New 75mm Hypergon Lens

[Dan Fromm]

I also doubt it would work with a 135° lens that is not retrofocus nor is made according to the Roosinov's aberrational vignetting principle (like all the Biogons, Grandagons, etc., etc.). A spherically curved filter may probably be a solution.... but certainly not a cheap one.

Please explain further. I ask because Rodenstock's Pantogonal, a 3 element lens and neither retrofocus nor Russar type, was offered with a flat center filter.

"This consists of a plano-convex lens of yellowish green glass which absorbs very strongly blue and violet light, and a plano-concave lens of a colorless, very transparent material. The optical constants of both these lenses is exactly equal so that the cemented lenses act as a plane parallel plate which is dark in the center and transparent at the sides."

[Mark J]

I also doubt it would work with a 135° lens that is not retrofocus nor is made according to the Roosinov's aberrational vignetting principle (like all the Biogons, Grandagons, etc., etc.). A spherically curved filter may probably be a solution.... but certainly not a cheap one.



Between the lens elements, the filter just wouldn't work as a center filter.

What I would like is to have an option to put yellow/green/orange/red gel filters inside the lens. It would be good for both modifying the image tones and cutting the influence of the chromatic aberration down.

I don't know what is meant by the 'aberrational vignetting' principle. Is it in his original paper ?
The problem with any flat filter used with the Hypergon is that for a ±67.5° field angle, the surface losses from a filter, either used outside or internally, become quite significant at the edge of the FOV and add to the fall-off. Even with AR coatings on the filter, most AR coatings are only better than bare glass out to about 60° incidence.
However, for a centre-filter ( US : center-filter ) , the effect of the internal gradation of the filter will still outweigh these surface-loss problems, so this is worthwhile.
For a coloured filter in the area of the stop, bear in mind that there will still be increasing surface-reflection losses beyond 60° , and that the light path through the filter is also longer for the edge-bundles, so the absorption will be a bit greater than in the centre, though this shouldn't be enough of a barrier to it working OK ... if you can compensate for the filter thickness affecting the effective air-gap.... Gels will probably be Ok though.

[ridax]

I don't know what is meant by the 'aberrational vignetting' principle. Is it in his original paper?

Yes that's the term Roosinov used to call his invention - both in his original papers and in his tutorials on optics (which I own and read from time to time but alas I'm not good enough in mathematics to get it all in). He chose that name because similarly to the common geometric vignetting that changes the lens pupil shape at the field edges, the aberrations he implied also changed the pupil shape off center. But while the geometric vignetting always make the pupil smaller causing light fall-off, the aberrational vignetting is able to increase the pupil size thus diminishing the light fall-off.

for a centre-filter ( US : center-filter ) <...> in the area of the stop, bear in mind that there will still be increasing surface-reflection losses beyond 60° , and that the light path through the filter is also longer for the edge-bundles, so the absorption will be a bit greater than in the centre, though this shouldn't be enough of a barrier to it working OK ...

Here is the most descriptive ray tracing diagram for the Hypergon that I found in my quick internet search:-


It clearly shows that at the place of the aperture, ALL the rays go through the same small circle in the center, and there is NO way to stop the paraxial rays while letting the oblique rays through. And that's true for any lens of any type.

[ridax]

Please explain further. I ask because Rodenstock's Pantogonal, a 3 element lens and neither retrofocus nor Russar type, was offered with a flat center filter.

"This consists of a plano-convex lens of yellowish green glass which absorbs very strongly blue and violet light, and a plano-concave lens of a colorless, very transparent material. The optical constants of both these lenses is exactly equal so that the cemented lenses act as a plane parallel plate which is dark in the center and transparent at the sides."

Please bear in mind I don't positively state it would not work; I just say I have serious doubts. Those are:-

(1) The surface reflection problem addressed by Mark J above (and that's one of the reasons to suggest a hemispherical filter instead of a plane one),

and

(2) the fact that a plane-parallel glass plate introduces its own aberrations. The thicker the plate, the more pronounced the aberrations are. And the greater the angle, the more pronounced the aberrations are, too. (An ultra-thin pellicle filter would probably work fine. But it would probably cost several times more then the Hypergon itself and be way more fragile than the original Hypergon star/fan.)

The Pantogonal center filter description suggests a pretty thick glass plate in front of the lens; that makes me suppose the Pantogonal was designed from the start to be the most sharp with the filter in place. The Hypergon was certainly not. Perhaps it is possible to modify the Hypergon to be good with the filter but that means lots of new calculations.

A prominent example of the plane-parallell glass plate influence are aireal lenses. Virtually all of them are designed to work with a thick glass plate just in front of the film plane, and perform more or less poorly without it. I've recently tested a bunch of the famous 125mm f/4 super-apochromatic CZJ Pinatar lenses that are often said to be the sharpest glass in the world. Those are for the 6x9cm format and really far from beeing wide-angle but without their 8mm-thick plane-parallel focal plane glass plates, they are really poor performers not worth their own weight....

[reddesert]

Roosinov's patent is US2516724A, https://patents.google.com/patent/US...n?oq=us2516724 You have to download the PDF as the OCR'ed text is very poor.

He uses "aberrational vignetting" to refer to the property that decreases the light falloff from cos^4 to cos^3. Today, it is often referred to as a tilting of the exit pupil; Roosinov's diagrams Figs 3-7 and 8-12 illustrate how the pupil shape is less cat-eyed off axis, but he doesn't use the wording of the exit pupil. I am not sure where he originated the phrase aberrational vignetting, but I think he's deliberately relaxing/violating one of the normal lens design assumptions; he speaks of maximum divergence from Abbe's sine condition for the half-lens as seen from the aperture.

The aperture stop in an imaging lens is more or less by definition the "waist" where all the ray bundles from both on-axis and off-axis sources cross. It's conjugate to the image plane (maximally out of focus) and so you can't introduce a center filter there. (Basically the same reason that you can stop down an ordinary lens without cutting off the corners of the field.)

However, there is a big difference between introducing a flat filter in front of the lens and behind the lens. A flat filter (plane parallel plate) will cause a focus shift, and for off-axis sources, introduce some astigmatism because the effect of refraction in the plate on the sagittal and tangential directions is different. In front of the lens, in pictorial use the object is typically very far away (many focal lengths) and the beam from an on- or off-axis source is nearly collimated. This makes the focus shift negligible and means the astigmatism has little effect, although it might still have some effect at the extreme field angles of a Hypergon, someone would have to try it to see.

Behind the lens, the lens to film distance is quite short and the beam for any source, on or off-axis, is strongly converging. This makes the focus shift significant and likely increase the effect of the off-axis astigmatism. That's why for normal, non-Hypergon lenses, we put glass filters in front of them without refocusing or off-axis image degradation, but if a lens is designed to have a rear filter, then it needs that rear filter or it won't perform right. (Common SLR examples include some mirror telephotos and fisheye lenses.) So I don't know if a front filter would yield acceptable results with a Hypergon-type lens, but it is a different case from a rear filter.

[ridax]

...So I don't know if a front filter would yield acceptable results with a Hypergon-type lens, but it is a different case from a rear filter.

Yes indeed it is different.
And I don't know the results of using a center filter in front of a Hypergon either....

Thank you for your thorough explanation.

[John Layton]

Totally naive question: what about treating the optic itself with a variable density coating?

[Mark J]

Thank you Ridax for the explanation of 'aberrational vignetting'.
I think he could have used a better description, really - after all, it's 'aberrational de-vignetting' ! He deserves to be better-known, for the fine work he did on wide-angle lens design.
It is little known, but there are sets of aberrations (like Seidels) for the appearance and position of pupils, in optics, as distinct from aberrations of the image spots.
The effect you see in wide-angle lenses is largely caused by pupil coma. It can easily be seen in wide-angle optics, looking at the front and tilting the lens around. You can observe the pupil expanding and turning towards you, and the expansion may be greater in the meridional section. This is the effect that enables cos^3 falloff or cos^2 fall-off in modern wide angle optics.
Unfortunately the Hypergon does not display this, and we are stuck with cos^4 ... !

John, it's a good question, it can be done and we have done it sometimes over the years, at work.
it does tend to be expensive though, it requires special masking and 'culotte' rotation in the vacuum-coating chamber, to do a set of lenses graded like this, and the 'plant fill' is much reduced, leading to a higher cost.
You also have to be careful if using a reflecting attenuation ( eg. a very thin layer of aluminium ) , it can cause ghosting problems.

[Mark J]

Thanks for the Roosinov patent link.

It's conjugate to the image plane (maximally out of focus) and so you can't introduce a center filter there.

Well.. it's not conjugate, and that's the problem !
The stop is the pupil plane, and there is no separation of the on and off-axis pupils there.
But clearly you understand this and the advice is correct.

I would definitely recommend a centre filter in front of the lens, but it needs to be big enough to cover all the field rays - ie. a lot bigger than the front element. They are not cheap, although some pop up on EBay.
It will reduce the vignetting, but not fully cure it.

[ridax]

Totally naive question: what about treating the optic itself with a variable density coating?

The idea came to me, too. But I am pretty sure such a cener-filter that can't be removed would make the ground glass too dim to see any image - to say nothing about focusing.