Small focal length, small exit pupil, large AFOV

[25585]

As focal lengths get shorter, field stops and exit pupils smaller, do wider angle views dilute the amount of light being spread over a larger field of view, at the cost of brightness and contrast, that disregarding number of lenses, coatings etc.

Is there an ideal, or best compromise, for FL, field stop and AFOV?

[notFritzArgelander]

As the number of different performance parameters you want to optimize increases so does the number of lenses in the design. You might consider aspheric surfaces. But one needs to express required performance parameters quantitatively to begin to answer your question. That’s the reason why the designs that aim in the direction of your question are so numerous and also have so many lenses.

If you value crisp high definition fields with maximum light throughput fewer lenses are more likely.

[Don Pensack]

As focal lengths get shorter, field stops and exit pupils smaller, do wider angle views dilute the amount of light being spread over a larger field of view, at the cost of brightness and contrast, that disregarding number of lenses, coatings etc.

Is there an ideal, or best compromise, for FL, field stop and AFOV?

No, if the focal length is the same, the brightness per unit area is the same.
A narrow and wide eyepiece go through the same exit pupil size but illuminate different areas on the retina.
So the wider eyepiece is not brighter or dimmer at the same exit pupil.

Technically, the wider eyepiece admits more photons into the eye, but the brightness per unit area on the retina is identical to a narrower eyepiece.

We don't see the total photons, we see unit area brightness, so a 110° eyepiece will be equal brightness to a 50° eyepiece.

In a non-tracking scope, there is a practical side to having the apparent fields of your eyepieces get larger as the focal length shrinks--the object can spend more time in the field of view as it drifts across the field.
In my dob, I go from 78° at 14mm/12.5mm, to 85° at 11mm, to 100° at 8mm/7mm/6mm, to 110° at 4.7mm/3.7mm, and that means I push the scope less than the magnifications would imply as the power goes up.

[turboscrew]

As focal lengths get shorter, field stops and exit pupils smaller, do wider angle views dilute the amount of light being spread over a larger field of view, at the cost of brightness and contrast, that disregarding number of lenses, coatings etc.

Is there an ideal, or best compromise, for FL, field stop and AFOV?

No, if the focal length is the same, the brightness per unit area is the same.
A narrow and wide eyepiece go through the same exit pupil size but illuminate different areas on the retina.
So the wider eyepiece is not brighter or dimmer at the same exit pupil.

Technically, the wider eyepiece admits more photons into the eye, but the brightness per unit area on the retina is identical to a narrower eyepiece.

We don't see the total photons, we see unit area brightness, so a 110° eyepiece will be equal brightness to a 50° eyepiece.

In a non-tracking scope, there is a practical side to having the apparent fields of your eyepieces get larger as the focal length shrinks--the object can spend more time in the field of view as it drifts across the field.
In my dob, I go from 78° at 14mm/12.5mm, to 85° at 11mm, to 100° at 8mm/7mm/6mm, to 110° at 4.7mm/3.7mm, and that means I push the scope less than the magnifications would imply as the power goes up.

Would it be right to say, that wider view eyepiece passes more light through the eye, but the light is spread on a larger area in the retina: double TFOW gives double-squared light in the eye, that is spread on double the area on the retina?
Light density (brightness) is the same.

[notFritzArgelander]

As focal lengths get shorter, field stops and exit pupils smaller, do wider angle views dilute the amount of light being spread over a larger field of view, at the cost of brightness and contrast, that disregarding number of lenses, coatings etc.

Is there an ideal, or best compromise, for FL, field stop and AFOV?

No, if the focal length is the same, the brightness per unit area is the same.
A narrow and wide eyepiece go through the same exit pupil size but illuminate different areas on the retina.
So the wider eyepiece is not brighter or dimmer at the same exit pupil.

Technically, the wider eyepiece admits more photons into the eye, but the brightness per unit area on the retina is identical to a narrower eyepiece.
….

Only if you neglect absorption losses due to more lenses and a longer optical path in glass, which absorbs a little bit.

[Don Pensack]

Only if you neglect absorption losses due to more lenses and a longer optical path in glass, which absorbs a little bit.

Well, yeah, but light loss is approximately 1% per INCH of glass, so it would be a pretty big eyepiece that lost more than, say, 3% from the small simple glass eyepiece.
And that is only about 0.03 magnitudes, which is less than the darkness fluctuation that happens on a half hour basis at most sites.
And such a difference isn't even important for a limit observation of the faintest targets.
Illumination on the retina isn't really determined in any way by the number of elements or thickness thereof as much as it is by reflection losses from the air-to-glass surfaces.
To wit: if the coatings are equal, a Plössl with 4 air-to-glass surfaces could transmit as much as 98.0%, ignoring glass loss.
A 9 element eyepiece with 6 groups might transmit 94.2% and lose, let's argue, another 3% in the glass, for a 91.2% transmission (that's smaller than most complex eyepieces actually transmit).
The difference in light transmission is still way less than a tenth of a magnitude, which isn't going to be noticeable in the field.

The differences that might be most noticeable relate to the surface polish on the lenses, as related to light scatter in the eyepiece. There, the multi-element eyepiece is at more of a disadvantage.
Fortunately for us, most of those eyepieces are likely to be higher end and have superior polish on the lenses (though not all do).

But, to all intents and purposes, the lens count of the eyepiece is not the important factor determining illumination at the retina in a given scope--exit pupil is.

[notFritzArgelander]

I know from direct comparison between Delos and Abbe Orthoscopics that the Orthoscopics beat Delos on transmission and contrast on faint fuzzy objects and shadow transits of Jovian moons. The difference is small but when I do the direct comparison it’s certainly noticeable. Path length through glass matters. Small difference is not zero difference especially at the limit of detection.

[Don Pensack]

And how did you measure transmission?
The brightness? You realize that the human eye in lab conditions cannot reliably see with more than random chance a difference of 10% in brightness.
The two eyepieces in question are only a few % apart in lab transmission, so the perceived difference is highly unlikely to be actual transmission.
Was something brighter in one than the other?
May I presume the focal lengths were exactly equal? If not, a brightness level could easily be in the exit pupil.
Not all orthoscopics are equal. Was it a high-end brand like Zeiss? Or the Fujiyamas?
How large was the scope used in the comparison?
Could you see a super faint star or galaxy in one you could not see in the other? You mention faint fuzzy objects. If talking planets, you aren't going to see a difference in transmission. Shadow transits are a matter of spot size, not transmission.
If talking faint moons, a difference in sharpness (spot size) could dictate the visibility of faint moons.

I'm not really being argumentative here just for the sake of argument, but I've seen lab measurements on a lot of eyepieces, and though, in general, small low element count eyepieces do have a slightly higher transmission, it's not a % large
enough to be seen with the eye unless one is significantly lower. I had one with a noticeably lower transmission and it measured only 81% transmission at 532nm. That is exceptionally low--it was a Meade Series 4000 14mm UWA.
Several internal surfaces were uncoated. But the average Delos is up in the 95% range and that is not enough difference to say they are lower in transmission.

That your impression is of lower transmission, I don't doubt. I would just like a better definition of what you mean by that. If it's simply, "Well, the stars appeared a bit brighter in one than the other.", that could have a lot of other reasons for the impression.
Like the spectrum of transmission versus the stars' or the sharpness of the star images, or even the focus in the scope.
But if there was a difference in exit pupil, then the difference is probably simply that.

[notFritzArgelander]

Mostly the comparisons that I recall were in my Z12 and VMC 110 L. I posted details here and at the former site.

The comparison on shadow transits of Jupiter were done with the Z12, using 6mm Delos, 6mm KK Fujiyama, and 6mm Baader Classic Ortho. The planetary detail was very even across the board with one noticeable exception, the shadow spot was defined better and darker in the orthos than in the Delos. I attribute this to better scattered light control.

Another comparison was done on the Merope nebula. This involved the TV Panoptic 19mm, TV Ploessl 20mm, Vixen NPL 20mm, BCO 18mm. The standouts were the Ploessl and the BCO. The detectable extent of the nebula in the direction of nearby stars was greater in the Ploessl. The nebula was darker in the BCO but had more well defined and richer internal details and somewhat greater extent detectable than the Panoptic and NPL which were dead even in both nebular extent and internal detail. I later came back to the same subject using the Panoptic as the control the 17.3mm Delos and the 18mm KK. The Delos had less nebular extent than the Pan. The KK had more. The KK also had more internal detail visible but not quite as much as much as the BCO.

[WilliamPaolini]

I know from direct comparison between Delos and Abbe Orthoscopics that the Orthoscopics beat Delos on transmission and contrast on faint fuzzy objects and shadow transits of Jovian moons. The difference is small but when I do the direct comparison it’s certainly noticeable. Path length through glass matters. Small difference is not zero difference especially at the limit of detection.

A very common experience for me as well when critically comparing minimum glass to maximum glass eyepieces. However, to dissuade the optical police I usually characterize the differences as apparent-transmission and apparent-contrast (including apparent-color saturation ), so whatever the drivers are of the difference visually is appears as a slight boost in brightness and more than slight increase in contrast to the view. Most notably seen on planetary observing, faintest portions of nebula, and ability to see detect faintest stars in clusters.

[Don Pensack]

Lacking lab reports, your subjective reports are about all we have.

Everything Bill reports could be attributed to superior contrast, and that is where the small, low lens count, eyepieces shine.
To extrapolate, a nebula filter makes a nebula appear larger and fainter parts become visible.
Transmission is actually lower with the filter in that case, yet the nebula is more visible. It's contrast, not transmission.

I saw this recently in a comparison between a 14mm Morpheus and a 14mm Orion LHD, in which a faint tidal tail on NGC3628 was visible in one with direct vision and only partially in the Orion with averted vision.
There was a lot of scattered light, and even EOFB in the Orion, but none in the Morpheus. Both eyepieces have the same number of groups and internal lenses. The difference was contrast.
And that certainly does make fainter details and objects visible.

[notFritzArgelander]

Lacking lab reports, your subjective reports are about all we have.

No that's not all we have. We also have physical optics. Coatings and other factors being equal, a longer path through glass implies greater absorption losses with resulting transmission loss and compression of detectible contrast ranges.

Everything Bill reports could be attributed to superior contrast, and that is where the small, low lens count, eyepieces shine.
To extrapolate, a nebula filter makes a nebula appear larger and fainter parts become visible.
Transmission is actually lower with the filter in that case, yet the nebula is more visible. It's contrast, not transmission.

Contrast and transmission are not so easily separated when the detector is the human eye. A little better transmission can make quite a difference in detecting contrast differences. We know for instance that in discerning contrast differences that the size of the region to the eye is quite significant. See for example https://clarkvision.com/visastro/index.html and references therein. So simply stating that a certain magnitude difference is all that matters is oversimplifying.

I saw this recently in a comparison between a 14mm Morpheus and a 14mm Orion LHD, in which a faint tidal tail on NGC3628 was visible in one with direct vision and only partially in the Orion with averted vision.
There was a lot of scattered light, and even EOFB in the Orion, but none in the Morpheus. Both eyepieces have the same number of groups and internal lenses. The difference was contrast.
And that certainly does make fainter details and objects visible.

All of which is irrelevant to my original point as not all other factors were equal. Different coatings, internal stops, etc. are all contributors. The fact remains that the physics of the optics means that a longer path in glass means a transmission loss and that will in turn compress the detectable contrast.

[j.gardavsky]

With the eyepieces (and other optics), the main losses of light are due to:
1. Absorbtion on the color centers, like the metal ions (from the metal oxides) in the glass matrix
2. Scattering on the microinclusions
3. Scattering on the bond materials (cements)
4. Reflection on the air/glass and glass/air surfaces
5. Diffuse scattering on the air/glass and glass/air surfaces
6. Other isses due to the striata, air bubbles, ...

ad1.: This cuts as a rule the UV and IR, and dims the blue/violet.
ad 1. and 2.: The absorption and scattering ad 1. and 2. are not linear with the glass thickness, definitely nothing like 1% per inch. The loss grows exponentially with the optical path through the glass.
ad 3. The scattering on the bond materials is suppressed in the modern designs, especially on the premium grade eyepieces like DOCTER, microscopy eyepieces, etc.
ad 4. The reflection on the interfaces is suppressed with the multicoatings for the on axis (paraxial) and for other normal incidence light rays. The multicoatings are less effective on the marginal light rays.
ad 5. The diffuse scattering happens partially already in the multicoatings stacks, then on the lens surface microroughness. How good suppressed, this depends both on the grinding standard (SCHOTT "K"-number), and on the final polish, and on the multicoatings design.
The HD (Leica, Zeiss), and maybe the HT (Baader) multicoatings seem to be the next step to reduce this loss of light.

The light losses 1. through 5. increase for the shorter wavelengths, and the losses are already pretty high at OIII, H-Beta.

The loss of light also means a loss of contrast of the view:

The preference towards the smart designs with minimum glass comes indirectly from the fluorescence microscopy, where the fluorescence is extremally faint. Nobody would come up with the idea to mount the wide and ultrawide eyepieces for these applications, even if the wide microscopy eyepieces have been designed and manufactured.

This is also the reason, why I take for the most difficult nebulae the eyepieces with a minimum of glass inside, and preferably eyepieces selected from the premium and research grade microscopes, like the Zeiss Axioskop and the Leica HC system,

Best,
JG

[notFritzArgelander]

Thank you @j.gardavsky for providing in detail the correct physics involved. In particular the key point you mentioned

ad 1. and 2.: The absorption and scattering ad 1. and 2. are not linear with the glass thickness, definitely nothing like 1% per inch. The loss grows exponentially with the optical path through the glass.

Just to give an example of what this means, if you double the geometric glass path length, the losses (due to processes 1. and 2. in your post) quadruple. The exponential increase in losses due to a thicker optical depth is a direct consequence of the equation of radiative transfer. Geometric optics and ray tracing are inadequate for energy transfer.

I used to make my living out of this book: https://www.amazon.com/Radiative-Transf ... 0486605906. So I'm thankful for your pointing this out and also a tad embarrassed that I did not hit on the exponential behavior of losses as the conceptual problem here.

Anyway the Wikipedia article on radiative transfer is too technical to post here but the concept of "optical depth" aka "optical thickness" can be read by folks who understand an exponential function. As a bonus there are images at the end of optical thickness of a martian dust storm.

https://en.wikipedia.org/wiki/Optical_depth

[j.gardavsky]

Hello not_Fritz,

I have been dead sure that you will join, when it is about the absorption, radiative transfer, etc.

I have spent my young years with the light absorption measurements, later with the theory, measurements, and applications of the scattering of light, before turning to the Fourier optics because of my loving attention towards the Fourier Theorems.
It is somehow the mathematics behind the physics, which is so much attractive.

Best,
JG

[Don Pensack]

All well and good from the standpoint of theory.
And yet, these thick, multi-element, eyepieces often eke out a 95% transmission at 550nm, compared to 98% for an ortho or Plössl.
So, to put it simply, light TRANSMISSION is a red herring in this argument. It is a contrast difference that is seen in the field.
Because it has been verified many times the human eye cannot see such a small difference in brightness,
but our eyes are superb contrast detectors.

In JG's list, every item on the list directly relates to a light scatter result, which robs contrast from the image.
In essence, a brightening of the background around the object or star and a blurring of detail.

I'm not arguing there is no difference in eyepieces here, merely that the differences we see (we do not see in the UV or IR) are due to contrast differences and not transmission.
Scotopic vision is almost entirely from 450nm to 550nm, with the ends of sensitivity at about 425nm and 600nm. Some individuals can see about 20-30nm deeper into the red.
What happens at shorter and longer wavelengths is not germane to the discussion. Plus, a 10% difference in transmission is still only 0.1 magnitude, which is not usually detectable by the eye outside of a lab.
The AAVSO says that 0.1 magnitude is about where the readings by observers pass to completely random chance.
And even the readings of spectrophotometers have to compensate for electrical and thermal noise in the instruments to get smaller readings (after compensation, a lot smaller, as in +/- 0.01 magnitude).

Ultimately, a transmission difference would have to be considered one factor in determining differences in eyepieces, but it is way way down the list of factors that have any importance in what we see through the eyepiece in a typical scope.
If you compare as many eyepieces as I have, you'd realize the visible differences between eyepieces has very little to do with the number of elements and a lot more to do with suppression of light scatter and the
adherence to a standard of correction in the eyepiece, which includes lens polish standards. If all things are completely equal, that gives the advantage to the small, low lens count, eyepiece, but things are never equal.

[notFritzArgelander]

I think I am seeing where we are talking past each other and perhaps this response will clarify.

All well and good from the standpoint of theory.

Theory is not to be disparaged. It enables one to correctly understand experimental and practical results.

And yet, these thick, multi-element, eyepieces often eke out a 95% transmission at 550nm, compared to 98% for an ortho or Plössl.
So, to put it simply, light TRANSMISSION is a red herring in this argument. It is a contrast difference that is seen in the field.
Because it has been verified many times the human eye cannot see such a small difference in brightness,
but our eyes are superb contrast detectors.

I think that the problem is that you are confusing two different meanings of transmission and this is misleading you.

One is an end-to-end test sense of transmission. You put in light at one end and see how much of it comes out the other. This is the sense you seem to be using. Let's denote that as test-stand-transmission. It has nothing to do with image quality and contrast.

The other meaning, which JG and I are using is the physics grounded sense of transmission that describes the energy flow through the system and determines image quality. It includes both absorption and scattering processes. Let's denote that as radiative-transmission. If someone is stuck on test-stand-transmission one will not understand the physics of image formation that is embodied in radiative-transmission, one will be misled by a total percentage throughput and ignore the image formation.

I think that a lot of the talking at cross purposes here is a result of the language of folks in the eyepiece biz not understanding the language of folks from the physical optics biz.

In JG's list, every item on the list directly relates to a light scatter result, which robs contrast from the image.
In essence, a brightening of the background around the object or star and a blurring of detail.

It is not true that "every item in JG's list directly relates to a light scatter result". Take the first item for example: "1. Absorbtion on the color centers, like the metal ions (from the metal oxides) in the glass matrix." That is NOT scattering. It is loss pure and simple.

I'm not arguing there is no difference in eyepieces here, merely that the differences we see (we do not see in the UV or IR) are due to contrast differences and not transmission.
Scotopic vision is almost entirely from 450nm to 550nm, with the ends of sensitivity at about 425nm and 600nm. Some individuals can see about 20-30nm deeper into the red.
What happens at shorter and longer wavelengths is not germane to the discussion. Plus, a 10% difference in transmission is still only 0.1 magnitude, which is not usually detectable by the eye outside of a lab.
The AAVSO says that 0.1 magnitude is about where the readings by observers pass to completely random chance.
And even the readings of spectrophotometers have to compensate for electrical and thermal noise in the instruments to get smaller readings (after compensation, a lot smaller, as in +/- 0.01 magnitude).

All of which reinforces the impression that test-stand-transmission is the idea you are stuck on. It's also not germane to the radiative-transmission ideas that are physics based and determine image quality including contrast.

Ultimately, a transmission difference would have to be considered one factor in determining differences in eyepieces, but it is way way down the list of factors that have any importance in what we see through the eyepiece in a typical scope.
If you compare as many eyepieces as I have, you'd realize the visible differences between eyepieces has very little to do with the number of elements and a lot more to do with suppression of light scatter and the
adherence to a standard of correction in the eyepiece, which includes lens polish standards. If all things are completely equal, that gives the advantage to the small, low lens count, eyepiece, but things are never equal.

Let me reword parts of this in my terms:

Ultimately, a test-stand-transmission difference would have to be considered one factor in determining differences in eyepieces, but it is way way down the list of factors that have any importance in what we see through the eyepiece in a typical scope.

You'd realize the visible differences between eyepieces has very little to do with the number of elements and a lot more to do with radiative-transmission and the adherence to a standard of correction in the eyepiece, which includes lens polish standards. If all things are completely equal, that gives the advantage to the small, low lens count, eyepiece, but things are never equal.

Radiative-transmission encompasses the scattering of light. Test-stand-transmission doesn't. That's the nub of the disagreement.

[j.gardavsky]

Thank you not_Fritz for the in-depth comments, and further explanations!

What finally happens, is the drop of the modulation transfer function MTF, and specifically on the high spatial frequencies, so important for a comfortable resolution of the details on the planets.

Best,
JG

[notFritzArgelander]

Thank you not_Fritz for the in-depth comments, and further explanations!

What finally happens, is the drop of the modulation transfer function MTF, and specifically on the high spatial frequencies, so important for a comfortable resolution of the details on the planets.

Best,
JG

Exactly so! The MTF also the related point spread function (PSF) are the best, most comprehensive quantitative means of describing and optimizing optical design and execution. They are faithful to the physics involved and take into account the interplay between transmission ( radiative i.e. Including scattering ) and contrast rendering that the test-stand transmission concept cannot handle.

[Bigzmey]

Great discussion! While I am a great fun of well made minimal glass EPs I end up having a few short FL widefield EPs. Notably, XW5 and XW3.5. After so many rounds of comparison on variety of targets I came to the conclusion that in 95% cases a well made widefield is as good as a good minimal glass EP, but provides better ergonomics. On occasion there is an extra faint galaxy or a tight double where Ortho or Plossl would push it slightly to a visible realm. So, most of my sessions are carried with modern complex EP designs (mainly for the comfort of observing) but I always keep a set of Plossls and Orthos in my EP case, and they save the day (or night) on occasion.