[Leica] Digital Leica ideas

[lrzeitlin at optonline.net (Lawrence Zeitlin)]

The topic of the full frame digital Leica keeps coming up. Here is a 
post I submitted to the LUG three years ago on various implementations 
of the full frame camera. Has Leica made any progress?

"I'm not an optoelectrical engineer but I did stay at a Holiday Inn 
last night.

Actually I didn't stay at the Holiday Inn but I had a couple of drinks 
at a bar nearby with some friends who are real optical engineers from 
IBM's Yorktown Heights Research Lab. After a couple of beers I posed 
the digital Leica M question. Specifically, could conventional lenses 
with a short back focus work with a full sized CCD sensor. The answer 
was the same as suggested by the nay sayers on the LUG - probably not.  
But - and there is always a "but" in discussions with engineers - if 
enough time and money were made available, a digital Leica M could be 
made to work. Here are a couple of the brainstorming suggestions that I 
remember:

Correction plate - a strong planoconcave correction plate in contact 
with the full sized sensor might be the least expensive but it would 
require considerable design work and would not work optimally with all 
lenses. The grazing angle of a 50 mm lens (back focus about 30 mm) is 
about 33 degrees at the corners of the image frame. The Kodak 
KAI-110000 sensor requires an angle no greater than 15 degrees. Thus 
the plate would have to deflect light to the vertical by 18 to 20 
degrees at the corners decreasing to zero degrees at the center. For 
lenses shorter than 50 mm, the corner deflection angle would have to be 
greater, longer lenses would require less deflection. A fixed 
correction plate would work best for only one focal length. One 
suggestion was to supply alternate plates for specific lens ranges, 
another was to use a single plate and correct for light fall off by 
software, the frame actuator triggering the right software mode. An 
alternative suggestion to the planoconcave correction plate was a 
fresnel lens with groove artifacts removed by software. These guys work 
for IBM, remember. The suggestion proposed in the Leica press release 
was to use individual micro lenses over each of the image wells in the 
sensor, most probably molded into a single pastic overlay. This is a 
modification of the fresnel lens method.

Image compressors - If we are not wedded to the concept of a full sized 
CCD and are willing to use a smaller sensor, albeit one with a high 
megapixel count, the easiest approach would be to just use the center 
1/2 to 2/3 of the image. This is a no brainer optoelectrically. It is, 
of course, the approach taken by Epson. When I suggested that many 
Leica users were wide angle fanatics and would object to a change in 
apparent focal length of their lenses, several suggestions for image 
compression that would give the full field of view were offered. The 
first of these was a fiber optic corrector plate. Imagine a fiber optic 
bundle the size of a 35 mm frame. Now heat and draw one end of the 
bundle out until it is the size of a smaller CCD sensor and optically 
fix it to the sensor. Any image formed on the large end will be 
reproduced in smaller size at the smaller end where it is picked up by 
the reduced size sensor. The large end of the bundle could be ground to 
minimize grazing angle and fall off effects. The result might be more 
even illumination than with film. The fiber bundle could be bent around 
from the Leica M inspection port area to the baseplate electronics bay. 
Artifacts and misplaced fibers would be rectified by software. Marty 
Forscher used the fiber optic bundle technique in the very first 
electronic cameras 30 years ago.

Relay lens - The aerial image of any Leica lens could be picked up at 
the film plane and directed by a relay lens to a sensor of any size. 
The optical twists and turns this requires are best left to your 
imagination. This approach is thoroughly practical and is used in many 
scientific instruments. It ain't cheap, however, and the relay optics 
have to be of the same quality as the prime optics.

Afocal lens attachments - A teleconverter type attachment inserted 
between the prime lens and the camera body could be used to increase 
the back focus to the point where a full sized sensor would be 
practical. Of course this would increase the apparent focal length, so 
an afocal wide angle attachment on the front of the lens could be used 
to decrease focal length to keep the status quo. Whether this would 
still be a Leica lens is another story.

Our booze abetted conclusion is that a digital Leica COULD be made 
given enough engineering talent and money. Whether Leica has enough of 
either is open to question. But there is no doubt that a digital camera 
that uses Leica lenses is within the realm of possibility."

Larry Z