A brightness and resolution question. (1 Viewer)

Apologies if this has been covered before.

The view through my various binoculars is significantly sharper in bright conditions. I'm just wondering if this actually makes sense.

In low light the pupil dilates so (age permitting ) you might use the full 5.25mm exit pupil of an 8x42 and get the full resolution benefit of the objective. In a good binocular that might be 3.5 arceconds of resolution. In bright sunlight if the pupil shrinks to 2mm you would only use 16mm of the objective, so the binocular resolution should reduce accordingly; to 9.2 arcseconds if nothing else changes. That might put it close to the close to to the acuity limits for those with the best eyesight at 8x and more of us at higher powers. Have I got that right? (So is resolution with an objective stopped down to 16 or 20mm more meaningful than full aperture? )

So why do my binos seem sharper in bright light if the binocular optical resolution is decreasing? Is it simply that my eyesight is better when it's bright (which it seems to be), or is there some factor involved I've misunderstood?

Thanks

David

Within any given glass, brighter conditions usually means sharper. The cones in the eyes which see detail need light, so this makes sense. A given glass seems to loose that crispness as it gets darker. At least this is my experience, Morning light usually gives the sharpest view for me.

I was thinking about cone (colour) or photopic vision, where the purpose of the pupil is to keep the the amount of light approximately the same over a wide range of light intensity. The rods do have a lower density, so our night vision (scotopic) is lower resolution.

My acuity improves a bit with bright conditions. That again appears to be the wrong way round. If resolution is based on entrance pupil diameter then acuity should decrease under bright condition. Maybe the aberrations in the eye increase with pupil diameter? Maybe I'm being dumb, but doesn't the theory suggest than both binocular and eyesight resolution be at a minimum under bright conditions?

David

I find the sharpest resolution on slightly overcast and breezy days. Good heat dispersion over the land is one reason but things just seem to pop out at me on overcast days as long as not too dark. Especially if rain has just cleared the air.

Is this anything like a camera lens, where the sharpest image is obtained at apertures a couple of stops under the maximum?

Dave

I've tried googling for published info.

As far as I can make out, acuity increases with background light intensity until it plateaus above ten millilamberts. Pupil area decreases with light intensity to a minimum above 1 millilamberts. While it's suggested that optimal lens performance is between 2 and 3mm pupil diameter, it seems that more cones are stimulated at light intensities that cause further pupil contraction increasing acuity further with maximum acuity between at about 2mm.

How bright is a millilambert???

So our eyesight is most demanding of the centre 16mm of any binocular objective?


David

Are you seeing increased resolution or increased contrast? Contrast tends to drop off in flat or dull light and seems at its best in bright, clear conditions.

james holdsworth said:

Are you seeing increased resolution or increased contrast? Contrast tends to drop off in flat or dull light and seems at its best in bright, clear conditions.

Click to expand...

James,

I was referring to acuity and resolution. I see plenty of contrast differences between pairs depending on their stray light control and colour balance etc.. Must admit I was wondering if my eyesight might be a bit weird, but Figure 1 of the following seems to illustrate what I see.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2140971/pdf/255.pdf

Guess I might be normal after all. Just haven't figured out what a millilambert 'brightness' might look like.

If I've understood things right, it now looks like at low to medium light intensities, eyesight might be limiting, but under bright conditions, particularly higher magnifications, the binocular resolution could be limiting and distinguishable between pairs. If Koenig is right there is a lot less variation between individuals than I might have guessed.

David

typo said:

...So why do my binos seem sharper in bright light if the binocular optical resolution is decreasing? Is it simply that my eyesight is better when it's bright (which it seems to be), or is there some factor involved I've misunderstood?

Thanks

David

Click to expand...

You misunderstand. Unlike your eye, the binocular is not an adaptive optic. Its resolution is fixed and determined by the objective lens size AND its optical quality. Its resolving power does not change with brightness.

Your eye is an adaptive optic and acuity (ie. optical quality) is optimized with smaller entrance pupils utilizing the central area of the lens so astigmatism and other vision issues that affect the outer eye lens are mitigated and the image is focused on the denser receptor clusters.

David,
A perfect 42 mm would have 2.4 arcsec resolution, but as you note, 3.5 is "alpha" quality. The aberrations that make the difference, resulting from very fast optics and imperfect construction, become less severe as the effective aperture is decreased, as when the eye closes down in bright light. In the best 8x42s, when the effective aperture is reduced to 22 mm by the pupil which contracts to about 2.5mm in bright light, the optics although small now will behave in a nearly perfect manner, resolving about 4.5 arcsec.

The normal eye resolves best at 2.5mm pupil size, because at this size, it's "camera like" depth of field, which compensates for its imperfections, reaches a balance with its own aperture limitation on resolution. In this bright condition, the naked eye resolves 60 arcsec, so at 8x will resolve 7.5 arcsec.

So, in dim or bright conditions, the resolution of the binocular/eye system is dominated by, or equal to, that of the eye, which resolves better in bright light. Only with a rather poor binocular is this not the case.

There is another contributing cause to the effect you describe that goes beyond optics. At a given pupil/quality/resolution, simply having more photons improves the signal/noise ratio, because in some ways the eye behaves like a photon counter, and follows the laws of Poisson statistics. Bright scenes appear bright, dim scenes dim, proving that the eye pupil's adjustments do not fully compensate for changes in ambient light. As a result, the retina recieves more photons when it is bright, and vision is improved. If the eye pupil is closed as tight as it will go, yet light intensity continues to increase, vision will improve to the threshold of pain.

So, what you observe is supported by all that I know (or think I know) about optics and vision. I am sorry this was so complicated and high and mighty sounding, and also apologize to those whom I have differed with, but that is my best understanding, resulting from an embarrassing amount of study required to learn but a little.

Ron

RJM said:

You misunderstand. Unlike your eye, the binocular is not an adaptive optic. Its resolution is fixed and determined by the objective lens size AND its optical quality. Its resolving power does not change with brightness.

Your eye is an adaptive optic and acuity (ie. optical quality) is optimized with smaller entrance pupils utilizing the central area of the lens so astigmatism and other vision issues that affect the outer eye lens are mitigated and the image is focused on the denser receptor clusters.

Click to expand...

Are you sure about this? I suspect your pupil becomes the aperture stop for the binocular. When conditions are bright, try placing black card in front of an objective with holes of different diameters. I don't have a set-up to do it with any accuracy, but as far as I can judge, it seems to confirm that the effective aperture changes with light conditions just like a camera lens. What I don't know is how much this is how much this contributes to the perceived view sharpness. I'm sure others will correct me.

I'm sure you are correct that the periphery of the eye lens contributes to aberration, and must in part explain higher acuity in bright conditions. That paper suggests that the density of receptors contributing to the image also increases with brightness, so the retinal resolution improves.

David

ronh said:

David,
A perfect 42 mm would have 2.4 arcsec resolution, but as you note, 3.5 is "alpha" quality. The aberrations that make the difference, resulting from very fast optics and imperfect construction, become less severe as the effective aperture is decreased, as when the eye closes down in bright light. In the best 8x42s, when the effective aperture is reduced to 22 mm by the pupil which contracts to about 2.5mm in bright light, the optics although small now will behave in a nearly perfect manner, resolving about 4.5 arcsec.

The normal eye resolves best at 2.5mm pupil size, because at this size, it's "camera like" depth of field, which compensates for its imperfections, reaches a balance with its own aperture limitation on resolution. In this bright condition, the naked eye resolves 60 arcsec, so at 8x will resolve 7.5 arcsec.

So, in dim or bright conditions, the resolution of the binocular/eye system is dominated by, or equal to, that of the eye, which resolves better in bright light. Only with a rather poor binocular is this not the case.

There is another contributing cause to the effect you describe that goes beyond optics. At a given pupil/quality/resolution, simply having more photons improves the signal/noise ratio, because in some ways the eye behaves like a photon counter, and follows the laws of Poisson statistics. Bright scenes appear bright, dim scenes dim, proving that the eye pupil's adjustments do not fully compensate for changes in ambient light. As a result, the retina recieves more photons when it is bright, and vision is improved. If the eye pupil is closed as tight as it will go, yet light intensity continues to increase, vision will improve to the threshold of pain.

So, what you observe is supported by all that I know (or think I know) about optics and vision. I am sorry this was so complicated and high and mighty sounding, and also apologize to those whom I have differed with, but that is my best understanding, resulting from an embarrassing amount of study required to learn but a little.

Ron

Click to expand...

Ron,

Thanks for that. I think what you you say is pretty much in line with my observation and limited understanding.

Like you say, I suspect a binocular objective resolution while decreasing, typically improves relatively as it's stopped down. I have no way to be sure, but I imagine that like a camera lens, eye lens etc., beyond a certain point the resolution could decrease again. Say, below a 2.5 mm pupil, does the relative resolution significantly deteriorate again? Generally it appears to be that visual acuity is the limiting factor, but under bright condition the two limits get closest. Do they overlap at any point?

Could you clarify the way you describe visual resolution. 20:20 vision is 120 arcsec/line pair, but 60 arcsec using opticians convention. 60 arcsec/lp or 30 optician speak is pretty rare. I'm never quite sure which the Dawes or Rayleigh numbers equate to. My best back yard figure was 87 arcsec/lp, but the conditions were no where near as bright as when I tried out all the top brands together. I felt at the time I was seeing the resolution limit on some of the pairs which puzzled me. On reflection it's possible I was if my pupil contracted below 2mm. The prize for 'sharpest' that day went to a Kowa.

David

David,
Good for you for calling me on my resolution figures of merit. I was playing fast and loose a little bit, and will try to explain.

Eyes are called normal if they can read the Snellen letters when each block of the 5x5 array subtends 1 arcmin. Although it's not strictly a resolution test, the 1 arcmin gap in many letters has to be resolved. That would be the same as 2 arcmin/line pair. Eyes are rarely tested any other way, except for research, so that's my basis for calling 1 arcmin resolution normal. Dawes's limit for a 2.5mm aperture, however, is a gap of 40 arcsec, or 80 arcsec/line pair, which would correspond to about 20/13 vision. Sounds like your eyes are better than normal and close to perfect.

Since I'm used to talking about astronomical scopes, I always think of instrument resolution in terms of Dawes's limit, and remember 4 inches resolves 1 arcsec. Scaling that to 42mm is how I got the 21arcsec resolution value for a perfect binocular. Dawes looked at two points, but the gap seems to me the important feature, so I believe 21 arcsec Dawes would correspond to 42 arcsec/line pair.

I thought it was ok, at the time, for to me to mix and compare these two measures, discerning the gap seeming to be the crucial performance, but I'm having my doubts now.

Could eyes like yours challenge a well made binocular, and see resolution differences in good lighting? This is a question that gets tossed around here occasionally, with simple optics/vision estimates almost always concluding "no", unless the binocular sucks royally, yet people continuing to report seeing differences in "sharpness", which are sometimes supposedly attributed to contrast differences, etc., ie. something besides resolution. If my original argument is repeated for the perfect 2.5mm eye, with 8x it should resolve 5.6 arcsec, still worse than the effectively 22mm but optically perfect 8x42 binocular, and the answer is still "no", but barely.

My argument about mixing Snellen and Dawes measures could have a flaw in it that is not obvious to me but make the little bit of difference needed to change the conclusion to "yes". The two measures are quite qualitatively different, and it may not be right to pick out the gap as the only important parameter. Certainly, for the 20/15 eye, it's awfully close to possible even by that possibly flawed reasoning, and seeing resolution differences seems possible indeed with less than perfect binoculars. I think you need to actually look at resolution charts with the different binoculars, mounted, to know for sure. I'd love to try that Kowa though!

To answer your first question, I have seen the point spread function of the eye's optics plotted vs the pupil opening. It's sharpest at 2.0-2.5mm, surprisingly close to perfect in the upper reaches, a bit fat around the base. For larger pupil sizes, aberrations kick in, and for smaller, it's the aperture limit.
Ron

Ron,

Many thanks for that. When I get a chance I'll see if I can measure some resolutions with reduced entrance and exit pupils to see if everything hangs together.

David

ronh said:

David,
Good for you for calling me on my resolution figures of merit. I was playing fast and loose a little bit, and will try to explain.

Eyes are called normal if they can read the Snellen letters when each block of the 5x5 array subtends 1 arcmin. Although it's not strictly a resolution test, the 1 arcmin gap in many letters has to be resolved. That would be the same as 2 arcmin/line pair. Eyes are rarely tested any other way, except for research, so that's my basis for calling 1 arcmin resolution normal. Dawes's limit for a 2.5mm aperture, however, is a gap of 40 arcsec, or 80 arcsec/line pair, which would correspond to about 20/13 vision. Sounds like your eyes are better than normal and close to perfect.

Since I'm used to talking about astronomical scopes, I always think of instrument resolution in terms of Dawes's limit, and remember 4 inches resolves 1 arcsec. Scaling that to 42mm is how I got the 21arcsec resolution value for a perfect binocular. Dawes looked at two points, but the gap seems to me the important feature, so I believe 21 arcsec Dawes would correspond to 42 arcsec/line pair.

I thought it was ok, at the time, for to me to mix and compare these two measures, discerning the gap seeming to be the crucial performance, but I'm having my doubts now.

Could eyes like yours challenge a well made binocular, and see resolution differences in good lighting? This is a question that gets tossed around here occasionally, with simple optics/vision estimates almost always concluding "no", unless the binocular sucks royally, yet people continuing to report seeing differences in "sharpness", which are sometimes supposedly attributed to contrast differences, etc., ie. something besides resolution. If my original argument is repeated for the perfect 2.5mm eye, with 8x it should resolve 5.6 arcsec, still worse than the effectively 22mm but optically perfect 8x42 binocular, and the answer is still "no", but barely.

My argument about mixing Snellen and Dawes measures could have a flaw in it that is not obvious to me but make the little bit of difference needed to change the conclusion to "yes". The two measures are quite qualitatively different, and it may not be right to pick out the gap as the only important parameter. Certainly, for the 20/15 eye, it's awfully close to possible even by that possibly flawed reasoning, and seeing resolution differences seems possible indeed with less than perfect binoculars. I think you need to actually look at resolution charts with the different binoculars, mounted, to know for sure. I'd love to try that Kowa though!

To answer your first question, I have seen the point spread function of the eye's optics plotted vs the pupil opening. It's sharpest at 2.0-2.5mm, surprisingly close to perfect in the upper reaches, a bit fat around the base. For larger pupil sizes, aberrations kick in, and for smaller, it's the aperture limit.
Ron

Click to expand...

Thanks to you and typo for a most informative thread. I always enjoy learning something new and I have already put this info to good use in my latest bino comparison :t:

Since that last post I've checking out for myself what effect light intensity has on my acuity, and generated a limited amount of data for some of my binoculars at full aperture and stopped down.

That data I mentioned from Koenig in 1896 clearly showed that acuity improves with light intensity up to a certain point. After a bit of googling and guess work on the conversions I believe that test subject's acuity plateaus above 2000lux with an acuity of about 70 arseconds/line pair or 20/12. (Why lux? It's the readout from my budget light meter.) 2000 lux is about what I got in the shade on a sunny April day in the UK.

Just using points inside and outside the house, I pinned up a copy of the USAF target, took a light reading, measured 15ft away and read the chart. Pretty primitive, but I got my best result at 3000 lux with 64 arcsec/lp or 20/11. The other points pretty much map directly to Koenig's data.

That acuity reading was rather a surprise. In my youth my vision was excellent, but I'm now in my late fifties and been wearing glasses for the last 15 years. In the past my home tests had been coming out at about 90arcsec/lp or 20/15, but I've now checked it a many times over the last few weeks when the light has been good, and got values mostly around the 20/10 to 20/12 mark, but there was one day when I repeated got readings at 10/8. I guess it's just a happy coincidence of my prescription currently perfectly matching the state of deterioration of my eyes at this moment and the mess of floaters I've got clearing a path that day. There is a lesson there; it pays to be on good terms with your optometrist.

The point of doing this was to really see how this relates to binocular performance, when your eyesight is optimal. I couldn't measure my own pupil diameters with any accuracy, but is is reported that optimal acuity occurs at about 2.5mm. With the weather we've been having I've not been able collect as much data as I wanted but a few things have emerged. For clarity I'll skip most of the numbers.

The binoculars I have, range from an optically excellent 12x50, two reasonable 7x to a pretty terrible 10x42. I've most data for the 12x50 and the 7x36. I've taken a number of unboosted and boosted reading at full and reduced aperture and compared it to my acuity at the time.

Unboosted, when the light is bright enough for optimum acuity, stopping down the aperture with black paper mask makes no difference to apparent brightness or apparent resolution, so the pupil must act to reduce the binoculars' effective aperture.

In boosted tests the 12x50 is very good at full aperture, and close to perfect when stopped down. The 7x36 and the 7x26 are quite poor at full aperture, but relatively improved when stopped down. The 10x42 is bad at full aperture and stays fairly bad when stopped down.

It's often stated that the true resolution of binoculars will normally far exceed the resolution of the eye and will not be limiting, but generally the full aperture resolution values are quoted as evidence. The Dawes limit for a 50mm objective is 2.32 arcsec, for a 36mm it's 3.6, but for 7x binocular for example for the optimum acuity 2.5mm pupil diameter it equates to only a 17.5mm objective which has a 6.6 arcsecond Dawes limit. That is much closer to the magnified acuity of the eye. How close is it and does it make any practical difference?

For someone with 20/20 vision the resolution magnified 7x should be 120/7 = 17 arcsec/lp which is still far removed from the theoretical limit of 6.6 for a 17.5mm objective but for someone with 20/10 vision it would be 8.5 which is much closer and is much more likely to see problems.

Just reading the apparent resolution from the chart with binoculars, multiplying the result by the magnification and comparing it to your acuity measured at the same time appears to be a good indication of whether the binoculars are limiting the resolution potential of your eyesight.

Admittedly with limited data so far it appears to hold up well. For my 12x50 the results with and without binoculars are virtually identical. For my 7x36 it is 25-40% worse with the binoculars, and for my poor 10x42 it is about 100% worse. It appears to be confirmed by data from stopped down boosted results, though that is of course trickier to do. It suggests that the cut off is very close to where the magnified acuity equals the stopped down resolution.

The results of course will be very much dependant on the eyesight of the user. For instance if my eyesight had been a much more common 20/15 when I did the test neither the 12x50 or the 7x36 would be limiting. If it was 20/20 then even my poor 10x42 would have been relatively OK.

I hope this was not too confusing. The bottom line is that stopped down boosted resolution testing appears to be a much better indicator of binocular performance than full aperture testing, and for many people the relatively simple method of reading a resolution chart with and without binoculars offers an easy way to determine if their binoculars optical performance is limiting.

David

typo said:

The bottom line is that stopped down boosted resolution testing appears to be a much better indicator of binocular performance than full aperture testing, and for many people the relatively simple method of reading a resolution chart with and without binoculars offers an easy way to determine if their binoculars optical performance is limiting.

David

Click to expand...

Ok yes. the bottom line is when you are aiming to be out in the field using the bins what you say here is the most practical test to give an idea of potential or even better forget the chart and just get out in the field.

Keep it simple!

If you want to design bins then thats something else but ultimatley its what they can do in the field, not in a lab and there ain't no lab is ever going to be a field condition with all its variations. Thats what probably fascinates us optic nuts but ultimately it should be what we can observe with most awe in the field dealing with all those different light variations. N'est pas!

Clive,

The background to this was out in the field when I found I was having trouble with sharp focus with that 7x36 when the conditions were bright. It's taken a bit of reading up on stuff, and messing around with different conditions, but that simple test was ultimately sufficient to tell me why. I like to know why.

David

Ok, we'll see.

Just keep cool. Thats an order.