I shoot digital too and have owned quite a few D cameras and to be
honest I would not want to wade through and scan 800 negatives, but
for one off shots I prefer MF or LF film.
No offense intended.
Hi Tammons,
None taken

What you've forgotten here is twofold. First,
there's the "theoretical" versus the real world then there's the
magic of interpolation.
It's not possible to "capture" more detail than a sensor can
resolve, but it's quite easy to "keep" what you have captured and
enlarge it to the point of seeing "marker pixels" (improperly
resolved detail). This pertains primarily to digital. With film
it's virtually impossible to exhaust resolution because you simply
run up against the wall of grain long before resolution exhaustion
becomes an issue and because of this you never see marker pixels -
just can't print large enough to reach that point. So the bottom
line is that film printing is grain limited and digital printing,
because of the extremely low noise, is resolution limited.
While my 1DS essentially exhausts its resolution for a wide angle
hyperfocal landscapes at a little over 16x24 inch print size, for
small fields of view and for subjects with less detail I can and do
make beautiful prints 8 feet on the long axis. We frequently print
life sized images of large tape libraries for electronic trade
shows captured with the 1DS at up to 8 feet on the long axis. These
are better in every way than I can get with my 4x5 because of
grain. The detail is excellent even on close inspection and the
clarity and sharpness is outstanding. Steve Eastwood printed
beautiful head and shoulders portraits made with his 3.5 megapixel
D30 at 75" on the long axis and said he could go larger. Try that
with a 645 and you will be sorely disappointed.
The problem with using a math approach to determining the possible
is that there are too many variables and it just doesn't obtain. I
remember with Dr. Clark was determining by extrapolation of image
resolution from a small HP and 2 megapixel digicam that to reach
35mm film quality digital would require 24 megapixel resolution.
Finally he actually used a six megapixel dSLR and the story is
drastically different now. Others estimated as much as 39
megapixels resolution. Obviously we all know that that argument
died when expectations based on numbers and the reality of
practical experience clashed.
Single pixel resolution? Yes, the 1DS Mark II will do single pixel
resolution. Actually, my Sigma SD10 will do single pixel resolution
(10.3 million sensors, 3.5 megapixel file size resolution). Check
Phil's review on the SD10....
An excellent interpolation algorithm depends on data integrity. As
long as there was "sufficient" resolution to properly define
boundaries of fine detail, there is no practical limit to digital
enlargement potential. How much resolution is "required" for this
depends on the detail in the subject and frame geography. A wide
angle hyperfocal landscape has incredible detail requirements. A
featureless piece of unicolor plastic filling the frame has nearly
no resolution requirements. A 2 megapixel digital capture of such a
subject could be enlarged to flawless bill board size. A 35mm film
capture of the same subject will still be overwhelmed by grain at
somewhat over 16x24 inches enlargement.
There are simply too many variables to make absolute statements
about digital/film equivalencies, but my comments were about
general photographic equivalencies. There are always exceptions on
either side.
Best regards,
Lin