New Binocular Resolution Test Results (1 Viewer)

solentbirder said:

I think I might throw some light on this (if you excuse the pun). I've been playing around with optics for as long as I can remember. This experience leads me to the following conclusion (which I expect will be strongly refuted !).

A larger objective lens has higher resolving power, regardless of the size of the your eye pupil. The 'field stop' argument only applies to the brightness of the observed image. Why is this so ? When you look through a pair of binoculars or a telescope you are viewing the image formed at the focal plane of the objective. The job of the eyepiece is simply to magnify this image.
There is nothing else going on. Imagine you inspect a very finely printed postage stamp with a magnifying glass. Imagine the stamp is the image at the focal plane of your binocular/telescope (in this analogy a larger objective 'prints' a more finely detailed stamp). Does the detail of the printing change under different light conditions because your exit pupil alters ? No.
Images formed at the focal plane of larger objectives always contain more detail than those from smaller objectives. The 'field stop' argument mistakenly treats the binocular as a single compound lens and misses the fact that you are actually viewing the image at the focal plane of the objective. At any magnification, a larger instrument will show more resolution than a smaller one (Assumptions: 1. The optical quality is equivalent. 2. The the lenses in your eyes are of good quality out to the edges. For a lot of people this is not true which is why some report a sharper image with binoculars having small exit pupils. In this case the beam of light is passing through the best central area of their eye lenses).

Let the slings and arrows begin !

Click to expand...

You are forgetting what the exit pupil of an optical system is. It is an image of the light coming through the objective, and there is an exact correspondance between the aperture and the exit pupil. The center of the exit pupil contains the light coming through the center of the objective. The outer part of the exit pupil contains light coming from the outer part of the objective (and light coming through each area contains light from the entire field of view).

You can easily prove this to yourself. Hold binoculars so you can see the exit pupil. Now place a something - perhaps the tip of a pencil - just in front of the objective. You can see that the placement of the pencil tip in front of the objective exactly matches the placement in the exit pupil.

The exit pupil is aperture, and stopping down the exit pupil is exactly the same as stopping down the aperture, and has exactly the same effect on light grasp and resolution.

I have been through this discussion several times in the past, and have gotten quite a bit of disagreement and even abuse in these discussions. During these heated discussions I talked with several folks who make a living designing and making optical systems, and their response was uniformly in agreement - a stop at the exit pupil has the same effect as a stop at the aperture.

Clear skies, Alan

AlanFrench said:

Jean-Charles,

I am a bit puzzled by these numbers.

The resolution, in green light, according to Dawe's criterion, of a 25mm aperture is 5.6 arc seconds. For a 50mm aperture 2.8 arc seconds.

Whose criterion are you using?

Thanks.

Alan

Click to expand...

I think you mean Dawes criterion, rather than Dawe's criterion. According to Dawes criterion the resolution in arc seconds is 4.56/D inches or 116/D mm Thus for 25mm and 50mm objectives we get 4.6" and 2.3". It was derived empirically and I don't think he mentioned the colour of the light.

Leif

Alan,

Leif is right.
I use a common approximation of the Dawes’ criterion : 120/D
The exact numbers are unimportant, only relative values matter.

Jean-Charles

Leif said:

I think you mean Dawes criterion, rather than Dawe's criterion. According to Dawes criterion the resolution in arc seconds is 4.56/D inches or 116/D mm Thus for 25mm and 50mm objectives we get 4.6" and 2.3". It was derived empirically and I don't think he mentioned the colour of the light.

Leif

Click to expand...

Leif,

Yes, indeed "Dawes." No, I don't think he did mention the wavelength, and I made things confusing by using the formula for the Rayleigh criterion and writing "Dawes" instead.

Just curious where the numbers came from, and I guess I'd better be sure I have the criterion and math in sync.

Thanks.

Alan

jcbouget said:

Alan,

Leif is right.
I use a common approximation of the Dawes’ criterion : 120/D
The exact numbers are unimportant, only relative values matter.

Jean-Charles

Click to expand...

Yes, thanks. I agree that the exact numbers matter little, and was mostly curious about where the numbers came from. Alas, I had calculated the Rayleigh criterion and wrote "Dawes" (as corrected).

Clear skies, Alan

rka said:

Wow ... a lot of reading in ths thread. I have one simple question though:

How do these resolution tests cater for the actual resolution seen when the binoculars are handheld and subject to minor shake. This scenario probably caters for 99.9% of binoculars usage. In this case, an IS unit with inferior glass may provide improved resolving to the end user.

Looking forward to feedback on this one.

Click to expand...

Personally, I think that the inability to hold the binocular steady is the most serious limitation on what we can see. There is certainly a difference in my view hand-held and my view with a tripod mounted binocular, and the difference is significant at higher powers.

Clear skies, Alan

solentbirder said:

I think I might throw some light on this (if you excuse the pun). I've been playing around with optics for as long as I can remember. This experience leads me to the following conclusion (which I expect will be strongly refuted !).

A larger objective lens has higher resolving power, regardless of the size of the your eye pupil. The 'field stop' argument only applies to the brightness of the observed image. Why is this so ? When you look through a pair of binoculars or a telescope you are viewing the image formed at the focal plane of the objective. The job of the eyepiece is simply to magnify this image.
There is nothing else going on. Imagine you inspect a very finely printed postage stamp with a magnifying glass. Imagine the stamp is the image at the focal plane of your binocular/telescope (in this analogy a larger objective 'prints' a more finely detailed stamp). Does the detail of the printing change under different light conditions because your exit pupil alters ? No.
Images formed at the focal plane of larger objectives always contain more detail than those from smaller objectives. The 'field stop' argument mistakenly treats the binocular as a single compound lens and misses the fact that you are actually viewing the image at the focal plane of the objective. At any magnification, a larger instrument will show more resolution than a smaller one (Assumptions: 1. The optical quality is equivalent. 2. The the lenses in your eyes are of good quality out to the edges. For a lot of people this is not true which is why some report a sharper image with binoculars having small exit pupils. In this case the beam of light is passing through the best central area of their eye lenses).

Let the slings and arrows begin !

Click to expand...

The image formed at the focal plane of the objective contains all the rays going through it. So the resolution of the objective is given by the radius of the Airy disk, which depends on the aperture of the objective.
But what matters is the resolution of your eye behind the binoculars, i.e. the resolution at the focal plane of your eye : the retina. Here the resolution is given by the size of the diffraction pattern on the retina. This diffraction pattern depends on the eye's aberrations and the size of the eye pupil. It is not directly related to what happens at an intermediate focal plane situated inside an auxiliary optical device in front of the eye.
You must consider that the diffraction pattern at the focal plane of the objective is virtual : if your eyes stop down the exit pupil, then some rays coming from the objective don’t enter the eye, and considering the diffraction pattern at the focal plane given by the whole objective is misleading.

I have to confess that the first time I read an explanation similar to yours, I found it excellent. But now I can tell you that it is wrong. I confirm that Alan French has spent a lot of energy to explain such things in other forums, and after thinking a bit about this question, I have changed my mind.

As a consequence, the fact that some people, including Kimmo Absetz, find that increasing aperture in low power instruments improves resolution, remains a mystery for me.

Jean-Charles

jcbouget said:

[SNIP]
As a consequence, the fact that some people, including Kimmo Absetz, find that increasing aperture in low power instruments improves resolution, remains a mystery for me.
Jean-Charles

Click to expand...

Jean-Charles,

This puzzles me too. There are a couple of possibilities. First, perhaps the eye is not working at the same pupil opening with both apertures. A smaller pupil would result in better acuity because they eye's aberrations are reduced. I have been meaning to look into what causes the eye to work at a certain pupil size and see if that might or might not provide a mechanism for differing pupil responses to different aperture and brightnesses.

Also, a larger objective tends to be a longer focal length, so stopping it down produces a slower system, reducing the aberrations of the objective and also tending to work better with the eyepiece,

Clear skies, Alan

AlanFrench said:

Personally, I think that the inability to hold the binocular steady is the most serious limitation on what we can see. There is certainly a difference in my view hand-held and my view with a tripod mounted binocular, and the difference is significant at higher powers.

Clear skies, Alan

Click to expand...

Agreed.

jcbouget said:

The image formed at the focal plane of the objective contains all the rays going through it. So the resolution of the objective is given by the radius of the Airy disk, which depends on the aperture of the objective.
But what matters is the resolution of your eye behind the binoculars, i.e. the resolution at the focal plane of your eye : the retina. Here the resolution is given by the size of the diffraction pattern on the retina. This diffraction pattern depends on the eye's aberrations and the size of the eye pupil. It is not directly related to what happens at an intermediate focal plane situated inside an auxiliary optical device in front of the eye.
You must consider that the diffraction pattern at the focal plane of the objective is virtual : if your eyes stop down the exit pupil, then some rays coming from the objective don’t enter the eye, and considering the diffraction pattern at the focal plane given by the whole objective is misleading.

I have to confess that the first time I read an explanation similar to yours, I found it excellent. But now I can tell you that it is wrong. I confirm that Alan French has spent a lot of energy to explain such things in other forums, and after thinking a bit about this question, I have changed my mind.

As a consequence, the fact that some people, including Kimmo Absetz, find that increasing aperture in low power instruments improves resolution, remains a mystery for me.

Jean-Charles

Click to expand...

It works very likely similar to a fotografic lens. Use of full aperture (as with smaller bins) leads to a good but not optimal picture, which is caused by a decrease of contrast, sharpness and resolution (blooming effects, softening effects a.s.o.). Stopping down aperture by one or two steps makes the optical quality optimal, because the edge rays are subdued and/or cut off. The stop down effect is done by the eye's pupil (round about 3mm) at bright light conditions. This increases everything: contrast, sharpness, perceived resolution. With smaller bins (EP of 3 or 4mm) there is not much to stop down.

Walter

jcbouget said:

As a consequence, the fact that some people, including Kimmo Absetz, find that increasing aperture in low power instruments improves resolution, remains a mystery for me.

Click to expand...

I have come up with a couple of ideas, which could at least partially explain this phenomenon. One (A) concentrates on the eye and the other (B) on the binoculars. I have not seen them mentioned before, but I wouldn't be surprised if they were.

(A) If the binocular has 2.5 mm exit pupil and the eye pupil is 2.5 mm, the eye can not see all the incoming light rays because of continuous rapid eye movements (microsaccades). I don't know how much larger the radius of this "kinetic iris" is than the physical aperture of the iris. When the exit pupil is large enough, the iris can move around and yet the retina is continuously exposed with "details".

(B) It has been discussed here and on Cloudy Nights forums that binoculars have varying amounts of vignetting (due to internal field stops, too small prisms or whatever). If a binocular transmits about 90% of the light, this is only at the central 50% (or something) of the field-of-view. Now a stopped down (by the 2.5 mm eye) 10x50 produces a completely unvignetted 10x25, whereas a compact 10x25 probably has significant vignetting.

These examples were just to show that there may be some areas (in addition to above mentioned) where a 10x25 "made from" a 10x50 may deliver more information to the eye than a real 10x25.

Ilkka

iporali said:

[SNIP]
These examples were just to show that there may be some areas (in addition to above mentioned) where a 10x25 "made from" a 10x50 may deliver more information to the eye than a real 10x25.
Ilkka

Click to expand...

Ilkka,

Excellent points. With 10x50 binoculars and a 2.5mm eye pupil it is almost certain that the binocular's exit pupil fills the eye's pupil. With 10x25 binoculars positioning is much more critical and it is likely that the binocular's exit pupil does not quite correspond to the eye's pupil.

Somewhere I recently read something about the minute changes of angle made by the "workng" eye. Now, if I can just remember where....

Clear skies, Alan

Thanks Alan - I have to admit that it is easier to sound convincing when the objectives are 25 and 50 mm than 32 and 42 mm.

Ilkka

AlanFrench said:

Ilkka,

Excellent points. With 10x50 binoculars and a 2.5mm eye pupil it is almost certain that the binocular's exit pupil fills the eye's pupil. With 10x25 binoculars positioning is much more critical and it is likely that the binocular's exit pupil does not quite correspond to the eye's pupil.

Somewhere I recently read something about the minute changes of angle made by the "workng" eye. Now, if I can just remember where....

Clear skies, Alan

Click to expand...

There is another reason why larger instruments might give a better view. I have noticed that small binoculars such as the Swarovski 8x20 seem to give exceptionally pleasing views, at least to my eyes. I think the reason is the narrow field of view: it is easier to design a high quality eyepiece with a narrow, rather than a wide, FOV. Larger binoculars also tend to have narrower fields of view than their ~40mm cousins, and hence the same argument should apply. (I suspect that the reason for the narrow FOV is that for a 50mm+ objective to provide a large FOV - e.g. 135m at 1Km - the prisms would be unwieldly, and hence the instrument would be too big and too heavy.)

Leif

kabsetz said:

Alan,

I think your puzzlement arises from not differentiating between systems that have relatively low magnifications and those that have relatively high magnifications.

With telescopes, you can obviously increase magnification until you get to the point where the full resolution of the objective is realized. With binoculars, the magnification is (unless the binocular really sucks) so low that human eyes won't be able to resolve the full resolution of the objective and will set the limit, but not the whole limit. What I'm trying to say is that in the latter case, that is, with optics that have a resolution limit beyond the smallest detail the eye can detect through them, the eye will nevertheless reach its full resolving capacity only if the resolving power of the optics has enough of a margin over what the simple (and simplistic) formula of: r(eye) = r(optics) times m(agnification) would propose.
[SNIP]
Kimmo

Click to expand...

Kimmo,

Yes, there is a difference between viewing at low powers, and high power views, but I do not buy the idea that the optics need a resolution limit in excess of the human eye at low powers. The optics can not deliver that resolution to the retina because of the additional diffraction effects caused by the eye's pupil. The full resolution of the 50mm objective is lost.

Clear skies, Alan

AlanFrench said:

Kimmo,

Yes, there is a difference between viewing at low powers, and high power views, but I do not buy the idea that the optics need a resolution limit in excess of the human eye at low powers. The optics can not deliver that resolution to the retina because of the additional diffraction effects caused by the eye's pupil. The full resolution of the 50mm objective is lost.

Clear skies, Alan

Click to expand...

Don't buy this either. With low power instruments for terrestrial observation contrast and a wide true field of view with limited optical aberrations are crucial. Resolving power of the instruments remain unused. Demanding an increased resolving power may be sort of senseless. Otherwise definition of resolving power is subject to change since the days of Mr. Abbe due to new techniques and due to observed objects. So I would like to stay cautious to agree or disagree until I get a full explanation.

Walter

I returned from my vacation to find many excellent posts on this thread, and in the end no consensus as to how much resolution can be seen or how much is really needed in low power binoculars. For anyone who can stand more on the subject I did an experiment to try to determine the relationship between the measured resolution of one barrel of an 8x binocular and the resolution I can actually see when looking directly through it. I went into this with the view that a binocular’s measured resolution only needs to match the observer’s eyesight acuity for no compromise in resolution to occur, so the result came as something of a surprise.

I fitted one barrel of a Zeiss 8x42 FL with a series of aperture stopdowns of decreasing size. Each reduction in aperture resulted in a measurable loss of resolution when viewing a 1951 USAF resolution test pattern at 10m through a 5x magnification booster positioned behind the binocular eyepiece. I also viewed the test pattern directly through the binocular with each aperture stopdown. The results are shown below. The first number is the measured resolution with the magnification booster and the second number is the resolution I could actually see on the chart looking directly through the binocular. BTW these numbers will look too good to those who are accustomed to Dawes and Rayleigh criteria. I am using what I understand to be the system used in ophthalmology since I want to relate eyesight acuity and binocular resolution using the same system. The resulting figures are about twice as good as Dawes/Rayleigh figures would be, so for an approximate Dawes figure just multiply by 2. This is also the system used by Kimmo Absetz in his Alula reviews, and appears to conform to Bawko’s figures at the beginning of the thread.

aperture- measured/ seen
42mm- 1.64”/ 5.8”
30mm- 2.06”/ 5.8”
21mm- 3.28”/ 5.8”
18mm- 3.67”/ 6”
16mm- 4.13”/ 6.3”
14mm- 4.64”/ 6.56”
11mm- 5.8” / 7.34”
8mm- 8.26”/ 9”

If my preconception had been correct I would have continued to see 5.8” until I reached the 11mm stopdown where measured resolution is 5.8”. Instead these results tend to confirm Kimmo’s view that a binocular needs to have resolution somewhat better than eyesight acuity for the combination of resolution and acuity not to be compromised. The very first tiny loss of visible resolution occured in this test when the binocular’s aperture had been reduced to 18mm and measured resolution had fallen to 3.67” . Multiply 3.67” by 8x and the result is about 29”. My best eyesight acuity was about 46” in this test (5.8” X 8x), so the result is not too far from Kimmo’s opinion that the resolution of a binocular needs to be about twice as good as eyesight. For my eyes Kimmo’s standard is a little better than needed but correct in principal. It’s interesting to see that when the binocular’s measured resolution had fallen to 5.8”, equaling my best eyesight acuity, the combination of binocular resolution and eyesight acuity was reduced to 7.34”. The combination of resolution and acuity together is worse than either taken by itself. I should mention that the image quality with the stopdowns was actually very high in the sense that it was quite clean, high contrast and aberration free. The 21mm stopdown actually produced a superior image compared to full aperture in every way except brightness. I used a very bright halogen task lamp about 6” from the chart to maintain as much brightness as possible at the small exit pupils. If the resolution had been limited by aberrations or defects rather than aperture the image would have looked brighter, but the image quality would almost certainly have been worse.

One thing that should be understood for placing these measurement in context is that 3.67” (the point at which I began to be able to see a tiny loss of resolution) is in fact a very poor measured resolution figure for any binocular with an aperture above 20-24mm. I’ve never measured resolution this poor for any binocular that wasn’t defective and even defective ones are usually better. I measured the two worst 8x binocular barrels I have on hand, one from a B&L 8x24 with severe pinching and the other from a Nikon 8x30 E with astigmatism. The B&L measured about 3.4” and the Nikon about 3.0”. Through both I could see the same 5.8” resolution I saw through the Zeiss FL. You can reasonably expect good non-defective 40-42mm binoculars to have have actual resolution between 1.9” and 1.6”. Good 30-32mm’s are usually between 2.3” and 2”. Any binocular 10x or less with these kinds of measurements would be good enough not to cause a visible problem with resolution, even tripod mounted. BTW it is quite impossible to see any of this without tripod mounting. Hand holding the Zeiss FL I found i could only see 8.26” continuously, with glimpses of 7.34”.

To the extent that this experiment can stand scrutiny I find myself in the happy position of being able to both agree with Kimmo that binocular resolution does need to be better than eyesight acuity, and still maintain my original opinion that the image through a good binocular contains details too small for the eye to discern when looking directly through it, at least as long as the measured resolution multiplied by the magnification produces a result about twice as good as an observer’s eyesight acuity.

Thanks, Henry. One good test is worth a thousand speculations...

I'll go on a limb now and try a quick explanation of why it is that when Henry has adjusted his 8x42 with an aperture stop to give a 5.8" resolution, his eye only sees 7.34 (or something thereabouts). We start with 5.8" (times 8, the magnification of the speculative binocular) being the limit of the eye in question. This limit is defined as the smallest bar target the eye can still see as resolved, i.e. the viewer can tell the orientation of the bars rather than seeing the target as a grey square. Obviously, for the eye to see this, the target must be well-defined, clear and have the bars clearly separated by equally wide spaces. If the resolution of the binocular would be 5.8," this means that the image provided by the binocular barely resolves this target, i.e. the bars are almost merging together, the spaces between the bars are effectively narrowed. For the eye, this makes the target much more demanding than a perfect 5.8" target, and consequently the eye can no longer resolve it.

Now, as to the difference between my guestimate that the resolution of a binocular needs to be about twice as good as eyesight for the eye to see an optimally sharp image and Henry's somewhat more conservative finding, I would like to offer two explanations. Firstly, Henry has done a proper, meticulous test (albeit with only one binocular) and consequently has a methodologically stronger case. I have based my guestimate on a more haphazard gleaning from various booster-measuerd and eye-measured resolution tests, NEED-tests and star-tests to assess aberrations.

Secondly, and perhaps now more in support of my more rigorous standard, Henry's test contains the systematic 'error' if I may so say, the one he himself points out: since he uses aperture stops to limit binocular resolution, lower resolution in his test also implies less aberrations. In the 'real world' of real binoculars, lower resolution tends to be the result of relatively more aberrations.

And finally a brief answer to Alan

[Originally Posted by AlanFrench
Kimmo,

Yes, there is a difference between viewing at low powers, and high power views, but I do not buy the idea that the optics need a resolution limit in excess of the human eye at low powers. The optics can not deliver that resolution to the retina because of the additional diffraction effects caused by the eye's pupil. The full resolution of the 50mm objective is lost.

Clear skies, Alan]


I actually agree with Alan and Walter in the sense that if diffraction were indeed the limiting factor, all the normal 7-10x30-50mm binoculars would be good enough for all of us.

I have never said that with a 7-10x binocular the full resolution of a 50mm (or even a 32mm) objective could be delivered to the retina, or to put it another way, that some of it would not be lost. What I have been trying to say all along is that for the eye to see all it can see, the binocular must be able to show significantly more.

Kimmo

Heny,

Very interesting experiment!

However, I don't quite understand this part and would appreciate clarification:

The 21mm stopdown actually produced a superior image compared to full aperture in every way except brightness. I used a very bright halogen task lamp about 6” from the chart to maintain as much brightness as possible at the small exit pupils. If the resolution had been limited by aberrations or defects rather than aperture the image would have looked brighter, but the image quality would almost certainly have been worse.

Click to expand...

My primary concern with the experiment is that brightness (and contrast) should have decreased in proportion to the square of aperture, if I follow this correctly. That would be a range of 15:1 from brightest to darkest. Hence, it seems to me that this would naturally result in lower visual acuity, independent of the eye's need for resolution within the optics.

Could illumination be controlled with a rheostat, thereby holding target brightness/contrast constant?

-elk

Atomic Chicken said:

Greetings...

Bawko

Click to expand...

After much deliberation, Bawko... I followed much of your advice. It was, in the end, between the Zeiss Conquest, Swaro and Leica. But the Ultravid 10x25 won in the end. What a stunner of a binocular it is. Right down to the dim light of early evening yesterday, they showed so very little difference from my Swaro 8.5x42 ELs, I was amzed, nay... I was utterly mystified by how bright and sharp those light little things are. It seems as if the physics must be wrong - the Swaros are vast and weighty things in comparison.

Thinking it was my older eyes fooling me, I asked my son to have a good look through the two, and by then it was getting yet darker. But he agreed with me. One third the price, too! I wish I had found them earlier. What a relief it will be to be able to carry these featherweights on my next long walk.