On 3/2/2012 11:37 AM, Chuck Norcutt wrote:
> Theory says a 4/3 camera starts to exhibit diffraction effects (even
> with a perfect lens) at about f/5.6. I say "about" since diffraction is
> frequency dependent and varies with color. f/5.6 fits the middle of the
> green channel, blue has a tighter spec and red a looser one. If Moose
> can see distinct softening at f/22 on 35mm then that probably does
> equate to a roughly equal effect on 4/3 at f/11.
I should make it clear that my comments aren't directly related to standard DOF
calculations
The standard calculations make some assumptions about image display size,
viewing distance and visual acuity.
On-line sites like Wikipedia <http://en.wikipedia.org/wiki/Depth_of_field> and
Cambridge in Color
<http://www.cambridgeincolour.com/tutorials/depth-of-field.htm> have excellent
articles on the subject - except they
don't address the source assumptions. Luminous Landscape does spell out the
assumptions used and their generic sources.
<http://www.luminous-landscape.com/tutorials/understanding-series/dof.shtml>
"*Definition: *"/A group of photographers sitting around trying to understand
Depth of Field/". — Just kidding.
You can't understand /Depth of Field /until you understand COC (/Circle of
Confusion/). The human eye has a finite
ability to see fine detail. This is generally accepted as being *1' (minute) of
arc*. Translating this to the practical
world, this means that at a normal reading distance the smallest object that a
person with perfect eyesight, under ideal
conditions can see is *1/16mm *in size. If you place two dots smaller than this
next to each other they will appear to
be just one dot.
The photographic industry has generally found though that this is too fine a
parameter, and long ago settled on *1/6th*
of a millimeter as the smallest point that can be clearly discerned by the
average person under normal conditions.
Expressed as a decimal, 1/6th of a millimeter equals *0.1667mm*.
Now, the camera industry figures for the purposes of calculating depth of field
(and therefore /Circle of Confusion/)
that an image is typically enlarged 5X from the negative to a print. This would
mean roughly a /*5X7"*/ print from 35mm.
So since*0.1667/5 = 0.0333*, this is the /Circle of Confusion/ that many 35mm
lens manufacturers use when establishing
their depth of field tables and lens markings."
Notice how fuzzy even this definitional material is. What's "perfect eyesight",
20/20? With visual acuity of 20/10, my
eagle eye can presumably resolve 1/2 min. of arc. Many others, particularly
among the older members of this list, may
only resolve 2 or 3 mins. of arc.
And what's 'normal reading distance"? The truth is that, when actually looking
at the whole image, not nit picking, we
tend to hold a 5x7 closer than an 8x10, than a 16x20. But probably not at
double the distance with each increase in
print size, which is the assumption of DOF calculations. So in the real world,
greater "DOF" is required for larger
prints in many situations.
So we've got the scientific vision researchers saying 1 min. of arc, but "The
photographic industry has generally found
though that this is too fine a parameter, and long ago settled on *1/6th* of a
millimeter as the smallest point that can
be clearly discerned by the average person under normal conditions." So they
are using almost 3 mins. of arc.
The obvious question is why that C of C basis? The answer is simple, they are
actually based on lots of human viewers
looking at lots of prints and what they thought was and wasn't sharp.
That's not a bad basis, hard to think what would be better for practical
picture taking. Still, it was long ago, using
B&W darkroom prints created using the cameras, lenses, films and papers of many
decades ago.
Would the results be different with today's high resolution digital imagers,
the best of today's lenses and color prints
from today's best ink jet printers? Before or after USM and/or deconvolution? I
tend to think so, but by how much, I
don't know. I think without arguing better/worse, we can agree that such images
just 'look' different, which probably
affects opinions about what is and isn't sharp enough.
So how to judge the dimensionless image file sitting one one's hard disk? The
way I've done that is to view test shots
at 100% to determine at what point a smaller aperture starts to make objects in
the focal plane less sharp. That doesn't
map in any direct way to standard DOF calculations.
So when I said the 135/4.5 starts to show diffraction effects at f22, I'm
really only talking about recent color
negative film scanned at 4000 dpi. I think they would hold up pretty well for
the 5D, although uncontrolled experience
suggests f16. if I were to repeat the tests. But with a 5DIII, I might well
decide that the effect was visible at f11,
maybe even f8.
How does any of this relate to an 8x10 print is not defined. Rather, the
lens/sensor limits are defined, which may be,
from a practical standpoint of getting the best image possible with one's
equipment, more useful than DOF calculations.
Dr. Alternate Focus
--
What if the Hokey Pokey *IS* what it's all about?
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|