) it is only function of the pupil opening. For a 4 mm pupil, the diffraction limit (Rayleigh) is ~1.22 lambda/D, or ~ 0.5 arc min for 500 nm wavelength (yellow). Our eyes can resolve beyond Rayleigh, closer to the Sparrow criterion (the factor would be 0.94 instead of 1.22) so make it 0.4 arc min.
Thanks for qualifying this a bit. Interesting.
 
Last edited:

Keyboard shortcuts

Back
Top