How are the image circles so big on xpan lenses?

So recently, I've been interested in different approaches to 35mm pano work. Obviously when one goes beyond the 24x36 the image circles of most lenses designed for that format go whack. My question is: how did the designers of xpan lenses enlarge the image circle so much without a significant increase in lens size (from other rf lenses in general)?

Is it purely smart optical design or is there some other factor that I am missing?

I believe the lens was designed for medium formats.
And since it does not have to have the shutter within the lens, the lens can be relatively slim.

I would think the relative slowness (f4 and 5.6) plus the sometimes necessary center filters are part of the tradeoff.

With the lack of prism in the camera you can have the rear element as close as you like to the film plane, that makes that the lens desing can be very compact. Think about the Angulon lenses orginally desing for 6x9 but able to cover 4x5'' stepping down with no movements. They are ridiculously tiny (filter thread M30.5).

I certainly think the lenses were designed for medium format (is xpan technically medium format?), and that the slowness is a tradeoff for larger image circle. I guess I just don't understand the basic principle behind designing lenses specifically for a larger image circle... For example, does a 90mm f4 lens designed for a 24x36 rangefinder have the same image circle size as the xpan 90mm f4??
If it doesn't, why?

It is not difficult, just different optical requirements and so a different starting point (I’m a lens designer).

From a design standpoint, 35mm camera objectives are one small category in the lens design universe. The 35mm frame is just one image size in a wide range of image sizes (and visible spectrum objective lens just one imaging application amongst many).

The diameter of the image circle is defined by the diagonal, and is a basic 1st order property established when the design work begins.

Like the design of an automobile begins with, say, how many passengers will it carry. Not a big deal at all do define it starting out, but a two seat sports car will have a different (yet similar) design path from, say, a bus. This isn’t a perfect analogy, because the two different optics will be more similar, but you get the idea: you won’t be able to bus a bunch of people around in a sports car even if they both have 4 wheels, an engine, door, etc.

Nodda Duma said:

It is not difficult, just different optical requirements and so a different starting point (I’m a lens designer).

From a design standpoint, 35mm camera objectives are one small category in the lens design universe.

Click to expand...

Very interesting. Any suggestions for reading materials on that general subject that aren't super very incredibly technical?

Kingslake “The History of the Photographic Lens”

Other than that, no. Lol. Optical design is incredibly technical as it’s rooted in physics.

For example, does a 90mm f4 lens designed for a 24x36 rangefinder have the same image circle size as the xpan 90mm f4??
Yes i think so, though in the case of a hypothetical '90' for 24x36mm you'd probably find mechanical vignetting would reduce the coverage to slightly less than 24x65mm. Maybe 24x60mm. There never was a 135mm lens for the XPAN but a Nikon can be fitted with an adaptor and covers most of the 'pan' format. Of course NOT rf coupled. It's a really good combination.

cfrye said:

Very interesting. Any suggestions for reading materials on that general subject that aren't super very incredibly technical?

Click to expand...

You could look for some college textbooks on physics. They're going to be kind of technical, but lens design is by it's nature, technical. You can't really understand it without first understanding the math behind it. Be sure to browse through the book first, as it's not covered in every college level physics book, but any engineering student will spend at least some time working with the basic principles of lens design.

Basic lens design isn't really all that difficult if you've got a pretty solid understanding of math. If you not very good with math, it's all going to be Greek. What makes lens design so difficult isn't understanding how it works, but finding a balance between all of the variables you have to work with to come up with a lens that meets all of your design goals. That's why computers are so useful to lens designers. They don't need help with doing the equations once, but computers can be really helpful when it comes to doing the same equations millions of times to figure out the best compromises between all of the variables. It's also why most successful lens designs are based off previously successful lens designs. The tradeoffs between the various principles aren't all connected to single variables (especially when your lens contains multiple elements), you can't just say something like, the bigger the front element, the faster the lens, and be right every time.

It's kind of like baking a cake. If you decide you want to invent a new kind of cake, it'll be a lot easier if you start off with a pre-existing cake recipe and tweak it from there. Your first attempt may not get you there, but you'll learn from it and eventually hone in on your goal. If you were to instead start off with every possible combination of ingredients at the grocery store, it could take a really long time before you come up with something edible.

They are enlarger lenses.

Ktcamera or Kt163 dot com (no need to surf its gone company gone) was a cameradealer at photokina. only dont remember year maybe 2000?
he showed C/Y-Xpan-adapter i had seen images online. pity i didnt bought. and one cannot find anymore and i have no contact to ask.
if i taking into account image-circle of Zeiss PC-distagon 35/2.8: 63mm it would result in +-5mm shift up and down without vigentting. max is +- 10mm(15° rotating steps).
Maybe with custommade center filter full 10mm is possible. Zoerk once built a 35mm shiftlens. based on mamyia 645 lens. +-20mm shift on FF.
Btw:Marco Kröger on FB (most genious camerabuilder-2D/3D cams) had twinned an exacta to get 24 x48mm. only dont remember which lens he use. no he does not sell any of his creations and the are really fantastic. he is working in a special new analogue-cam shop in Berlin-Germany.