[LRflex] Re: R-Option Downdate/Douglas Sharp
- From: "wildlightphoto@xxxxxxxxxxxxx" <wildlightphoto@xxxxxxxxxxxxx>
- To: leicareflex@xxxxxxxxxxxxx
- Date: Mon, 20 Sep 2010 11:07:59 -0400
Neil Gould wrote:
>>>
Could you write a little more about your statement that "A
digital camera will also demand much more from a lens than
film did,,,"? It is my impression that the opposite is true,
at least until the 50+ megapixel sensors become standard
fare.
<<<
I get more detail from the DMR @ ISO 400 than I did from Kodachrome 25.
The web-published tests that suggest a 50+MP sensor is required to equal
film involve black-and-white (no gray) test charts that have no
relationship with real-world subjects. These tests are optimum for B&W
film, and a worst-case scenario for the typical digial camera, because film
at the microscopic level is either 0N or OFF, more binary than digital
cameras! To make grays between black and white a cluster of silver
crystals is required, some ON and some OFF, the proportion varying
depending on what gray you want. This reduces resolution because you need
multiple silver crystals to represent a single gray point. Add color
information (Red, Green, Blue) and resolution is 1/3 of the gray-scale
resolution to represent a single color point.
A typical digital camera incorporates an RGBG filter in the sensor, and
each photosite can record multiple tonal values. Converting the image from
a digital camera to B&W throws away up to 3/4 of the image data (and
resolution), and this is before considering the individual photosite's
ability to record multiple tonal values vs. the need for film images to
clump several silver grains to represent interemediate tones.
>>>
The oblique light ray issue wouldn't affect DMR users
because the sensor is smaller than 35mm, so the image circle
would have mostly (if not all) direct rays. I suspect that
this was a design decision on Leica's part
<<<
It was a determined by the size of the R8/R9 film gate. Leica has
demonstrated with the M9 that full-frame images with much shorter exit
pupil-to-sensor distances (resulting in more oblique geometry) can produce
outstanding image quality.
>>>
a 35mm sensor worthy of R lenses is likely to be too unique to warrant the
production costs.
<<<
An M9 sensor would be fine.
>>>
Frankly, I don't care for the results of R
lenses on full frame Nikon or Canon bodies,
<<<
I don't either. The CaNikon cameras' AA filter, shallower bit depth, and
the sensor's color filter array optimized more for high-ISO performance
than for color quality may be to blame.
>>>
I think the Leica R "solution" is the
S2, as every aspect of that camera and its lens line appear
to support the Leica legend (price included).
<<<
The S2 is not targetted to the R market. Leica is aware that an adequate
solution involves using existing R lenses, which the S2 cannot do.
Doug Herr
Birdman of Sacramento
http://www.wildlightphoto.com
--------------------------------------------------------------------
mail2web - Check your email from the web at
http://link.mail2web.com/mail2web
------
Unsubscribe or change to/from Digest Mode at:
http://www.lrflex.furnfeather.net/
Archives are at:
//www.freelists.org/archives/leicareflex/
Other related posts:
