Optical Prescription Puzzles

[bclaff]

From time to time I encounter optical prescriptions in patents that are incorrect and that I am unable to repair. In the past a few collaborators have helped solve some issues with programs like Zemax so I thought I would try to cast a wider net.

I have restricted my puzzle list to only those that appear to match production lenses and have made a Puzzles Dropbox folder.



In each of these folders are a set of files; for example:

[ATTACH alt=" View "original size" to see better"]3351576[/ATTACH]
View "original size" to see better

PDF is the patent document. Note that I may have added leading zeroes.
TXT is the optical prescription in my format. It's useful to consult the [Notes] section.
ZMX is the optical prescription in Zemax format.
The PNG files are the lens cross sections.
The PNG without a trailing 'P' is the patent figure and the one with the trailing 'P' is the manufacturer diagram that looks like a match.

So, if you have access too tools like Zemax and enjoy puzzles I would appreciate any help. Normally I try to stick as closely as possible to the patent so for example I don't substitute "real" glass (unless the glass is totally missing!).

The Dropbox link is read-only so contact me if you need help getting proposed solutions back to me.

Feel free to pass this along on other fora even beyond dpreview.

--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

 

[bclaff]

Solved.



--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

 

[bclaff]

Just added. See Notes section of the .txt files for details.

 

[bclaff]

More additions

Zeiss Distagon T 21mm F2.8 Contax
Zeiss Tele-Tessar 200mm F4.0
Nikon AI AF Nikkor 85mm f1.8S

 

[Luxo Sydney AU]

In the astronomical community where the optics tend to be simpler than camera lenses but the degree of correction expected is extremely high, there are quite a few examples which suggest the design described in the patent is close but not quite the full thing.

They’re close enough that the patent can be defended, but the designer keeps a bit of “secret sauce” up their sleeve that isn’t disclosed. The result being it remains very difficult to come up with a real design in OSLO or ZEMAX that performs as well as the real thing.

this is common in eyepieces (Televue) and also in APO refractors with 4…6 elements, which is where the problem becomes non-trivial mathematically.

In addition there are issues with the optical properties of the actual glass batches used; while catalogs give a general prescription of a glass type, the smart opticians measure each batch to one or two decimal places beyond what the catalog states and tweak the design accordingly to achieve the best results.

Hence the patent is a close (but inexact) prescription and the actual design lenses from each batch of glass are tweaked. If this is done for complex camera lenses (10 elements or more) to be honest I’d say you have very little chance of deducing the “secret sauce” applied by say the likes of Leica or Zeiss, starting from the patent disclosure.

…and that’s even with the help of automatic optimisation…

 

[bclaff]

In the astronomical community ... there are quite a few examples which suggest the design described in the patent is close but not quite the full thing.

...

I'm not aware of any patents in this area that appear to correspond to actual product since there is a dearth of manufacturer diagrams to match patents to.

Do you have some information?

 

[Luxo Sydney AU]

Ja... look up the eyepiece patents assigned to Televue, their designer is Al Nagler.

When the designs as stated are raytraced they're OK, but they're not as good as the actual eyepieces Televue sells. Over 40 years many have tried to reproduce the designs - notably the Chinese - but they're invariably inferior.

For a long time it has been evident Televue tweaks their designs in some way to compensate for the aberrations of the most common telescopes used by their primary market (fast newtonians, popular in the USA) but exactly how they do this is unclear.

I expect the mature lens designers in the camera industry do similar - ie patent a design that is good, but not stellar; while knowing all along that some specific tweaks do make the design stellar - but those tweaks are not obvious and you're not likely to find them using automatic optimisation in ZEMAX or OSLO.

 

[bclaff]

Ja... look up the eyepiece patents assigned to Televue, their designer is Al Nagler.

When the designs as stated are raytraced they're OK, but they're not as good as the actual eyepieces Televue sells. Over 40 years many have tried to reproduce the designs - notably the Chinese - but they're invariably inferior.

For a long time it has been evident Televue tweaks their designs in some way to compensate for the aberrations of the most common telescopes used by their primary market (fast newtonians, popular in the USA) but exactly how they do this is unclear.

I expect the mature lens designers in the camera industry do similar - ie patent a design that is good, but not stellar; while knowing all along that some specific tweaks do make the design stellar - but those tweaks are not obvious and you're not likely to find them using automatic optimisation in ZEMAX or OSLO.

I'll take a quick look out of curiosity but without manufacturer provided lens cross sections it's supposition that an optical prescription in a patent corresponds to a production lens.