analogue lens with digital back

[Emmanuel BIGLER]

Digital light rays must be perpendicular to the sensor.

I disagree, but depends on what 'digital' means.
If digital means 'any kind of silicon image sensor with a Bayer aray', then I definitely disagree, look at the specs of this Kodak 39 Mpix sensor: silicon alone is perfectly able to detect slanted rays. See figures 18 & 19 page 8 of this paper.
www.kodak.com/ek/US/en/31Mp_and_39Mp_Full-Frame_CCD_paper.htm
The 39 Mpix sensor does not have micro-lenses.and still exhibits an efficiency of 70% for slanted rays at 40° of incidence. This loss of efficiency is less that the "natural" loss of any classical large format lens @40° of incidence (80° of total field angle).
For example a classical 105° wide-angle lens only delivers 45% of illumination at 40° off-axis (data taken from Schneider's specs for the SA 5,6-75mm classical WA lens). It means that one can use this sensor up to 45° of incidence (total field 90°) without noticing that some kind of "digital absorption of rays" occurs inside, the phenomenon will be hidden behind the natural light fall-off common to all quasi-symmetrical WA lenses.

Hence, micro-lenses are the issue, i.e a component 100% analog-optical and 0% digital, not silicon, and not the analog-to-digital converter behind

Not kidding : medium format sensors used in professional backs do not all feature micro-lenses. Sure, the ISO sensitivity, in consequence, is low, but not lower than our good ol' film.

The last generation of "digital" lenses from Schneider-Kreuznach and Rodenstck deliver superb results to sensors with no micro lenses. The design of those modern WA lenses belongs to the family of quasi-symmetrical designs although a tiny amount of retrofocus design is included in order to keep the back focal distance, i.e. the (last lens-element to sensor) distance manageable at infinity.
And you can correct light fall-off either with a centre filter (Rodenstock recently introduced a new kind of optical centre filters), or by post-processing, as usual.


P.S. to the best of my knowledge, Kodak sold its silicon sensor division to semeone else at the end of last year ...

[Frank Buchner]

I found the article very useful where there is something about this. Someone more familiar with this type of information on the web?
http://www.luminous-landscape.com/re...al-ebony.shtml

[Frank Buchner]

Nice information about Rodenstock sironar and Schneider super symmar on this forum http://forum.getdpi.com/forum/medium...rmat-lens.html

[Michael E]

A thought for your further testing: During the Photokina 2010 I talked to a Rodenstock engineer. He took over an hour to explain the latest "digital" lenses to me. One thing that stuck to my mind: These lenses have to be used at wide apertures. The examples that he showed on a computer screen illustrated that the full resolution was available in the center at f4, the corners were fine by f5.6, everything over f8 severely suffered from diffraction. This is not compliant with the old ways of LF photography, so make sure to check for the best aperture of any tested lens.

Michael