What is the 'sweet spot' exactly? (1 Viewer)

Alexis Powell said:

The term "sweet spot" was never intended to be a precisely defined quantity that is subject to absolute measure, it's just a quick way of making relative comparisons between binoculars. Typically, the term is used when discussing the relative merits of two binoculars that differ greatly in field of view. For example, "Model A has a very wide field of view but the sweet spot is so small, I'd rather have model B because although it has a narrower FOV the sweet spot is much wider." --AP

Click to expand...

One of the issues with "sweet spot" is that FOV tends to get involved. Your statement is a good example, — and as usual I come away perplexed. A fixed retinal projection area affording a high level of visual acuity (as Henry described) probably would be perceived as relatively small within a wide field and relatively large within a narrow field. Assuming that to be a perceptual fact, I don't follow how one can compare the merits of two binoculars on this basis.

Elk

elkcub said:

I don't follow how one can compare the merits of two binoculars on this basis.

Click to expand...

You just describe what you see.

Leif said:

You just describe what you see.

Click to expand...

Precisely!

I prefer to estimate the sweet spot size as an area of apparent field in degrees. A 30 degree sweet spot is the same size in any binocular no matter how wide or narrow the total apparent field may be.

I think all of us who have experienced various sweet spot sizes in the field recognize that the concept is generally valid if not precisely measurable. After using many binoculars and trying to measure their sweet spots I've formed my own personal set of loose standards based on how my pupils tend to roam about within the field. For me a binocular with a sweet spot of 20 degrees or less will certainly be unpleasant to use because my gaze will frequently stray out of the sharp area. A sweet spot of 35 degrees or more will never cause any problems, but I've seen sweet spots that wide in only a few binoculars that use so called "field flatteners" in the eyepieces. Somewhere between 25 and 30 degrees seems to be the threshold between comfort and discomfort in the field for me. To that I would add that binoculars in which off-axis astigmatism is the dominant aberration limiting the size of the sweet spot tend to be more unpleasant for me to use than binoculars in which field curvature dominates.

henry link said:

... A 30 degree sweet spot is the same size in any binocular no matter how wide or narrow the total apparent field may be.
...

Click to expand...

Precisely, — although it may also be perceived as smaller within a larger apparent field.

I was thinking along similar lines, and tempted to write that when the SS is defined as the largest angle within which a criterion level of acuity can be maintained, its size should be inversely related to observable head motions. I'm glad to hear your experience tends to confirm this, as does mine.

henry link said:

...A sweet spot of 35 degrees or more will never cause any problems, but I've seen sweet spots that wide in only a few binoculars that use so called "field flatteners" in the eyepieces. Somewhere between 25 and 30 degrees seems to be the threshold between comfort and discomfort in the field for me.

Click to expand...

Just to follow up on this important point, your angular estimate jibes quite well with how the eye/head normally work. In general, there is a built in coupling between eye motion and head motion. Considering the lateral dimension (vertical is similar), the tendency to move the head increases as ocular rotation approaches 20˚ from center, and becomes compelling thereafter due to facial geometry, particularly the nose. So, the range of what seems "natural," would be captured by about 35˚, i.e., ±17.5˚, of the apparent field. Below approx. 25˚, i.e., ±12.5˚, more head motions would be required than are "natural," and there might well be increasing unpleasantness associated with that.

My reason for capitalizing on your comments is that this aspect of binocular assessment is, in many ways, more approachable within the framework of bio/behavioral science than optical science, per se. In that light, this aspect of design may also be more important than first imagined.

Elk

The attachment is taken from Human Engineering Design Data Digest, MIL-STD-1472.

henry link said:

I prefer to estimate the sweet spot size as an area of apparent field in degrees.

Click to expand...

Exactly! And such comparisons do not require that the angle be measured. When two binos are compared side by side, it is easy to see which renders a larger amount of a large object in pleasing focus/image quality. The field of view does not complicate such comparisons.

--AP

Quote: 'In the case of a 10x32 this would be close to the Dawes limit so I would not expect to get to twice visual acuity. The ISO standard for 10x32 would call for resolution better than 9.38 arc seconds.'

Ron

HNY

This is often where the difficulty lies with meaningful discussions around binocular and scope resolutions, and what individuals prefer to use as a test.
With my background, diffraction is the bottom line. An instrument can be tested to see whether it meets the resolution limit set by Rayleigh or Dawes, and as such can be described as being limited only by diffraction. Similarly, I can place an instrument on a testbench and look at the figure, or perhaps use an optical interferometer. This will enable a fairly safe assessment of whether the instrument will meet the diffraction limit of that aperture or not, or even what level of contrast can be expected that is due to figure. 99.9 times out of 100, if an instrument is tested for diffraction limit afterwards, the assessment proves correct. Sure, individual acuity may interfere with diffraction testing, but at higher powers this is less likely.
Then we get into the more murky area of bar charts and Snellin charts, or even sinusoidal testing with tests of lowering contrast. I say murky because with a testbench, interfering obstructions such as the eye (acuity, diameter of pupil etc), play much less of a role so the test results are much less open to interpretation.
Internal reflections can reduce centre-field contrast, but do not (usually) affect a diffraction limit test.

The problem with just describing "what you see" as another has suggested, is that the results can not be trusted. There are simply too many variables - exit pupil versus eyepupil, aberrations in the observer's eye, environment - heat conditions. Then we have to trust that the observer can accurately judge what is 25 degrees or so in a 60 or 70 degree apparent field. The state of health of the individual also makes a difference.
Here's an example. A 90mm Meade Maksutov Cassegrain is used as a spotting scope. The observer wishes to test the "sweet spot" with a favourite eyepiece - lets say a converted 21mm Leica eyepiece. At about 60X he decides to use an evenly illuminated brick wall as a target. An internal reflection creates unfocused diffuse light at the focal plane. The amount of diffuse light creates a condition where the boundary between the central area of sharp detail and where it ceases to be sharp becomes a little blurred. He estimates about 30% of the central field is satisfactory. Now he adds a dewshield (lens shade) of about 5" length. When he looks again, contrast has improved quite notably, and the sharp area of the field has risen to about 35%.
A decade or so ago, I used a pair of old Swarovski 8X30 SLC binoculars frequently for birdwatching. I used to take around with me a pair of homemade lens shades in the form of black painted 3" long cardboard rolls. In certain conditions they reduced internal reflections dramatically, improving contrast considerably, and making detail away from the centre of the image visible.

As an engineer you will appreciate the importance of scientific and objective testing and information. We can all judge by our own standards, but then there are as many standards as judges.
For me, a proper testbench method is the only meaningful method for assessing any real optical aspect of an instrument. Other than that, point sources in darkness at least allow the condition where the exit pupil is the defining pupil.

andytyle

andytyle said:

The problem with just describing "what you see" as another has suggested, is that the results can not be trusted. There are simply too many variables - exit pupil versus eyepupil, aberrations in the observer's eye, environment - heat conditions. Then we have to trust that the observer can accurately judge what is 25 degrees or so in a 60 or 70 degree apparent field. The state of health of the individual also makes a difference.

Click to expand...

I have to disagree with you. I have no objection to having objective tests, but IMO an experienced observer can reach an assessment of an instrument that often surpasses the results of tests.

For example, some instruments behave fine when tested for resolution and contrast, but in certain situations they fall to pieces. That fact would not be picked up by most tests despite the fact that the tests are far more accurate, precise and objective than any human observer. As you say, "there are simply too many variables".

I will give you an example. The Zeiss 8x30 Classic binocular is a very fine instrument, and can give lovely views. But use it at dusk, and you may see serious flare that interferes with observations. The Swarovski 8x5.42 EL is another very fine instrument, but in some circumstances it can exhibit substantial CA both on an off axis, and I find that objectionable. Admittedly many if not most users do not seem to notice the CA, but it exists and I do not not like it. In other words, it is very hard if not impossible to design a set of tests which gives an accurate assessment of an instrument.

Getting back to the sweet spot, in some instruments such as the Swarovski 8.5x42 EL, the image degrades gracefully away from the optical axis, and is still okay at the edge. In the Nikon 8x42 HG the image is remarkably good to the edge. No doubt in some others the image is almost perfect to half way to the edge, then degrades rapidly. So how do we rate each instrument? Do we favour a slow graceful decline, or do we favour no decline to half way to the edge, followed by a rapid decline? Clearly the choice is subjective. I suppose we could draw graphs showing resolution as a function of angle from the axis, and allow the reader to decide. But would that really give them a good feel for the instrument?

Or we would just describe what we see, and when several people start saying the same sort of thing, we can be reasonably confident that a) the observer is a reasonable judge and b) the observed behaviour is characteristic of a given model and not just one sample. Alternatively just try the instrument for yourself and ignore other peoples prejudices.

I’d like to add just a few comments to the nice and lively discussion going on this resurrected thread.

First, to Andytyle concerning mainly the following quote as well as what you say in post #68:

[So, if we accept that a binocular being manufactured to a standard where a diffraction limited area at center-field is very rare, (lets say the prisms are the problem in higher quality binoculars where the objectives may well be of good enough quality). The "sweet spot" is thus not a universal concept, because of, as you say, area of OG being used because of pupil size, individual user acuity etc. So perhaps it is meaningless to even mention it.
However, the problem is that visual acuity is a factor in every case of using an optical instrument, to some extent. The only way to determine whether concepts like "sweet spot" are meaningful, is to treat each instrument as a black box.
Its that test bench thing again. All else is opinion and interpretation.]

As the discussion between S.I. (in post #19) and Alexis, Elk, myself and others shows, current test bench procedures do not seem to be quite universal enough in their applicability to average (or exceptional) users’ perceptions. The test bench procedures used by Zeiss seem perfectly reasonable on the face of it, but when the hard proof they provide is uncorroborated or contradicted by the end users’ experiences, I find I cannot easily dismiss more subjective evaluations in their favor. So, even if a “sweet spot” is not easily quantifiable or is precisely quantifiable only for a particular individual under particular viewing conditions, I still find it a useful concept. In fact, much the same arguments as presented against the SS could be leveled against visually measuring edge-of-field resolution in general. For a user, it does matter how far off-axis and how easily one can look and still perceive a satisfyingly sharp image. If a reviewer or a user discusses these factors, it can (and should) certainly be called opinion and interpretation, but for another user it may nevertheless be highly valuable and sometimes (unfortunately) more valid than a full lab report of hard data when the procedure for obtaining the data turns out to miss (or misrepresent) some important aspects of real human use.

Another brief comment concerning your bracketed sentence about problems in higher quality binoculars. Although prisms can sometimes be the main culprit, I find (based on booster-assisted point source evaluation of visible aberrations) that the most common cause for less-than-optimum optical performance is misalignment of elements in the optical train. At least for the top manufacturers and their binoculars, prism manufacturing tolerances seem to be “good enough” not to introduce marked aberrations. With telescopes, the situation is not quite so good (because the magnifications are higher), but with binoculars usually the only prism-induced artifact I see is a very thin diagonal line of light splitting the image of a bright point source if I view the point source in darkness or near darkness and place the point source on the prism roof edge.

Next, just briefly, to Surveyor,

[In the case of a 10x32 this would be close to the Dawes limit so I would not expect to get to twice visual acuity. The ISO standard for 10x32 would call for resolution better than 9.38 arc seconds.]

The Dawes limit is based on double-point-source separation, and since the Snellen visus is based on single-line width, a 10x32 can in fact provide twice visual acuity for all but the very most fortunate among humankind. The real boosted resolution limit of the best 10x32’s I have measured is just under 2 arc seconds in single-line width, so you still have plenty of margin left.

Lastly, I can just quote Henry as I find almost nothing to add to his summary in post # 64,

[I prefer to estimate the sweet spot size as an area of apparent field in degrees. A 30 degree sweet spot is the same size in any binocular no matter how wide or narrow the total apparent field may be.

I think all of us who have experienced various sweet spot sizes in the field recognize that the concept is generally valid if not precisely measurable. After using many binoculars and trying to measure their sweet spots I've formed my own personal set of loose standards based on how my pupils tend to roam about within the field. For me a binocular with a sweet spot of 20 degrees or less will certainly be unpleasant to use because my gaze will frequently stray out of the sharp area. A sweet spot of 35 degrees or more will never cause any problems, but I've seen sweet spots that wide in only a few binoculars that use so called "field flatteners" in the eyepieces. Somewhere between 25 and 30 degrees seems to be the threshold between comfort and discomfort in the field for me. To that I would add that binoculars in which off-axis astigmatism is the dominant aberration limiting the size of the sweet spot tend to be more unpleasant for me to use than binoculars in which field curvature dominates.

Perhaps a reasonable definition of the "sweet spot" in binoculars would be that area of the field in which the smallest details that can be resolved are limited only by the eyesight acuity of the viewer. Outside the sweet spot the smallest visible details would be limited by the off-axis aberrations of the binocular's optics.]

I’m wondering if, just for practical purposes, we might try to come up with some kind of consensus definition of sweet spot parameters, at least between those of us who are interested in using such a subjective concept and measurement. I think that Henry’s definition in the last quoted paragraph is perhaps a bit too stringent, especially considering that for testing for the area limited only by the viewer’s eyesight acuity we will have the binocular tripod-mounted but for general viewing usually not. To come up with an alternative suggestion, I would have to do more testing, but I could imagine it would be something like considering the sweet spot as the area of the subjective field where resolution has not dropped by more than one element on the USAF chart.

Kimmo

Quote: 'I have to disagree with you. I have no objection to having objective tests, but IMO an experienced observer can reach an assessment of an instrument that often surpasses the results of tests.'

Really, how? How does the subjective assessment of an instrument surpass the accuracy of measured results?

'For example, some instruments behave fine when tested for resolution and contrast, but in certain situations they fall to pieces.'

Can you give me a further example other than the ones you gave?
The two examples you gave - the Zeiss 8X30 flaring is (I presume you are talking about an internal reflection) something that can be seen in other light conditions besides dusk, and can be shown on a test bench. The Swarovski, I would suggest, has the chromatic aberration it has. The fact that some notice it and some not is the very point of this discussion. In one person's opinion it is objectionable, in anothers it is not. This then becomes a subjective assessment, not particularly useful. This makes the case for an objective test.

Any optical problem can be seen and often measured on a test bench, even chromatic aberration. Chromatic aberration can even be seen very easily on something as simple as a Ronchigram. A Koster's Prism interferometer can directly measure it.

'Getting back to the sweet spot, in some instruments such as the Swarovski 8.5x42 EL, the image degrades gracefully away from the optical axis, and is still okay at the edge. In the Nikon 8x42 HG the image is remarkably good to the edge. No doubt in some others the image is almost perfect to half way to the edge, then degrades rapidly. So how do we rate each instrument? Do we favour a slow graceful decline, or do we favour no decline to half way to the edge, followed by a rapid decline? Clearly the choice is subjective. I suppose we could draw graphs showing resolution as a function of angle from the axis, and allow the reader to decide. But would that really give them a good feel for the instrument?'

This paragraph doesn't make much sense. There is nothing subjective about it at all. We rate each instrument's "sweet spot" by measuring it. 'Slow graceful declines' is not a useful description. There will be a point within the field of view beyond which the image is not sharp. This can be demonstrated quite easily on a test bench, and furthermore, what the aberration is. This is not possible in the field.

'Describing what we see' is not good enough. It is OK for conversation between birders, or even chat on a forum, but little else. It is entirely subjective and not helpful when definitive and objective information is called for. 'Alternatively just try the instrument for yourself and ignore other peoples prejudices.' Fine, but isn't that what people do anyway? What we are talking about is whether we can agree on a universally acceptable test which will give an idea of which binoculars have the greatest "sweet spot" and whether this is consistent within each model.

andytyle

Quote: 'As the discussion between S.I. (in post #19) and Alexis, Elk, myself and others shows, current test bench procedures do not seem to be quite universal enough in their applicability to average (or exceptional) users’ perceptions. The test bench procedures used by Zeiss seem perfectly reasonable on the face of it, but when the hard proof they provide is uncorroborated or contradicted by the end users’ experiences, I find I cannot easily dismiss more subjective evaluations in their favor. So, even if a “sweet spot” is not easily quantifiable or is precisely quantifiable only for a particular individual under particular viewing conditions, I still find it a useful concept. In fact, much the same arguments as presented against the SS could be leveled against visually measuring edge-of-field resolution in general. For a user, it does matter how far off-axis and how easily one can look and still perceive a satisfyingly sharp image. If a reviewer or a user discusses these factors, it can (and should) certainly be called opinion and interpretation, but for another user it may nevertheless be highly valuable and sometimes (unfortunately) more valid than a full lab report of hard data when the procedure for obtaining the data turns out to miss (or misrepresent) some important aspects of real human use.'

I agree entirely with the individuality of user perception of an image, but user perception is what it is. This may well be fine for most of us, we choose our instruments based upon how we find the optical performance. Nevertheless, the measured result on a testbench will give a definitive result (coma, astigmatism, LSA and HSA, Secondary spectrum, lateral colour, spherochromatism, distortion, field curvature). If a definitive test by a manufacturer on their own product or another testing is not particularly accurate, or if instruments are not consistent then it is not surprising that sometimes users prefer a subjective assessment. I would have to say that a testbench can reveal any optical issue, whether this issue is important to individual users, or whether the issue is objectionable in different circumstances, is quite another matter.

'Another brief comment concerning your bracketed sentence about problems in higher quality binoculars. Although prisms can sometimes be the main culprit, I find (based on booster-assisted point source evaluation of visible aberrations) that the most common cause for less-than-optimum optical performance is misalignment of elements in the optical train. At least for the top manufacturers and their binoculars, prism manufacturing tolerances seem to be “good enough” not to introduce marked aberrations. With telescopes, the situation is not quite so good (because the magnifications are higher), but with binoculars usually the only prism-induced artifact I see is a very thin diagonal line of light splitting the image of a bright point source if I view the point source in darkness or near darkness and place the point source on the prism roof edge.'

Yes, I should have mentioned that we should at least expect manufacturer's to align instruments properly - Quality Control again. Of course, assembly problems are often the biggest cause of errors, the entire image being comatic. My train of thought was more concerned with fabrication errors and their measurement - the Seidal aberrations (apart from on-axis coma). If we can't get the manufacturer's to up their game with regards to assembly QC, then all other talk of testing is somewhat academic (for accurate measuring).
Having said that, it is still possible to recognize the other Seidal aberrations even when on-axis coma dominates.

andytyle

I should state (should have stated before) that my faith in a testbench giving me objective answers is through experience of how accurate predictions of performance are, based upon measured results. The results of tests on various instruments gives us an idea of exactly what the optical problems are, and thus what can be done about it. It should not necessarily be the deciding factor about which instrument to choose for personal use, there are many personal reasons why we choose. But it is useful to know because it gives us some power as consumers. We are always being told that the customer is always right. Well, here is a chance to see if that saying is right. If a product is sold as being 115mm long, and we measure it to be 113mm, are we wrong?

andytyle

AndyTyle: I do not doubt that measurements can be of use. For example, the amount of CA, the transmission, the centre field resolution, and the amount of scattered light are all interesting quantities, and they are measured by manufacturers. They can give an indication of quality as high transmission usually correlated to an optic of high optical quality. But such measurements are far from complete.

"the Zeiss 8X30 flaring is (I presume you are talking about an internal reflection) something that can be seen in other light conditions besides dusk, and can be shown on a test bench. "

That is most probably true but there are an awful lot of tests that you would have to perform to catch everything. Or alternatively you rely on a small set of tests and risk missing important aspects.

"Any optical problem can be seen and often measured on a test bench, even chromatic aberration. "

That is no doubt true but there are several issues. The first issue is choosing which tests to perform. There are an awful lot that you would have to carry out if you wanted to pick up everything. Unless of course you are simply interested in testing the instrument in one artificial situation on a test bench e.g. examination of a test subject with lighting from the front and to one side. The second issue is how to present and interpret the results. In other words, how do you convert figures into something meaningful.

And regarding flare, you would have to test the instrument with all kinds of lighting. You would for example have to test with light from above to check that there are not internal reflections in the objective cell which leads to flare in the visual field. I believe that Canon recalled the initial batch(es) of 24-105mm F4 L lenses due to internal flare resulting from reflections off an unpainted screw head inside the lens. The flare was only seen in specific lighting situations. It was picked up by users. I think Zeiss had an issue with flare inside the original Victory (not the current one) due to reflections. I recall being in a forest with a non Zeiss instrument, and suddenly seeing massive flare in my binoculars. The cause was a break in the canopy overhead, in an otherwise dark wood.

"This paragraph doesn't make much sense. There is nothing subjective about it at all. We rate each instrument's "sweet spot" by measuring it. 'Slow graceful declines' is not a useful description. There will be a point within the field of view beyond which the image is not sharp. "

Yes it does make sense, there is something subjective about it and 'Slow graceful declines' is a useful description, at least for me. The decline is not necessarily linear. Or graceful. The subjective comes in how to define what is acceptable or preferable.

For example, the Zeiss 8x42 FL binocular has (IMO) a decent sweet spot, but the image gets very soft towards the edge. Some people object strongly to this, and do not like the binocular, considering the sweet spot too narrow. Others such as me are more forgiving. However, the softness is in part due to field curvature. The human eye can of course to some degree correct for field curvature. So, how do you design a test to account for one instrument which has edge softness due purely to field curvature, and another which has edge softness due to field curvature and an aberration such as spherical or astigmatism? Do you assume they are comparable due to the same amount of softness in a test? Or do you favour the instrument whose edge softness is purely due to field curvature, and hence many/most users will be able to adjust their eyes to correct for it?

"'Describing what we see' is not good enough."

It looks like we will have to disagree on this. I would say that a brief description can be extremely informative, despite the subjectivity.

Photography is an area where huge amount of work has been done to assess optics from various manufacturers. Although MTF plots are useful they do not tell you everything, and often do not tell you about the most important aspects of a lens. One of the most respected lens reviewers is B. Rorslett, and the following might be of interest to you:

http://www.naturfotograf.com/lens_surv.html

Note that although he does take test photographs, his lens reviews are essentially short informative descriptions, not masses of data.And they are very useful.

I have read reviews by Kimmo and liked them a lot, as well as other's using USAF charts etc. and would say I have never read any review of binocular or spotting scope using what Andy describes. Telescopes maybe. I would much sooner be able to read what others say as well as test, than nothing at all. Why doesn't Andy post some of his reviews?
Regards,Steve

Quote: 'AndyTyle: I do not doubt that measurements can be of use. For example, the amount of CA, the transmission, the centre field resolution, and the amount of scattered light are all interesting quantities, and they are measured by manufacturers. They can give an indication of quality as high transmission usually correlated to an optic of high optical quality. But such measurements are far from complete.'

Not quite sure what you are saying here. Do you mean that manufacturer's tests of their own equipment are far from complete? Or do you mean that the variety of ways to test an optic still leaves an objective description of that optic far from complete? If its the former, then yes, it may well be true with some manufacturers. If its the latter then I'm sorry but it simply isn't true. ANY aspect of the performance of an optic can be seen on a testbench. Not only that, but better than it is seen in the field, because it can be controlled and quantified. Any stray light or condition where internal reflection or diffuse light can affect the image can be provided. All else simply requires collimated light.


'"the Zeiss 8X30 flaring is (I presume you are talking about an internal reflection) something that can be seen in other light conditions besides dusk, and can be shown on a test bench. "

That is most probably true but there are an awful lot of tests that you would have to perform to catch everything. Or alternatively you rely on a small set of tests and risk missing important aspects.'

No different to testing in any other condition. No important optical aspect is missed. The benefit on a testbench is that there is control, and hence greater accuracy for measuring. If this were not so, there would be no point in scientific bench tests of any kind, in any industry. Subjective assessments are just that, nothing more. Not useful because unlike a scientifically controlled test, a subjective assessment is not repeatable.........because its subjective.

'"Any optical problem can be seen and often measured on a test bench, even chromatic aberration. "

That is no doubt true but there are several issues. The first issue is choosing which tests to perform. There are an awful lot that you would have to carry out if you wanted to pick up everything. Unless of course you are simply interested in testing the instrument in one artificial situation on a test bench e.g. examination of a test subject with lighting from the front and to one side. The second issue is how to present and interpret the results. In other words, how do you convert figures into something meaningful.'

This paragraph is essentially a repeat of the points you made in the previous one.
The conversion of figures into something meaningful happens every time any optic is tested on an interferometer, thats what an interferometer does. It gives values to all the aberrations it tests for and then gives an RMS figure for the optic. The meaning is how the aberrations affect the image. That comes from a knowledge of what the aberrations are and what affect they have. Testbench testing and how the results correspond to performance is all very well known and practised in every professional optical shop virtually every day.


'And regarding flare, you would have to test the instrument with all kinds of lighting. You would for example have to test with light from above to check that there are not internal reflections in the objective cell which leads to flare in the visual field. I believe that Canon recalled the initial batch(es) of 24-105mm F4 L lenses due to internal flare resulting from reflections off an unpainted screw head inside the lens. The flare was only seen in specific lighting situations. It was picked up by users. I think Zeiss had an issue with flare inside the original Victory (not the current one) due to reflections. I recall being in a forest with a non Zeiss instrument, and suddenly seeing massive flare in my binoculars. The cause was a break in the canopy overhead, in an otherwise dark wood.'

Then this goes to show that satisfactory tests were not carried out by the manufacturers.....not that tests don't reveal all the errors you want them to reveal.

'"This paragraph doesn't make much sense. There is nothing subjective about it at all. We rate each instrument's "sweet spot" by measuring it. 'Slow graceful declines' is not a useful description. There will be a point within the field of view beyond which the image is not sharp. "

Yes it does make sense, there is something subjective about it and 'Slow graceful declines' is a useful description, at least for me. The decline is not necessarily linear. Or graceful. The subjective comes in how to define what is acceptable or preferable.'

I find this statement difficult to answer. I'm not sure what you are trying to say. Are you suggesting that we should prefer one type of boundary between what is diffraction limited (or near to diffraction limited) and what is not, and another type of boundary? Whichever, "slow and graceful" is hardly a term which has any meaning in a scientific sense, particularly in context with measuring a geometrical optical field.

'For example, the Zeiss 8x42 FL binocular has (IMO) a decent sweet spot, but the image gets very soft towards the edge. Some people object strongly to this, and do not like the binocular, considering the sweet spot too narrow. Others such as me are more forgiving. However, the softness is in part due to field curvature. The human eye can of course to some degree correct for field curvature.'

Only with extreme accomodation. Field curvature is normally bad enough to need re-focusing.

'So, how do you design a test to account for one instrument which has edge softness due purely to field curvature',

Simple, you measure focus shift from edge to centre.

'and another which has edge softness due to field curvature and an aberration such as spherical or astigmatism?'

One aberration will dominate, all three can be seen or measured using either interferometry, Foucault or Ronchi testing.


'Do you assume they are comparable due to the same amount of softness in a test? Or do you favour the instrument whose edge softness is purely due to field curvature, and hence many/most users will be able to adjust their eyes to correct for it?'

From what I have seen of field curvature in the many binoculars I have used, I would not expect anyone over the age of about twelve to have the necessary accomodation.
By "softness" (another ambiguous term), do you mean defocus or loss of image integrity due to aberration?

'Photography is an area where huge amount of work has been done to assess optics from various manufacturers. Although MTF plots are useful they do not tell you everything, and often do not tell you about the most important aspects of a lens. One of the most respected lens reviewers is B. Rorslett, and the following might be of interest to you:

http://www.naturfotograf.com/lens_surv.html

Note that although he does take test photographs, his lens reviews are essentially short informative descriptions, not masses of data.And they are very useful.'

Not sure what this demonstrates. This appears to be some guy who tests nothing but Nikon lenses. He even suggests in no uncertain terms that his test ratings are 'subjective'. The very thing I suggest is rarely useful when describing the performance and testing of geometrical optics.

andytyle

Quote: 'I have read reviews by Kimmo and liked them a lot, as well as other's using USAF charts etc. and would say I have never read any review of binocular or spotting scope using what Andy describes. Telescopes maybe. I would much sooner be able to read what others say as well as test, than nothing at all. Why doesn't Andy post some of his reviews?
Regards,Steve'
--------------------------------------------------------------------------------

Steve

To answer the first part of your point. In the birding consumer optics market, no-one has ever bothered (as far as I am aware) to use a proper test bench for assessing an instrument. You need to understand how to do it, and apart from professional opticians, I don't know anyone else who does. Even amongst all optical companies there are only a small handful of people who are the optical designers and testers, and even here, in a company that mass produces instruments, these people have very narrow areas of expertise. Very, very few people have the interest and experience to understand many areas of geometrical optics, and then you have to be very fortunate for those people to be also interested in subjects like astronomy and birding.......hobbies that require good quality optics.

The second part of your question. I have never been asked to review binoculars or spotting scopes using test bench methods. It can be done, but so far, testing has been accomplished using test charts. This is useful for comparing instruments (providing there are strict environmental conditions, and a level playing field), but what it doesn't do is highlight which particular aberrations are the most destructive.
I have carried out interferometry and testbench tests like this before on a variety of instruments, but am not at liberty to post the articles, as they legally belong to a company.

This topic has been covered a few weeks ago on another thread, and it might be that during 2008, a few members of the forum begin testing using non-conventional methods. We will have to see what occurs.

andytyle

The conversion of figures into something meaningful happens every time any optic is tested on an interferometer, thats what an interferometer does. It gives values to all the aberrations it tests for and then gives an RMS figure for the optic. The meaning is how the aberrations affect the image. That comes from a knowledge of what the aberrations are and what affect they have. Testbench testing and how the results correspond to performance is all very well known and practised in every professional optical shop virtually every day.

Click to expand...

Andy,

I think you were responding to Leif in this quote, but the same basic theme comes through several posts (which I seriously enjoy reading). The theme seems to be that interferometer analysis potentially (if not actually) says everything one needs to know about the optics. Apparently, it can determine all the third-order aberrations (or higher), assembly errors, glare resistance, and so forth — to a gnat's eyelash or better.

A moment's reflection will show, however, that such results must constitute a multi-dimensional description of the optic under evaluation, the dimensions being the separate aberrations, assembly errors, environmental conditions, etc. We also know, however, that: (a) many of the dimensions are not physically/optically independent of one other, and (b) decisions were made by some designer or design team to stress or suppress certain dimensions in favor of others. We generally refer to these as "tradeoffs." Third order Seidel aberrations, for instance, can not all be eliminated simultaneously, so the question switches to the balance of aberrations as opposed to their presence.

It's generally understood that the design process involves contending with these tradeoff decisions, and the most important factors entering into the process ultimately involve the user population. The most salient of these is probably the acuity of the eye, but, even this must be put into context. Is it day or night, pod mounted or hand held, and exactly what is the observer expected to be doing, e.g., finding enemy troops on the battlefield or birds in the foliage? Each task condition defines its own optimum optics, and these optima (plural), in turn, relate back to the multi-dimensional response surface that is measurable with an interferometer, or related equipment, on a stark lab table.

What can't be determined on an optical bench, are these performance optima. They must be provided from the outside, specifically by the people who will use the equipment for various tasks or task combinations. But how will they do that? The easiest way is by expressing subjective opinions about how well a given optical design suits their needs. Function trumps form. These opinions can be validated (objectified) through expensive behavioral studies, but not on an equally expensive optical bench.

So, as I see it, the reason why subjective evaluations and descriptions are essential is that users who perform similar tasks, such as particular types of birding or astronomy, or hunting, can communicate efficiently without first understanding the myriad of pscho-physical functions relating the physics of their optics to the desired performance characteristics (that is, if anyone knew them). Steve Ingraham coined the phrase a "Better View Desired," I'm inclined to think, to emphasize that the ultimate criterion is the human one, i.e., the better "view," not the better optic.

I'm sure you realize these comments are only intended to enliven the discussion.

Blue skies, and HNY,

Elk

Elkcub: Well expressed.

As you indicate a binocular is a compromise, in that there will be aberrations and faults, and the preferred balance between them is subjective.

AndyTyle: "Or do you mean that the variety of ways to test an optic still leaves an objective description of that optic far from complete? If its the former, then yes, it may well be true with some manufacturers. If its the latter then I'm sorry but it simply isn't true. "

The number of degrees of freedom is too large to carry out complete and sufficient tests. In practice you do a sample of tests, to gain a reasonable assessment. And that choice of tests is subjective, based on experience and knowledge. And of course the interpretation of the results is subjective as someone has to decide what is acceptable and what isn't, given that the instrument is to be used by a human, and not just to carry out some well defined and routine task, such as creating an image of a mask on a silicon chip.

AndyTyle: "From what I have seen of field curvature in the many binoculars I have used, I would not expect anyone over the age of about twelve to have the necessary accomodation."

I am middle aged, and I can re-focus my eyes to counteract some of the field curvature present in various binoculars I have owned. I can also twiddle the focus knob and correct for the softness when it is due to field curvature. Not so for other aberrations.


AndyTyle: "Not sure what this demonstrates. This appears to be some guy who tests nothing but Nikon lenses. He even suggests in no uncertain terms that his test ratings are 'subjective'. The very thing I suggest is rarely useful when describing the performance and testing of geometrical optics."

Yes it is 'some guy' who tests mainly Nikon lenses. The point I was making is that subjective tests are useful, and Rorslett is an example of an experienced and respected photographer who presents subjective tests which are widely quoted by amateur and professional photographers alike. I find that his subjective tests are usually far more accurate than supposedly objective tests such as those on www.PhotoZone.de. The problem with many/all objective tests is that they give an incomplete picture.


I guess you propose performing a test of an optic not only with an 'infinite' target, but with an intermediate and near ones too? And then performing tests of the amount of scattered light (flare) under goodness knows how many different illuminations? It would require huge amounts of work, would take ages to do, and it would generate massive amounts of data. And then you would have to make subjective decisions when presenting the data e.g. how much CA is acceptable, how much field curvature is acceptable etc. Hence you are back to the fuzzy subjective domain rather than the cold precise and rational objective one.

Alternatively why not just pick up the instrument, and spend a few months using it, then tell us all what you think of it? Sure you might have older eyes than the average, and your acuity might be a bit low, and you might only use it in daylight, but your experience when combined with others would be valid.

Incidentally posts here on BF indicate that sample variation is a significant issue for both spotting scopes and binoculars. That objective test tells you about one sample of an instrument. A range of 'horribly subjective and worthless' user opinions gives an idea of how multiple samples perform.