What do you call lenses that can focus like this?

[Frank Petronio]

I'm optically ignorant and have never studied any of this (no silly photo school for know-it-alls), so I am not going to be able to use the correct terminology. But I think most lenses must focus in an arc, everything being in focus at the same radius distance out from the center of the lens. Right? Example in figure 1.

But has anyone ever made a lens that focuses on a flat field, like figure 2?

I think the practical application would be if you were using very wide lenses or close up lenses wide open, where you might lose focus on the edges of some large flat plane, like a tower, shot vertically? Or crowd scenes, horizontally.

[rdenney]

Actually, most lenses are corrected for a flat field. Few aren't--maybe a Petzval or something similar. Not all do so perfectly, of course, but that's a design objective for all modern lenses. Lenses intended for copy work will be most carefully corrected for flat field and zero geometric distortion.

Rick "infinity is flat" Denney

[Michael Clark]

Actually, most lenses are corrected for a flat field. Few aren't--maybe a Petzval or something similar. Not all do so perfectly, of course, but that's a design objective for all modern lenses. Lenses intended for copy work will be most carefully corrected for flat field and zero geometric distortion.

Rick "infinity is flat" Denney

Rick,are you saying that the center field of view of the lens when focused at say 2 1/2 ft the outer field of view will be in focus at 3 1/2 feet? as in Franks example.

Mike

[Frank Petronio]

Oh jeez... Almost 30 years of this and never ran up against it.

OK so if most lenses are flat-field, then when I use my rangefinder to shoot a portrait, and my eye and camera are at waist level, if I use my rangefinder to focus straight ahead, centered upon my subject's belly button - belt buckle area, the lens might be 5-feet away....

And then when I tilt my camera and rangefinder up, to focus on the subject's eyes about 24-inches higher, the distance between the lens and the rangefinder might be about 6-feet apart.

So if just focus on the belt buckle - 5-feet away - I can trust that the person's eyes, 6-feet away, will also be in critical focus, even if I use a fast lens wide-open, for shallow depth of field?

Or should I focus on the eyes, 6-feet away? Even if the camera is pointed and centered on the belt buckle?

I know, a ground glass and loupe solves this dilemma. So does Polaroid. But it's late.....

[Michael Clark]

Never thought about it till I saw your drawings Frank.

[Oren Grad]

When you're working at close range with fast lenses used flat out, focus-and-recompose can indeed be a problem.

The belt buckle should be fine - if it's really in the same plane as the eyes, and if the lens really does have a flat field. In fact, many small-format lenses have enough residual field curvature to matter. You really do have to get to know your specific lenses. Yes, easier to run a bunch of tests to figure it out if the lens is on a digital camera.

Science fiction comes to life - if you have $$$$$ to burn, the latest generation of Hasselblads will take care of the problem for you:

http://press.hasselblad.com/media/1938766/victoronline_10_2009_news_truefocus_phocus.pdf

[genotypewriter]

But has anyone ever made a lens that focuses on a flat field, like figure 2?
If you're referring only to the flatness of the field, it's like what others are saying... most lenses are like that. This is more important for lenses used in projectors than for photography (explained below). You might also want to look up objective plane flatness/curvature vs. image plane flatness/curvature.



I think the practical application would be if you were using very wide lenses or close up lenses wide open, where you might lose focus on the edges of some large flat plane, like a tower, shot vertically? Or crowd scenes, horizontally.I don't think you'd shoot a tower with a close up (close focus only) lens. Shooting a crowd with a close focus lens might get you a restraining order or two :)

The curvature of field matters very little in everyday photography. Say, if you're shooting something like tower, you're going to be relatively far away from it (i.e. relative to the focal length) to fit it in the shot. And then you will have a lot of DOF and using a wide aperture won't really provide much benefit to the shot (other than perhaps an aesthetic effect)... so you can stop the lens down... and then you get even more DOF. So at the end of the day, the rate at which DOF increases, as the subject distance increases, helps in minimising the negative effects of field curvature.

The exception to this is in astrophotography where field flatness has to be specifically corrected for in telescopes. Why DOF doesn't take care of things here is a bit beyond... me. Perhaps it has to do with extreme magnifications. Haven't really looked in to it, to be honest.

Coming back to Earth, field curvature can mostly be a problem if you're shooting something really up close without a good close focus/macro lens that is optimised for such distances. E.g. when you're taking a picture of a coin or a stamp at a high magnification using an ordinary lens, etc.

GTW

[rdenney]

Rick,are you saying that the center field of view of the lens when focused at say 2 1/2 ft the outer field of view will be in focus at 3 1/2 feet? as in Franks example.

Yup.

Rick "noting that large-format lenses tend to be better corrected for field flatness than are strongly retrofocus lenses for SLRs" Denney

[Frank Petronio]

Thanks. I realized that in the real world it wasn't impacting my normal photography very often but perhaps with the rangefinder/portrait example it has? Because I didn't know any better after 28 years of shooting?

[rdenney]

And then when I tilt my camera and rangefinder up, to focus on the subject's eyes about 24-inches higher, the distance between the lens and the rangefinder might be about 6-feet apart.

Just as with a view camera, the focus plane intersects the lens board plane and the image plane. When the image and lens planes are parallel, as they are in a fixed-body camera, the focus plane will also be parallel. So, if the image plane is parallel to the plane containing the eyes and belt buckle, and if the lens is well corrected for flatness, both will be in focus. If you tilt the camera (image plane), then maybe not.

The only lens I've owned that shows a strong field curvature is a 45mm Mir 26--a Ukrainian medium-format wide-angle lens. At wide apertures, the corners are fuzzy--unless there is something in subject field in the right place.

http://www.rickdenney.com/leesburg_courthouse_fence_45_lores.jpg

Notice how the branches in the upper corners are in focus, even though the focus plane is on the nearby iron fence? That's field curvature. It gets better on this lens as you stop down. I've never seen a hint of that on my Super Angulons or other LF lenses.

Rick "who posted that picture before having a calibrated color management" Denney

[William Whitaker]

I'm optically ignorant and have never studied any of this (no silly photo school for know-it-alls)...

Gosh, Frank... And you live at the Optical Center of the Universe! Why don't you hie yourself over to UR before drop/add is over?

I really like your illustrations! Interesting the way you draw hair... :)

[Steven Tribe]

"Rick,are you saying that the center field of view of the lens when focused at say 2 1/2 ft the outer field of view will be in focus at 3 1/2 feet? as in Franks example."


No, the outer field of view will still be in focus at an arc 2 1/2 feet from the camera lens. But our brains are built to view pictures and cinema screens that are flat - the real distance to our eyes vary between the centre and edge of the screens. Actually, the edges in the example have a nearer focus - perhaps 2 feet.

I am sure all simple lenses, have lots of uncorrected curvature of field. It has been used to enhance portrait composition - saving time with camera adjustment.

[Frank Petronio]

So... next time I am using a Crown Graphic, Technika, or my trusty Littma... with rangefinder focusing... assuming Symmar or Xenar type lenses... WHERE should I focus?

Straight ahead, figuring the subject's belt buckle and face are in roughly the same plane?

Or should I tip my camera and rangefinder up, focusing a slightly longer distance, so that the eyes are perfectly focused?

(I already know that I should trust and verify, confirm on the ground glass... but what do you think?)

[Brian K]

If I recall correctly a true flat field lens will focus on a flat plane. Consider an enlarger. When making an enlargement most of us focus on the center of the projected image. And assuming that you are using a glass carrier and your enlarger, lens and easel are all perfectly aligned, then the projected edges of the image will be sharp even though you focused on the center and regardless that there might be a substantial difference between the edges and the center to the lens.

However I do not think, as someone wrote in this forum that most lenses are corrected for flat field. There's little requirement for that except in shooting copy work or other flat subjects. Most scenes are 3 dimensional, and if the lens were corrected for flat field, then the DOF scales on lens are pretty useless. Because if your lens is flat field then an object that is the same distance away as the plane that you focused on, but not actually on that plane, will be out of focus. In Frank's drawing, a flat field lens aimed at and focused on the center of that arc would mean that the higher parts of the arc, the ones that maintain the same distance to the camera would be out of focus if he kept the lens aimed at the center of that arc. That has not been my experience when shooting still life with sironars.

[rdenney]

"Rick,are you saying that the center field of view of the lens when focused at say 2 1/2 ft the outer field of view will be in focus at 3 1/2 feet? as in Franks example."


No, the outer field of view will still be in focus at an arc 2 1/2 feet from the camera lens. But our brains are built to view pictures and cinema screens that are flat - the real distance to our eyes vary between the centre and edge of the screens. Actually, the edges in the example have a nearer focus - perhaps 2 feet.

I am sure all simple lenses, have lots of uncorrected curvature of field. It has been used to enhance portrait composition - saving time with camera adjustment.

I have photographed many flat objects from two or three feet away, and I've never yet met a lens with a spherical curvature of field. If the center is 2.5 feet from the camera and the edge, because of the wide field and near position, is 3.5 feet, the camera will be focused on both, assuming (as in Frank's diagram) that both are in a plane parallel to (or intersecting with) the lens and image planes. The photo I showed has curvature that goes the other way--the corners focus on a distance farther than the center focus point.

Rick "believing the evidence of his own eyes" Denney

[rdenney]

However I do not think, as someone wrote in this forum that most lenses are corrected for flat field.

If you qualified that as "perfectly corrected", I would agree with you. I have done quite a bit of lens testing as a hobby, and have tested the flatness of field of some truly great and truly crappy lenses. Most lenses are reasonably well-corrected for flatness of field, to the level of accuracy implied by Frank's question.

Any of us can test this with a loupe on our ground glass. A truly curved field is quite easy to see. Just focus the camera on a wall that is in the Schmeimpflug focus plane.

Frank, you have discovered one reason why rangefinders have limited near focus distance. There is no only parallax on the view, but also parallax on the focus if you aim the focus spot around too much. Focus on the subject plane that is parallel to the image plane on your fixed-body camera.

Rick "thinking of the classic newspaper-on-the-wall quickie lens test" Denney

[Brian K]

Rick the sample image you show would indicate to me a reverse field curvature. The focus is on the fence which is much closer than the trees of which the plane that the trees are on is out of focus in the center. That says to me that the field curvature goes away from the camera.

I think what happens in most circumstances is that normal acceptable field curvature is simply not noticeable. In order to really see the affects of it you need to show a very large amount of the arc indicated in Frank's drawing. Because the lens has a narrow angle then the amount of curvature being depicted is such a small amount and not so radically different in distance than the plane that is being focused on as to not be noticeable. And in order to show enough of that wide arc you'd have to use a very wide angle lens which inherently had tremendous DOF.

When I was shooting a lot of cosmetics, the distance from the various subjects in the image was just as important as was the plane of focus. If I focused the camera on a lipstick in the center of the image, another lipstick on the far right side on a different plane but the same distance would also appear to be in focus.

[GPS]

...

However I do not think, as someone wrote in this forum that most lenses are corrected for flat field. There's little requirement for that except in shooting copy work or other flat subjects. Most scenes are 3 dimensional, and if the lens were corrected for flat field, then the DOF scales on lens are pretty useless.
...

So in your opinion the DOF of a lens is achieved by curving the field of focus? And once you have a flat field lens the DOF is not there anymore? Good grief...:rolleyes:

[Bart B]

When I worked at Hewlett-Packard, a good friend designed chips that were made in the site's integrated circuit facility. Each layer for the chips was made by taking a picture of a drawing of that layer focused on the silicone wafer. He said the steppers used lenses very well corrected for field flatness were used so the chip's layers would perfectly align with each other. Nikon made those steppers whose lenses resolved 1000 to 2000 lines per millimeter on the wafer.

The site's documentation group used an copy camera with a "copy" lens to photograph 20x25 inch flat artwork.

In both instances, the image plane (silicone wafer or film) was exactly parallel with the subject plane (chip layer drawing or document page artwork) The photographer in the site's graphics group told me that while most pro camera lenses are fairly well corrected for field flatness, flat art copy camera and IC stepper lenses are better corrected for field flatness. But good handheld camera lenses' (the 35mm Nikons and and 4x5 Graflexes HP used for magazine ads, user and service documentation) field curvature usually isn't enough to be noticable.

He also mentioned that lenses used to focus best at fixed distances (IC stepper lenses at several inches and aerial camera lenses at infinity) for good field flatness do better for use at one distance. And stepper lenses are made for only one light wavelength to get best resolution on a flat surface of a flat subject at the high numbers they get. Nikon's IC stepper lenses resolve at 65 nanometers or better these days; that's about 15,000 lines per millimeter (if my math's right). That has to be why a 128GB compact flash card has as much memory as it does.

[Oren Grad]

So... next time I am using a Crown Graphic, Technika, or my trusty Littma... with rangefinder focusing... assuming Symmar or Xenar type lenses... WHERE should I focus?

Straight ahead, figuring the subject's belt buckle and face are in roughly the same plane?

Or should I tip my camera and rangefinder up, focusing a slightly longer distance, so that the eyes are perfectly focused?

(I already know that I should trust and verify, confirm on the ground glass... but what do you think?)

Verify the behavior of your camera before you enter the field of battle. Next time you buy a 4x5 RF camera and before you start to use it to expose film, set up the camera on a tripod in front of a flat subject at a typical working distance, grab a loupe and see for yourself on the GG what happens to different parts of the subject plane as you focus in different places.

[Brian K]

So in your opinion the DOF of a lens is achieved by curving the field of focus? And once you have a flat field lens the DOF is not there anymore? Good grief...:rolleyes:

Where did you get that from?

What i'm saying is that if focus follows distance and not just a flat plane regardless of distance, then the use of distance for a DOF scale as the means to determine DOF is flawed, as a DOF scale would then only be reliable at determining DOF on a flat plane and not by the same distances of any other subjects within the photograph.

In the diagram I include, points A and B are the same distance, point C is obviously further away. With a non flat field optimized lens, I would assume that point B should be in focus because it is the same distance from the camera as point A. If it is not in focus then the use of a DOF scale on a lens is a flawed tool.

With a flat field lens, I would assume that the plane of focus has been flattened to allow point C to remain in focus and point B would go out of focus.

[Stephane]

Petzval lens for portrait (@f/4): easy to see the curved field...
Is that a good example?

http://farm4.static.flickr.com/3496/3972749816_768fb45e69_o.jpg

[jnantz]

frank

enlarger lenses are corrected for flat field and copy work.
not many of the tominon lenses will cover a full sheet of film
except doing close up/copy work, except for the 127 ...
it covers without problem and is flat field in a copal press shutter
that can also be removed from the shutter
and you can replace it with a g-claron, also flat field.

have fun !
john

[Helen Bach]

I agree with everyone.

Field curvature and correction for astigmatism are related - simple lenses can be corrected for astigmatism, but that causes field curvature. Lens designers try to correct for both, of course. They may not achieve perfection! Hence a lens that does have correction for field curvature may have residual curvature, but the plane in focus will not be spherical (with the front nodal point as the centre of the sphere) at all focus distances.

Correction for field curvature is likely to be distance-dependent, because the relative differences between on-axis and extra-axial object-side distances changes with distance (ie it's relatively greater at short distances than long). This means that the plane in focus can be backward curving at long distances and inward-curving at short distances. (It can even be wavy, in fact.)

It's best to learn the characteristics of your lenses at the distances you use them - this comes naturally.

Errors using 'focus and recompose' (which supposes a spherical field) are fairly obvious at wide apertures to users of small and medium format rangefinders.

Best,
Helen

[Steve Barber]

I don't know nothin bout no flat field focus. Maybe he do, but don't go blamin me iffn he got it wrong:

http://web.archive.org/web/20090202222234/http://wisner.com/myth.htm

[aduncanson]

Please pardon my ardor, but of course most photographic lenses are corrected to image a subject plane onto the film plane. Remember the terms used in expressing the Scheimflug Principal, "subject plane", "lens plane" & "image plane"? Thankfully they are not "subject sphere" & "image sphere" - what a nightmare that would be.

There are exceptions, Petzvals perhaps, (I have no experience), very inexpensive lenses, (like the Holga's), lenses designed to image a curved subject, like a cardboard mounted 35mm slide or the face of a CRT, onto a flat screen or film. Also, many wide angle lens designs have a significantly curved field. As in the example provided by Rick "Who usually has something worthwhile to say" Denny. the subject field of these lenses is often convex from the camera's perspective.

I remember buying my first super wide angle lens, a 90mm/6.3 Komura (for a princely sum since this was quite a while ago.) It came in the house and as soon as I could I slapped it on to a camera and pointed it out the window toward the row of trees at my property line. I was sickened by the lack of corner sharpness. Later I read somewhere about the practice of "focusing in (http://www.largeformatphotography.info/forum/showthread.php?p=31522&mode=linear#post31522)" often required with wide angle lenses and I made peace with my Komura.

Since Rick mentioned never seeing this on his Super Angulons, I just checked my 47/8 SA on a 6x9cm camera and could not detect the phenomenon on the ground glass. That does not mean it does not exist.

- Alan

[engl]

That Komura 90/6.3 is a retrofocus design, one of very few such lenses for large format. Non-retrofocus wides, which is the norm for large format lenses, suffer much less from field curvature.

[Michael Clark]

Very nice thread, thanks people for the replies.Thanks for posting Frank, nice cartoon also.

Mike

[GPS]

That Komura 90/6.3 is a retrofocus design, one of very few such lenses for large format. Non-retrofocus wides, which is the norm for large format lenses, suffer much less from field curvature.

In the light of the recent discussion of retrofocus LF lenses I have to ask you - what exactly do you mean when saying - LF retrofocus design lenses? What order of retrofocus do you have in mind?
LF lenses with a "slight" retrofocus effect (several cm) are very common, which contradicts what you say. "True" LF retrofocus lenses (using the scholarly definition) are non existent - which contradicts you too.
Non retrofocus wides suffering less from field curvature? Isn't it just the opposite that the negative front element reduces considerably the field curvature of the lens..?

[Colin Robertson]

Frank,
http://www.rokkorfiles.com/24mm%20VFC.htm
just for fun

[Brian K]

Petzval lens for portrait (@f/4): easy to see the curved field...
Is that a good example?

http://farm4.static.flickr.com/3496/3972749816_768fb45e69_o.jpg

I like the affect! What length and flavor lens in that?

[aduncanson]

In the light of the recent discussion of retrofocus LF lenses I have to ask you - what exactly do you mean when saying - LF retrofocus design lenses? What order of retrofocus do you have in mind?
LF lenses with a "slight" retrofocus effect (several cm) are very common, which contradicts what you say. "True" LF retrofocus lenses (using the scholarly definition) are non existent - which contradicts you too.

I used to assert that the Komura 90mm/6.3 was a retrofocus design but I was chastened by one of the better informed participants here. So I limit myself now to only those things which I can observe directly. In that vein, my Komura 90/6.3 requires about 130mm from the front of the lens board to ground glass to focus to infinity. When I have the very stiff standard bellows on my camera, this turns out to be quite a nice feature.

[Stephane]

I like the affect! What length and flavor lens in that?

9" petzval, at f/4 on 8x10
lens was in his face

[Sevo]

I used to assert that the Komura 90mm/6.3 was a retrofocus design but I was chastened by one of the better informed participants here. So I limit myself now to only those things which I can observe directly. In that vein, my Komura 90/6.3 requires about 130mm from the front of the lens board to ground glass to focus to infinity.

IIRC the Komura is unusually long for such a slow and short lens, so that may mean the node is still well within the glass. I'd only consider it a designed for retrofocus if there should be more than 90mm between focal plane and rear element at infinity. A rear node half way down the rear group is perfectly normal for large format wides, and a side effect of low light falloff lens designs rather than done for retrofocus purposes.

[Ken Lee]

If most View Camera lenses didn't focus mainly along a flat field, there would be little point in using a View Camera.

See View Camera Focus (http://www.trenholm.org/hmmerk/HMbook18.html) by Harold Merklinger

http://www.kenleegallery.com/images/forum/FilmPlane.jpg

[Oren Grad]

If most View Camera lenses didn't focus mainly along a flat field, there would be little point in using a View Camera.

Why? Unless the field projected by a lens is so severely distorted that it has no practical use for photographic imaging at all, it seems to me that movements still add flexibility in solving imaging problems, and thus would still have utility.

[rdenney]

I agree with everyone.

Field curvature and correction for astigmatism are related - simple lenses can be corrected for astigmatism, but that causes field curvature. Lens designers try to correct for both, of course. They may not achieve perfection! Hence a lens that does have correction for field curvature may have residual curvature, but the plane in focus will not be spherical (with the front nodal point as the centre of the sphere) at all focus distances.

Correction for field curvature is likely to be distance-dependent, because the relative differences between on-axis and extra-axial object-side distances changes with distance (ie it's relatively greater at short distances than long). This means that the plane in focus can be backward curving at long distances and inward-curving at short distances. (It can even be wavy, in fact.)

+1 Helen.

Rick "whose experience is that high-end MF and LF lenses do pretty darn well in this category" Denney

[GPS]

IIRC the Komura is unusually long for such a slow and short lens, so that may mean the node is still well within the glass. I'd only consider it a designed for retrofocus if there should be more than 90mm between focal plane and rear element at infinity. A rear node half way down the rear group is perfectly normal for large format wides, and a side effect of low light falloff lens designs rather than done for retrofocus purposes.

That's correct. The scholarly definition of a retrofocus lens has its sense.

[GPS]

If most View Camera lenses didn't focus mainly along a flat field, there would be little point in using a View Camera.

...

Well, view cameras well used a lot even before lenses started to be corrected for flat field...:)

[aduncanson]

This caused me to look up the formal definition of a retrofocus lens (in published books, not just from the internet) and found Both Kingslake & Cox defined a retrofocus, inverted telephoto, or reversed telephoto lens as one with back focus (distance from the rear element to the image at infinity) is longer than the focal length. In other words the near node is outside of the rear element. By this definition the Komura 90/6.3 is not a retrofocus lens since the rear node is about 40mm behind the front of the lens board, while the rear element extends to about 54mm behind it.

The Komura is a lens with an exceptionally long flange focal distance for the 90mm focal length. Again, I measured it to be 130mm for the Komura compared to 99 or 100mm given by Schneider for the older Super Angulon 90mm lenses and 94 or 98mm (per Rodenstock) for the 90mm Grandagon-N lenses.

Both Kingslake and Cox also incorporated into their definition of retrofocus the design feature that a negative or diverging component precedes a positive or converging one. Of course all of the super wide designs (Biogon, Grandagon, Super Angulon, etc.) are better understood as a double reversed telephoto with large negative elements at both ends.

[Ken Lee]

I stand corrected.

I guess that even with a curved field, shift and rise are just as useful. With tilt and swing, we end up adjusting the curve - rather than the plane - of focus.

[GPS]

I stand corrected.

I guess that even with a curved field, shift and rise are just as useful. With tilt and swing, we end up adjusting the curve - rather than the plane - of focus.

Even with the shift and rise of not flat field lenses you adjust the "curve" rather than the plane of focus...:)

[Leigh]

Actually, most lenses are corrected for a flat field.
Mr. Denney and I agree.

Drum roll and trumpets, please. :D :cool: :rolleyes:

- Leigh

[rdenney]

Mr. Denney and I agree.

Drum roll and trumpets, please. :D :cool: :rolleyes:

- Leigh

Who knew?

Rick "scared to go outdoors for fear of lightning" Denney

[Randy]

So... next time I am using a Crown Graphic, Technika, or my trusty Littma... with rangefinder focusing... assuming Symmar or Xenar type lenses... WHERE should I focus?

On the nipple...I'd say...

Oh for gad sakes Frank, close down to f/32 and shut up :)

[Joe Forks]

On the nipple...I'd say...

Oh for gad sakes Frank, close down to f/32 and shut up :)

I was going to suggest the opposite. Hang an Eidoscope on the end of that thang, and open it all the way up ;)

[johnielvis]

I've run into the same problem...lots of work arounds....pre-shifting the focus....or just learning to back up just slightly (when you tilt the camera back down after focusing)...it all depends on how close you are....you absolutely can't go wrong by say focusing on the nose rather than the eyes to start out...and you tilt down...this piuts the eyes closer to where the focus is (which is now behind the nose)...you gotta learn to compensate for the camera/lens/distance you're using...

the only foolproof way is to have a rangefinder that is tiltable so you can synch it 'as composed"....I think if you just practice focusing closer at things and backign up if necessar...like stick your neck OUT to focus...tilt...neck back...see what happens..you can get a guage---it's totally learned behaviour---that's why tlr/slr rules for this kind of thing.

I also considered some kind of lens tilt to compromise---maybe I'll do the calculations some day to see if there's any kind of simple method to set it so you're always on.

[ic-racer]

But has anyone ever made a lens that focuses...

In terms of Curved Field lenses, the first that comes to mind is the Kodak projection Ektanar lens, taking into account the film is curved in the mount.

A more sophisticated approach was taken by Minolta in this interesting curved field prototype. From what I recall, Fuji made a prototype like this also:

http://img.photobucket.com/albums/v670/ic-racer/MinoltaCurvedField.jpg