For the sake of argument (or non-argument), I'll agree with you,
Chuck. We'll make the assumption (which may or may not be true) that I
misfocused for the original test. Since I didn't have usaf charts in
the scene, I was at the mercy of the details in the scene--which in
this case is the grass. If I did misfocus, this does point out just
how haphazard our focusing really is and we're just getting it close
or maybe not-so-close most of the time. I did think my "real world"
landscape shot with ultra-detail was a good indicator of lens
performance, though.
As to the 5D Mk2's live-view--it is my understanding that this camera
doesn't use all pixels for live-view and skips pixels. This, again,
would get you in the neighborhood but is anything, but exact. To
differenciate a focus point at 5 vs. 6 meters with a "normal" focal
length, the difference with most lenses is a whisper of a nudge. Like
C.H. says, 0.5mm or so. The problem is that wide-open, there is
usually just enough CA or other issues with the lens which prevents
precise focusing to this level. As mentioned, stopping down the 24mm
F2.8 lens to F4 acually make precise focusing more feasable. The
line-pairs on the usaf lit up like a Christmas Tree when in focus. I
could conclude, therefore, that it is possible the L1 has a superior
live-view than the 5Dmk2.
Here's the rub. Let's say we adjust the focus point to, say, 3 meters.
This is a nice point close enough to the camera that focusing should
be easier. But wait! What is the accuracy of focus? If you can't
differenciate between 5 and 6 meters, then at 3 meters your accuracy
is probably no closer than 5 cm. This would throw off the test too.
Therefore, I can only conclude that it is impossible to focus any lens
to any degree which would allow us to compare DoF, Bokeh or any other
trait. We're pretty much hosed. We're obviously deep inside the margin
of error.
BTW, a nifty trick for focusing an SLR in the dark. Shine a laser
pointer through the viewfinder. When the dot on your subject is sharp
you are in focus. :)
So, here we are--a lens conundrum. How to test these lenses in a way
that either proves or disproves the theory. Empiracal evidence is no
evidence at all. The emperical evidence shows that there is something
beyond the simple algebraic calculations which assume simple optical
traits. But Jan, Chuck and I'm sure others are screaming at their
computers right now saying "just trust the numbers".
I don't think I've ever claimed that one lens has more DoF than
another, I've stated pretty clearly, I hope, that one lens has an
apparant DoF greater than another. What is causing this, I don't know.
That's what I'm trying to figure out. But then there is a side
question of "What is DoF?" If DoF calculations are based on a
subjective analysis of a printed image and the "apparant" increase in
DoF is visible in the printed image doesn't that therefore defeat the
existing formula for calculating DoF? Again, if the formula is
subjective analysis based and the increase is also subjective, then
the formula is wrong.
The formula is wrong or defeated by diffraction. Also, how does the
formula change in relation to the position of the nodal points to the
film plane? Most formulas use the front nodal point. One reason why I
selected such a long distance was to cover up the variations in the
nodal point positioning. Closer focusing will greatly alter the
position of this front nodal point to the level that it will mess with
any usable conclusions. The 50mm tests were done closely enough that
the change in front nodal point was enough to change subject size.
Recently I did a test of 50mm focal lengths. It was quite amazing the
differences between the size and shape of the bokeh at identical
apertures. Some of the lenses with aspherical elements folded the OOF
highlights in on themselves. Instead of a symmetrical growth pattern,
it instead remained small. If you could develop a lens with multiple
aspherical elements which successfully folded the OOF highlights back
to a point then you'd have a deep-focus lens. In fact, check this out:
http://www.freepatentsonline.com/1627892.pdf
There is obviously then a way to defeat the DoF calculations with
aspherical elements and other nifty tricks. A cross-section and ray
analysis of some of our modern lenses may show deep-focus lens traits.
But unless we can figure out how to accurately focus our lenses, we're
pretty much hosed in ever knowing.
AG
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|