• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
ZEISS DTI thermal imaging cameras. For more discoveries at night, and during the day.

Sharpness and resolution, one subject or two ? (1 Viewer)

Renze,

Thanks for the link to Tobias' very interesting article.

Below is an MTF plot from H. Suiter's book "Star Testing Astronomical Telescopes". It shows the detrimental effect on contrast from one almost universal aberration in binoculars, lower order spherical. If we assign to this plot a 20mm aperture to correspond to the effective aperture of an 8x binocular stopped down by a 2.5mm eye pupil in daylight then the 100% spatial frequency would be around 6 arc seconds. A person would need a little better than 20/8 eyesight to resolve that spacial frequency when it's multiplied by the binocular's 8x magnification to 48". Very few humans have eyesight acuity that good. A more typical human acuity (around 20/15) would fall at a spatial frequency of about 50% of that or around 90-100 arc seconds. I placed a red line on the chart at approximately that acuity. If we accept the Panavision standard of 30% contrast for a "sharp" appearing image it can be seen that an instrument corrected to 1/4 wave can meet that standard (for 20/15 acuity), but a 1/3 wave instrument can't, even though no smaller line pairs would be resolvable through the better instrument since the eye can still resolve line pairs at 20% contrast or even worse.

Most binoculars have a large amount of spherical aberration and many individual specimens have other defects on top of the SA. Total aberrations approaching 1 wave are common at full aperture (even in "alpha" brands) and when the effective aperture is reduced in daylight most are still worse than 1/4 wave. It's a wonder that any of them look "sharp".

Henry
 

Attachments

  • DSC_0979.JPG
    DSC_0979.JPG
    300.9 KB · Views: 80
Last edited:
Henry,

In theory with an 8x my magnified acuity should get me to about 7.5" or fraction better on a good day. It's puzzled me how it is possible that I can with care tell that a binocular with a 6.2" stopped down resolution is better than 6.8" for example. I can think of one model I tried where from just viewing the USAF chart it appeared that the MTF curve might be quite concave like the 1/2 wave curve or worse, but generally the contrast reduction rate appeared very similar and quite acute for the better binoculars I've tested. It might be the relative contrast at the acuity limit that is actually the most important criterion as you suggest, but with only one exception so far the actual effective resolution appears to be a pretty good predictor of perceived sharpness.

David
 
Last edited:
David,

I know you have extraordinarily fine acuity. I can only guess that you are picking up on differences in tiny bars that would be quite invisible to me in the 10-20% contrast range.

Henry
 
Awesome data, Henry!
That's for a single objective lens, though, correct?

Most optical systems play various aberrations off against each other..
so things aren't quite that bad. The field lens and/or the flattener
before it bear most of that load. Still, those compensations have their
patterns as well (not the same periodicity in that graph),
thus the phenomena like the "fuzzy donut", and the
severe deterioration of most binoculars at the edge.

If we only had 2 elements, things would be bad indeed,
and the eyepiece element would actually multiply the
errors you show. The rest is adding things to wrestle with the beast.

Note that a test with many many line pairs is much more challenging than
detecting a single 'half pair', like the edge of a beak against the snow.
The many line pairs model fills in from two peaks instead of one,
so the contrast is a lot less. In a nutshell, most object you see
have dark|bright edges, not the dark|bright|dark|bright...etc of the tests.
(so everyone typically sees with more sharpness for most things).
 
Last edited:
Henry,

I'm know some of the youngsters will do better still. I certainly need a lot more light than I used too.

For the finest distinctions it's easier to use a chart for comparisons, but if the the light is good it's usually sufficient to pick out some feature in the tops of trees or something to spot differences of 0.5" or more. That covers a lot of binoculars! The thing that took me some time to realise is how critical the light level is when making comparisons. I've found the meter I was using can give spurious results in some circumstances so I've been working on a better way of estimating luminance and will probably post something when I've done a bit more checking.

David
 
Last edited:
I'm a believer in fonts. Partway between dozens of lines and the actual critters you look at.
Optometrists say.... resolution is 1/5 if the sans-serif font height (to be able to read the words).
It seems more practical than twin star points or many many lines blurring.
You see with different resolution for different target signatures, just like musicians,
sonar, spy scopes, radio signals, etc... The edge of a bird's eye isn't ||||||||||||||.

Not saying you can't use USAF numbers....in signal-processing theory and engineering, though,
single-edge detection is about twice as fine as line-pairs.
 
Last edited:
OPTIC_NUT said:
Awesome data, Henry!
That's for a single objective lens, though, correct?
Click to expand...

No, the figure from Suiter's book shows the effect of spherical aberration on the MTF of any image forming optic, whether it's a mirror, a lens or a complete telescope like a binocular. Star-testing a binocular for spherical aberration means looking through the whole thing, not just the objective lens.
 
Henry and David know this, but for many readers of this thread the following elaboration might be useful.

When looking at the highly illustrative figure Henry provided, we should keep in mind that it still shows a simplified scenario where there is only one aberration, SA, present to degrade the image. Aberrations stack, so the end result will be the sum of all aberrations, not just the limit set by the worst one. With even very high quality telescopes, there are often small amounts of other aberrations present in addition to SA, and with binoculars there practically always are, sometimes quite severe. Most common of these are miscollimation (de-centring of optical elements within a single binocular barrel) and, in roof prism binoculars, astigmatism-type light scatter from a prism roof edge that is not precise enough.

The presence of additional aberrations will make the MTF curves distinctly worse than the ones in this graph, and unlike SA which does improve with reduced eye pupil, they tend to be constant across pupil diameters.

One aberration that people usually tend to forget about when considering what is a "good enough" quality level for binoculars or telescopes is de-focus. We of course try to optimally focus our binoculars on the target, and if we succeed the de-focus aberration is zero for that distance. For any other distance, de-focus sums with the other existing aberrations, and the more of these there are the more precisely does one need to focus and the shallower the effective depth of field becomes.

Kimmo
 
Last edited:
Thanks Kimmo for explaining that.

I've not seen published comparisons of MTFs for varying objective diameter but Kimmo may well be right that other aberrations may be more constant. From my own tests a bad resolution at 42mm can be either good or bad at 20mm though a very good result at 42mm will usually still be good at 20mm of course. I'm more than happy if it's good at 20mm.... usually.

There was a fairly inexpensive 8x42 binocular I had to review which in casual testing might look quite good for one view and quite poor for another. It turned out that 20mm it gave a fairly typical resolution figure for the price but at 25mm and 30mm the results were significantly worse before recovering a bit at 42mm. I was getting quite obvious variations in sharpness depending on the luminance of the scene. Good when it was bright, poor when it was gloomy and OK at twilight. I wouldn't know if this was spherical aberration or some other problem but at least one other user reported a similar pattern. That was an extreme case, but I suspect smaller degrees of variation are quite common.

David
 
Last edited:
kabsetz said:
considering what is a "good enough" quality level for binoculars or telescopes is de-focus. We of course try to optimally focus our binoculars on the target, and if we succeed the de-focus aberration is zero

Kimmo
Click to expand...

Hi Kimmo,

I, of course, agree with you and Henry on this. In order to simplify and make MTF plots more useful for general comparisons I will relate some shortcuts I have found useful over the years.

First, a general premise or assumption or two that I regard as accepted. Any aberration present results in a defocus of part or the entire image. The various lumps, bumps and bulges and other displacements are the result of these accumulated errors.

For the sake of simply comparing one instrument to another and keeping track of them in the same terms for consistency I just compare my plots to the general defocus chart instead of trying to rate each by SA, astigmatism or whatever.

I think you understated the importance of proper manual focus; this is the most common and largest source of error in a lot of the plots.

When I do a test I first focus the instrument with a highly boosted image and have a micrometer drive attached to the focus control, leave the optic at that setting while positioning the imaging system. The target is at the focus of a collimator that is pretty close to ISO standards. Once I have the IS in place I then take images at 1, 2 and 3 mm either side of my starting focus point. Without disturbing the IS as much as I can, I take the files to the computer and run a rough curve reduction using the assumption that the best focus results in the best curves, If there is to great a spread in the curves I go back at start at the best point and go in 0.1 mm steps either side and repeat the process. Once I am satisfied I have the best manual focus I then image at vertical, horizontal and 45° and take the middle curve as average. Although I have a very stable imaging platform it is not unusual for different setups to be off infinity focus by several tenths of a diopter, worse with wide lenses. Occasionally even this procedure is inadequate.

I have attached a pdf from the book Henry mentioned for those interested to look at. Also attached a file of some of my older tests.
 

Attachments

  • MTF Suiter.pdf
    1.9 MB · Views: 49
  • Pocket Bins Average 1.jpg
    Pocket Bins Average 1.jpg
    76.1 KB · Views: 59
henry link said:
No, the figure from Suiter's book shows the effect of spherical aberration on the MTF of any image forming optic, whether it's a mirror, a lens or a complete telescope like a binocular. Star-testing a binocular for spherical aberration means looking through the whole thing, not just the objective lens.
Click to expand...

I am confused...a complete telescope has the same curve?
you are saying the curve for an entire system with many elements
is identical to one element? If I were to look at the eyepiece response for
a simple combination, where aberrations were pitted against each other,
a Ramsden, Kellner, Plossl, Erfle.... there are mutiple ripples.
Those designs do not function without that interplay.

Unless.....you mean that the MTF is dominated by one element (?)
That's what I see in the plot....just one element's errors.
That's sort of true ... the longer focal length of the objective controlling
the results. Seems like a major missed opportunity in tuning the system.
In that case, a middle-ground element like the "field flattener" would
seem like just the tip of the iceberg for designing a more wholistic response.
 
Last edited:
Ron,

Thanks for the material.

I'm sure I understated the importance of focus if we are thinking of obtaining actual MTF curves with a test setup. There, obtaining a precise focus would be paramount. My starting point here was rather what happens when we use the binoculars, and I'm using MTF curves here rather as a diagrammatic way to better understand and "visualize" what happens with the image. As far as I know, you are the only one of us here on BF who has been able to put together a test setup and has learned the necessary skills to actually do MTF testing at home. The rest of us can only study model curves and do simple resolution testing. I have enormous respect for what you have been able to achieve.

In visual use, our eye accommodation to a larger or lesser extent compensates for the defocus aberration, but my point is that the more there are of other aberrations, the less effective this becomes and the more problematic the image.

Kimmo
 
OPTIC_NUT said:
I am confused...a complete telescope has the same curve?
you are saying the curve for an entire system with many elements
is identical to one element? If I were to look at the eyepiece response for
a simple combination, where aberrations were pitted against each other,
a Ramsden, Kellner, Plossl, Erfle.... there are mutiple ripples.
Those designs do not function without that interplay.

Unless.....you mean that the MTF is dominated by one element (?)
That's what I see in the plot....just one element's errors.
That's sort of true ... the longer focal length of the objective controlling
the results. Seems like a major missed opportunity in tuning the system.
In that case, a middle-ground element like the "field flattener" would
seem like just the tip of the iceberg for designing a more wholistic response.
Click to expand...


OPTIC_NUT,

Take another a look at the plot. There's no mention of it applying only to a single element. The curves simply show the effect on MTF from varying amounts of wave front error due to SA. It doesn't matter whether the error is after light has passed through one lens or is the total error after light has passed through a series of lenses.

Henry
 
It might be pointless to point it out, but serious
equipment has different MTF charts.
I think I have been spoiled by working alongside some
mil-spec engineers. Equal-ripple responses are designed,
and this is far away from those systems.


It does bother me that might be all you're getting for $2000.

I would be interested to see the plots of various eyepieces
with long-barreled systems (so the objective is less important).
This isn't the place, though.. different assumptions and practices.

I still suspect some finer designs do have more complex OTFs.
A 'field flattener' is just the beginning.
 
Last edited:
Ron,

I'll second Kimmo's appreciation for your excellent work in this area. It's a shame we aren't supplied with MTF's for binoculars, although I can imagine problems. For instance, I'm sure what we would get from manufacturers would be ideal MTFs of perfectly executed designs, something we seldom see in the real world. On the other hand MTFs of actual binocular specimens in reviews would be subject to the usual sample variations.

Henry
 
Last edited:
Henry, post 35,
I can understand very well, that manufactureres do not supply us with MTF plots, since probably 99% of the customers will not understand these curves and it may lead to many confusing discussions in consumers magazines is my guess.
Gijs
 
henry link said:
Gijs,

I'm sure you're right about that, but for some reason the manufacturers of camera lenses do supply MTFs. I don't know what kind of confusion results from that in the photography world.

OPTIC_NUT,

I don't think I'll be hiring you for any engineering work. I'd be afraid it would turn out like this: ;)

http://upload.wikimedia.org/wikiversity/en/9/91/Poster.gif

Henry
Click to expand...


Heh...the systems I've seen are little messier than
binoculars, but not much. I cannot seem to explain it here, though.
The sacred scrolls are too old.

Paradigms are funny things, as are wants.
They get you places, but in a tunnel.
Thus my astonishment when I look out across this winter
forest with Meopta Meopros (2013) and Scope 3010s(1970s?),
and wonder what all the fuss was about. Under some conditions,
the latest stuff seems a lot less sharp and saturated.


But, to focus on spherical aberration:

1) Would you deny that eyepieces have elements
that compensate for each other's aberrations?
I keep reading descriptions of eyepiece designs that mention that.

2-a) Do you believe the MTF chart for all lenses and all systems of lenses
looks the same, just scaled somehow? There doesn't seem to be a point
in measuring something that's pretty much the same.
...or is it the general shape...that there is always a singular fall-off to a
diffraction-pattern appearance?

2-b) Actually........this could be true for almost all binoculars made,
from the same desisgn paradigm. So....maybe that is true,
but not for everything optical. Or...maybe it's possible to be
different but way too costly. I didn't ask about costs....but still,
there are $2500 binos.

I think you are saying 2-b, sort of, but you say that covers all optics everywhere.

Something's missing from this picture.
 
Last edited:
All,

After reading Suiter's "Star Testing Astronomical Telescopes," 1994, several years ago, it became clear to me that his emphasis was entirely on telescope objectives. Briefly reviewing his logic:

  • In the last paragraph on Pg. 37 he says: "The filters covered in this book concentrate on the center of the stack, from the atmosphere down to the image inspected by the eyepiece." [see below for the "stack" of filters]
  • On Pg. 38 (top) he says: "Eyepieces are neglected here..."
  • Throughout the book he relies on photographs, implying that the camera is an adequate simplification of the eye.
  • His regard for the eye in the astronomical observation task is foreclosed in the next to last paragraph on Pg. 37: "The only things that can be done ... for the awkward distribution of low-light sensors on the retina is not to look directly at dim objects." [hic !]
  • With regard to MTF graphs he states (Pg. 53): "... This book concentrates on the theoretical description of ideal cases." In other words, the graphs are generated from theoretical equations, not empirical data.
So, when he gets to the topic of "Filtering of Spherical Aberration" (Pg. 176, Fig. 10.5), which Henry included above, we are looking at his generalized theoretical curves pertaining to telescope objective aerial images, devoid of influences by eyepiece or eye.

For these reasons I'm befuddled by how far these graphs can be generalized to low-power birding binoculars, much less how perceptions at the end of the filter stack can be predicted or better understood. Suiter simply doesn't mention terrestrial binoculars, ... the ones we look through that have all those nasty effects he avoided.

Don't get me wrong, I'm not against star testing. But I do appreciate Warren J. Smith's admonitions in "Modern Optical Engineering," (1990), pg. 501. (second attachment)

Ed
 

Attachments

  • Book pg. .jpg
    Book pg. .jpg
    117 KB · Views: 50
  • Star Test Smith .jpg
    Star Test Smith .jpg
    13.9 KB · Views: 38
Last edited:
Ah, that's sort of an even more holistic view of the confusion I am having.

I believe Henry is concentrating on the spherical aberration because he assume it's dominant.
That assumption seems about right.

ElkCub: you were questioning doing things like star-splitting with binoculars.
The makes sense given the lower powers and the eye's acuity .... seems like it's in the wrong ballpark.
Things like 'seeing' also seem mostly beneath relevance.

If we look at the actual resolution of most binoculars, it almost always finishes up
well inside the apparent angle for most people's acuity, so it's clear the total aberration is OK.
I think many binoculars push that limit by reducing the length as much as possible.
Low power is a defining feature for binoculars, but so is compact length.



Henry:

As I read more and more, I realize my fuss over the shape of the MTF doesn't mean much
in general and for portable sporting optics especially. Thank you so much for your patience.
I do think it can't be too alarming if actual binoculars have sub-human-acuity resolution anyway.

An interesting topic creeps in as I look at your graph and the measured graph, though:
the outer fringes of the MTF have consequences on contrast. I hadn't realized that
before, but it does show up for reflectors, as one example. I worry about that more,
the accumulation of all the fringes of all objects with the MTF.
 
Warning! This thread is more than 9 years ago old.
It's likely that no further discussion is required, in which case we recommend starting a new thread. If however you feel your response is required you can still do so.

Similar threads

P
Resolution sharpness and contrast.
2 3
Replies
51
Views
4K
Maljunulo
M
G
Leica, Zeiss & Swarovski binoculars compared: A personal report
2 3
Replies
42
Views
4K
Paultricounty
P
P
Subjective, non-analytical comparison of three 832 Premiums.
2
Replies
22
Views
3K
Paultricounty
P
dorubird
Swarovski Swarovision EL 8x32 vs Nikon Monarch HG 8x30
2 3
Replies
45
Views
5K
Owlbarred
Owlbarred
looksharp65
EDG 7x42 and Meostar B1.1 8x32 head to head
2
Replies
22
Views
2K
Scott98
S

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top