Which lens design is shown in this diagram? f/4.5, 10 elements, 8 groups, 78°

[genotypewriter]

101379
Left: Object end
Right: Image/film end

Max aperture: f/4.5
Construction: 10 elements in 8 groups
Angle of good coverage at ∞: 78° from f/4.5
Optimised focal distance: ∞ and thereabouts
Fwiw, it's also multi-coated.

Is there a close match or is it comparable to something that you've seen? I find the flatter front surface interesting/different.

Thanks in advance!

G

[Jac@stafford.net]

It looks like a Fujinon 55mm EBC for MF.

[Taija71A]

It looks like a Fujinon 55mm EBC for MF.

____

WOW!!!
Great 'Eye' Jac! :)


http://www.gf670.com/downloads/GF670W_Brochure.pdf

________

[genotypewriter]

Thanks :) but I was hoping if anyone knew if it's similar to any classic designs like biogon, etc. The lens looks similar to a reversed Voigtlander Ultron 28/2 but some +/- surfaces are -/+ so I don't think they came from the same place.

Being a non-retro focus rather modest 78-degree wide-angle, I find it interesting how it has 10 elements when even a 105-degree f/4.5 Super Symmar XL has only 6 elements. Any thoughts?

[Bob Salomon]

Thanks :) but I was hoping if anyone knew if it's similar to any classic designs like biogon, etc. The lens looks similar to a reversed Voigtlander Ultron 28/2 but some +/- surfaces are -/+ so I don't think they came from the same place.

Being a non-retro focus rather modest 78-degree wide-angle, I find it interesting how it has 10 elements when even a 105-degree f/4.5 Super Symmar XL has only 6 elements. Any thoughts?

Some Rodenstock wide angles for digital have up to 15 elements. Several have 10.

[EdSawyer]

Sometimes elements are used for things like field flattening, as opposed to correction. Also, modern lenses don't have the same choice of glasses for construction as they did pre-2002 or so, so optical formulas have a smaller choice of glass types and thus other compromises to a design may need to be made.

[EdSawyer]

Sometimes lens patents can be illuminating also , as to why certain design choices are made. Fuji obviously spent some money on this lens. It would be interesting to see how it stacks up against say the 50 and 65 from Mamiya for the mamiya 7.

[Jac@stafford.net]

Thanks :) but I was hoping if anyone knew if it's similar to any classic designs like biogon, etc. angle,

With respect, it does not matter how the lens is described. If you wish to know how it renders you have to use it yourself. ...and it is obviously not a Biogon type.

[genotypewriter]

Thanks for the inputs everyone. I didnt have a look at the Digitar and Digaron era designs. Didn't think they would be any more relevant than SLR designs because they're made to have a long back focus. This GF670W or Bessa 667/IIIW lens is only for film afterall and the lens isnt sitting suspiciously further away (or too close) going by the focal length. But I want to get to the bottom of this so I will have a closer look at digital stuff too.

Jac@stafford: oh I got a brand new one last week and love the images im getting with it. It's like a Leica for people who appreciate formats larger than 35mm :) I only asked this question to understand the lens design. I find this lens interesting because of the large number of elements used while having a fairly conservative f number and an angle of view, being optimised for low magnification and made only for film. So I was wondering if anything like this is there in LF too.

Would like to hear more thoughts from people... especially about the flat front element and the bulging rear element. What reasons are there to possibly use a lens in the reversed position if it's to be used for low magnifications?

Thanks again.

[Len Middleton]

Actually have an old Nikkor-UD wide angle 20mm f3.5 lens that will do you one element more (11), but that is an early pre-AI retro-focus design that followed their old 21mm design that required mirror lock up.

[ridax]

Being a non-retro focus rather modest 78° wide-angle, I find it interesting how it has 10 elements when even a 105° f/4.5 Super Symmar XL has only 6 elements.

First, 78° is wide enough for a lens intended to be used wide-open. Fewer elements LF lenses are made to be used well stopped down; a lens fast not only for focusing but for shooting too, has to imply way more elements for decent sharpness.

Second, the above does look as a retrofocus (though not as much retrofocus as SLR lenses are of course). The front negative element is perhaps more powerful then the 2nd and 3rd positive ones combined. Besides, the 4th concentric cemented group is probably negative, too. Retrofocus design in often useful in non-SLR cameras also.

And BTW generally, more elements do not mean better performance. In lens design, substituting two or more elements for a single one is actually a cheaper Ersatz for meticulously recalculating the original few-element design to perfection (go compare Tamron to Leitz for example). The results are mediocre but way way cheaper and faster to obtain... Well, sharpness can be OK nevertheless but out of focus rendition of a multi-stand-alone-elements lens is almost never good, as messing with more and more elements means introducing more and more higher-order aberrations, and those are far from being any good for the resulting picture beauty.

[Sevo]

Second, the above does look as a retrofocus (though not as much retrofocus as SLR lenses are of course). The front negative element is perhaps more powerful then the 2nd and 3rd positive ones combined. Besides, the 4th concentric cemented group is probably negative, too. Retrofocus design in often useful in non-SLR cameras also.

Probably right. Not for placement reasons, the GF670W does not even fold, but (retrofocal) asymmetry can be used to improve illumination.

[Jac@stafford.net]

Jac@stafford: oh I got a brand new one last week and love the images im getting with it. It's like a Leica for people who appreciate formats larger than 35mm :) I only asked this question to understand the lens design. I find this lens interesting because of the large number of elements used while having a fairly conservative f number and an angle of view, being optimised for low magnification and made only for film. So I was wondering if anything like this is there in LF too.

Excellent points, thank you. I look forward to learning more of the same.

While it is not LF, I enjoy 6x12cm work using Rodenstock Grandagon lenses. We might look to Grandagons, etc. for more information regarding LF applications.

[EdSawyer]

"a lens fast not only for focusing but for shooting too, has to imply way more elements for decent sharpness." - Not necessarily true. Look at the Xenotar design, for example. Or any gauss design. Plenty fast, sharp, and only 5 or 6 elements in some cases. Another example would be fuji's own 65mm as used on GSW690III - only a half stop slower, and only 6 elements.

Can't say whether this lens is retrofocal or not - but if it is it's not by much from the looks of it. Clearly based on the camera it's sitting pretty close the film plane.

re: "more elements do not mean better performance" - this is not something that can even be generalized I don't think. More elements = more performance in some cases, and not in others. Rather than counting the elements it's more useful to evaluate the design based on other factors. Counting elements/groups certainly is one of those factors but only one out of many.

ultimately what matters are the results and how it renders.

-Ed

[Dan Fromm]

What reasons are there to possibly use a lens in the reversed position if it's to be used for low magnifications?

Thanks again.

Why do you think the lens is reversed? It is oriented the way its designer intended.

[ridax]

"a lens fast not only for focusing but for shooting too, has to imply way more elements for decent sharpness." - Not necessarily true. Look at the Xenotar design, for example. Or any gauss design. Plenty fast, sharp, and only 5 or 6 elements in some cases. Another example would be fuji's own 65mm as used on GSW690III - only a half stop slower, and only 6 elements.

Xenotar is 55°. Not a beast to compare to wideangles at all. And so are other fast double Gauss lenses.

The 65mm on GSW690III is f/5.6. That's no way a half-stop from f/4.5. The very basic mathematics reveals those are good 2/3 of a stop. To understand the meaning of a difference that great and the amount of difficulties in designing a lens that faster, think of the difference between f/1.8 and f/1.4 50mm small format lenses, for example - or between f/1.5 and f/1.2. That's NOT small.

And yes the rule that a fast wideangle lens for wide-open usage has to imply way more elements for decent sharpness, is necessarily true. (Well, actually, a single aspherical element may be substituted for several spherical ones to make the total amount of elements far less. But there are no asphericals in the lens discussed. And BTW, an aspherical surface is even worse a killer of decent out of focus rendition then the higher-order aberrations introduced by too many spherical surfaces.)


re: "more elements do not mean better performance" - this is not something that can even be generalized I don't think. More elements = more performance in some cases, and not in others.

I've made it clear enough what I meant by performance. Yes sharpness in a multi-group lens may be good. But out of focus rendition is NEVER anywhere near the one of carafully designed simpler lenses.


Rather than counting the elements it's more useful to evaluate the design based on other factors.

Such as...?


ultimately what matters are the results and how it renders.

A universal truth, no doubt. But not too new... and neither too informative I'm afraid. And IMHO not too appropriate in this topic as this one looked as a pretty theoretic one from the start.

[Sevo]

And yes the rule that a fast wideangle lens for wide-open usage has to imply way more elements for decent sharpness, is necessarily true.

No, not really. Once you throw in low falloff/even illumination, that statement is somewhat closer to the mark - but at a penalty of low speed and quite nasty falloff, two group/four element lenses of Topogon/Metrogon type were already capable of very high resolution (and flatness of field).

[ridax]

at a penalty of low speed and quite nasty falloff, two group/four element lenses of Topogon/Metrogon type were already capable of very high resolution (and flatness of field).

Yes - low speed. And we are talking about fast lenses.

But your point on light fall-off is very true. In a wideangle lens design, a lot of effort (and still more lens elements) goes to diminish this particular fault.

And BTW a lot of fall-off really means a lens in not actually fast - except in the very center of its field...

[ridax]

I find it interesting how it has 10 elements when even a 105° f/4.5 Super Symmar XL has only 6 elements.

I've just realized my words above apply to the Super-Symmar XL, as the Super-Symmar is actually aspherical -


a single aspherical element may be substituted for several spherical ones to make the total amount of elements far less.

[genotypewriter]

Thanks everyone for the interesting points raised.

I'd like to add that while the SSXL (4.5 or otherwise) has aspherical elements which likely reduced the number of spherical elements needed for comparable correction, still the difference between 78° and 105° is huge and would have needed more elements anyway. So the SSXL having I don't quite agree that the SSXL having an explicitly stated recommended working aperture of f/16 (or is it f/22?) means that the Fuji lens here is optimised to be shot wide open at f/4.5. Also it's not very clear that the Fuji lens is retrofocus because as I said, it sits a decent way away from the film plane although the proportions of the camera makes things look like the lens is very close to the film.

So my conclusion is... it's difficult to draw a lot of insight from existing lenses without making apple-to-orange comparisons. And with lens designs being so closely related it's sometimes not easy to make distinctions.

Dan: I was asking about any lens, not just this lens.

Thanks again for all the inputs... would also be interesting to hear thoughts from anyone who has designed lenses on a ray tracer.

G

[genotypewriter]

Thanks everyone for the interesting points raised.

I'd like to add that while the SSXL (4.5 or otherwise) has aspherical elements which likely reduced the number of spherical elements needed for comparable correction, still the difference between 78° and 105° is huge and would have needed more elements anyway. So the SSXL having I don't quite agree that the SSXL having an explicitly stated recommended working aperture of f/16 (or is it f/22?) means that the Fuji lens here is optimised to be shot wide open at f/4.5. Also it's not very clear that the Fuji lens is retrofocus because as I said, it sits a decent way away from the film plane although the proportions of the camera makes things look like the lens is very close to the film.

So my conclusion is... it's difficult to draw a lot of insight from existing lenses without making apple-to-orange comparisons. And with lens designs being so closely related it's sometimes not easy to make distinctions.

Dan: I was asking about any lens, not just this lens.

Thanks again for all the inputs... would also be interesting to hear thoughts from anyone who has designed lenses on a ray tracer.

G