[Leica] Learning a lens

[Mike Johnston <michaeljohnston@ameritech.net>]

>>>Let me finish by saying
that while my photo track record (pictures published) is excellent only
because I have an excellent eye for composition, but, I am a real bozo
when
it comes to anything technical and that is why I am attracted to a
baseline
number<<<


Steve,
I find it poignant that you could say such a thing. You have the
misfortune to be "just" a talented photographer, but you don't have an
encyclopaedic knowledge of technique that would allow you to discern
meaningless distinctions that have very little indeed to do with
photography...how much better off would you be if you were a polymath
about phototechnica but couldn't take a picture worth a damn?

<g>

Okay, I'll quit it.

The problem with assigning a number to a lens is that when you take
pictorial pictures, you are concerned with pictorial effect. Today we
have any number of super-competent and affordable Japanese lenses
designed using MTF as the basic evaluation tool. They're all "good"
lenses but many photographers feel they are characterless and bland in
terms of pictorial effect. And then, consider the characteristics of a
lens that go into evaluating its performance. In a version of the old
joke about the blind man and the elephant, one tester might look only at
resolution at infinity. But say you take all your pictures from ten feet
and less. Does the test tell you anything? (We used to have the opposite
problem, in that photographers liked to test their lenses by
photographing sheets of newspaper taped to the wall. Which of course led
them to possibly different conclusions than if they had been testing the
lenses for resolution at infinity.) Another tester is concerned mainly
with distortion. He gives a high mark to a wide-angle lens that renders
a straight line straight at the edge of the frame. But say you're a John
Brownlow / Johnny Deadman type shooter, doing quick grabshots out on the
street at night. How important is distortion to you then? Another tester
loves color transmission, saturation, and discrimination
characteristics, and give much weight to those things. But you
photograph exclusively in black-and-white. Another tester (_Consumer
Reports_? <s>) is much concerned with value--features for the dollar.
But photography is your life, and you have lots of money; the difference
between a $300 lens an a $900 lens is essentially meaningless to you.
Another tester loves wide-open performance, and downgrades fast SLR
lenses severely based on poor full-aperture performance, reasoning that
no one would buy a fast lens unless they intended to use it wide open.
But perhaps you prefer the wide aperture for focusing ease, and never
use full aperture. Another tester is much pre-occupied with ultra-fine
detail. But you shoot on P3200 film, and fine detail below a certain
level is typically lost on the film. Yet another tester feels that a
lens should be an all-purpose tool, and gives high marks to consistency
up and down the aperture range and from near to far focus. But say you
are a real-estate photographer, and you always shoot at f/8 in daylight
from roughly the same distance, parked at streetside.

And of course you could turn all these examples around the other way. I
could also think up as many more as appear above.

One of my favorites is size and weight. I have a friend--this is a true
story--who, a dozen years ago, researched 135mm lenses thoroughly and
bought the AIS Nikkor 135mm f/2, which he had determined after
fastidious research to be "the best" (9.6? 9.8? <g>). But the damn thing
was such a brick he never took it with him anywhere. Know how many times
he used it? Twice. I am not making this up. So he obviously ignored one
very important characteristic of the lenses he was evaluating.

Do you see what I'm saying? Even in subjective evaluation you can be at
cross-purposes with the tester or reviewer. Portrait lenses are a fine
example of this. A tester light like a lens that is sharp and
technically good, but your clients might like a lens that has fine
subjective qualities but is not so very sharp! (I always valued the 90mm
Summicron-R because it is not so sharp wide open but still yields
fine-looking pictures--perfect for portraiture.)

And then we get into properties that are essentially unmeasurable. How
well does a particular lens "match" to a certain type of film? What are
its flare characteristics? (I have a battery of different tests I use to
appraise flare performance, since I think it's an important aspect of
lens performance in practice--and all of it is unmeasurable.)

I personally have a problem in that many of my pictures have large
amounts of out-of-focus area, and I am charmed or repulsed by the
quality of the  _bokeh_ or blur. It's not a trivial or an arcane
issue--it's all over many of my pictures, right there for the eye to
see. And although most Japanese magazine tests make an effort to
describe and demonstrate the quality of the blur, American magazine lens
tests ignore it entirely. So I am one photographer for whom lens tests
simply fail to describe what I most need to know about a lens. The
solution? I have to look for myself, with my own eyes. (It's why I'm
often eager to see multiple samples of what someone else has done with a
particular lens--I can often get a basic handle on the _bokeh_ of a lens
without having to use it myself.)

Arthur Kramer once quoted to me a Leica lens designer he interviewed for
_Modern Photography_. Unfortunately, Arthur could never give me the
exact quote. (So yes, literalists, I paraphrase, and thus do violence to
strict journalistic and scholarly standards--please bite me). "All lens
tests are shortcuts," he said. And they definitely are. They may tell
you some things you need to know. They also may not tell you what you
most need to know.

The way to learn a lens is to use it, and study what it does. A lens
that does what you want, whether it is the worst lens in the world or
the best, is the lens to have; and a lens that earns the highest "grade"
in a published test _may_ be, but is not _necessarily_, the right lens
for you.

So tell me, is a lens that earns a "9.6" really "better" than a lens
that earns a "9.5"? That's like saying that you like wine better than
Ted Grant likes scotch or that red is better than blue. It's just
meaningless.

The ONE way in which an ultimate grade ranking has merit is that it
might help establish bragging rights for status-seekers. "Ooh, you have
the Aspherilux Astro-Noctilar. I see that lens got a 10.1 on a ten scale
from _Zippon Camera_ magazine last month. You must be a discerning and
serious photographer indeed!"

Pshaw.

- --Mike