Re: [Leica] SHARPNESS

["Joe Stephenson" <joeleica@email.msn.com>]

Dear Mr. Schiemier,
An interesting post. It made me get out my Ansel Adams book ("The Camera")
and look up 'sharpness.' He starts by noting that content is far more
important than the impression of sharpness in a print. There is no way to
measure 'sharpness' in a lens. Acutance, yes. Resolving power, also.
I find the "proven facts" that you mention rather unimportant. There are a
variety of subjective factors that are of greater importance, in my opinion.
Art (what I try to do) is subjective. If I photographed test charts, I would
consider test results of taking photos of test charts important. But I
don't. I take photos of people and things in a variety of situations and
settings. My experience has convinced me that Leica equipment is
particularly good for this purpose, so I keep using it. I do use a 4X5 and
an slr from time to time, when needed, but mostly I use my 35 and 50mm
Sumicrons and am quite satisified with the results.
And, as most lugger's realized, the lens is just the starting point. The
film, processing, technique, etc. all have a great influence on the final
print or slide. For my part, I'm glad that photography is not as cut and
dried as discussions of test charts, "sharpness," and the like suggest. I
suppose the very best quality would be obtained with a process camera, but
it hardly seems like a kit that most of us would find useful.
Sincerely,
Joe Stephenson
=============
Mr. Schiemier wrote:
Have read with great interest the continuing threads regarding ‘sharpness’
of
lenses, as well diversions spawned from same.
When I first became interested in photography a mentor taught me to evaluate
lenses not on ‘reputation’ but on proven fact (results).  In all this time,
in
all the directions imaging has taken me, that standard of proof has been a
constant.  It does not change with whimsy or get modified by marketing
jingoism.  [Although I’ll admit to falling for some whoppers!]
One of the miracles of modern photographic technology sits before you in
your
computer.  LSI and VLSI (large scale integration) relies on an exacting
science of reproducing images perfectly in making ‘road maps’ for printed
circuits (to the power of 10000x+).  Part of the process includes
determining
resolving power of lenses utilizing a known base of performance.
Also, many of  you probably know someone who has had angiography done, it’s
a
pretty common procedure these days.  In setting up an imaging system a
certified standard for resolution (called ‘line pairs’) is used by
technicians
to establish optimal performance.  Observing secondary arterial stenosis
needs
to be pretty exact.
Subjective qualifiers about specific camera lenses which have no basis, no
constant, and no reference are bandied here on LUG as easily as shelling
nuts.
It is all homogenized by personal preference, and -in consideration of the
very diverse results general photography can produce- infinitely variable.
While Joe over here may refer to an image he took of a stand of trees as
‘sharp as a tack’, and Bill over there may refer to a shot he got of a
peacock
in full strut as ‘sharp as a tack’-  those two tacks are as different as
night
and day.
In the final analysis, neither may be as sharp as they could have been.
The ability of one viewer over another to determine sharpness of a resultant
print or transparency is, at best, obscure.  Another person needs to know
the
observers ability to discern a difference first, so that a statement of
acuity
can be taken at face value.
Joe shot his trees with a 135mm from 175 feet, and the peacock Bill got was
done with a 50mm at five feet;  but what f-stop, what shutter speed,
tripod?,
what film, what processing, what enlarging lens, what paper??  Heck, what
time
of day was it taken??
Because Joe is a pro, and we know it, do we take what he said as gospel?
Bill, a rank amateur and unknown entity, should have his comments
discounted-
he couldn’t possibly know what a tack is.
What if both Joe & Bill could benefit from a sharing of knowledge, thereby
enhancing their results?  Ultimately getting that image REALLY ‘sharp as a
tack’ {with the tools at hand.}
I’ve always used a linearity chart (from film and video applications) for
determining optimum aperture and resolving power of lenses [sometimes used
in
collimators].  I’d think using an item like that could help define the
playing
field for our future reference.
I’ve done some research in the area and found Ed Romney has an inexpensive
48
page periodical called “Evaluating and Testing Photographic Lenses” ($9.00
USD), which includes a reprint of the Reckmeyer article (c. 1934) and the
National Bureau of Standards lens chart with description and instructions.
His URL is http://www.edromney.com/, tel. no. 864/597-1882.  [I neither work
for, nor obtain any financial or other remuneration of any kind for this
recommendation.]
Don’t care much that the Reckmeyer article is sixty years old, nor what
anyone
in particular may think of, or about, Mr. Romney- it is, simply, the NBS
chart
and instructions for its use that is of primary concern here.
Until a standard is set for the mythical lens bar, any such commentary will
always be open for derision.  No matter what you say, or who you are, it is
all subjective conjecture without a clinical analysis.
Use the chart to measure your lens, post the results (complete with all
factors as applicable), and then we’ll be able to really know what ‘sharp’
is,
or isn't.
In the meantime it’s all a bunch of hearsay, and should be taken with a
modicum of suspicion.
For instance; quoting here from a publication I have;  “Optimum aperture is
the lens stop at which lens performance is at its best.  It is judged above
all according to contrast and uniform sharpness distribution.”   Yet, I saw
few, if any, references to optimum aperture in any of the threads.  Curious,
wouldn’t optimum aperture be a significant factor in determining how sharp a
lens is at other apertures?  Or, in the least, how far off the mark of
perfection a particular image may have been -for that lens?
Think of it like this;  there are some six or seven hundred folks subscribed
to LUG (I don’t know, I’m guessing).  Most of them have (or want) Leica
cameras and lenses.  Factor the cumulative gear investment quotient (GIQ)
represented by the LUGites (quick calculator check makes it, on best
guesstimate, $2.5M USD), wouldn’t you think it’d be worth our while to
establish exacting standards for lens performance?
What value a discovery the 90 Elmar is good from f-2.8 to 8, but really
great
at f-11, then falls off significantly from there?... I don’t know this, I’m
making it up, speaking rhetorically.
Rather than relying on relative quantification why not work together to
establish a standard (or use one already established- NBS resolution test
chart), and post the results?  In that way we wouldn’t have to use
adjectives
like ‘good’, ‘great’, ‘fantabulous’, or whatever- it would be a mathematical
reference- like 80 lines per millimeter at one enlargement to the second
power
(2X).
Everybody gets better, gray area fades and contrast gets brighter.