I missed commenting on some recent threads about megapixels, resolution
and prints. But, surprise, surprise, I have some thoughts and opinions
on the subject.
---------------
Digital camerainfo.com says:
"The 5D does in fact offer more edge-to-edge resolution than the D200,
and the difference is not just academic. While the results plotted in
the charts above may appear as an insignificant difference between the
10.2 megapixel D200 and the 12.8 megapixel EOS 5D, when observing 8 x
10-inch prints made from each camera - held roughly a foot away - the
difference is clearly noticeable. The 5D produces sharper images with
more observable detail, even at 8 x 10. The results obviously become
more pronounced as the print size is increased."
http://www.digitalcamerainfo.com/content/Nikon-D200-vs-Canon-EOS-5D-Head-to-Head-Review-/Performance--Image-Quality.htm
Am I to believe this? Well, let's see, in their comparison, they draw
conclusions about sensor resolution using 50/1.4 lenses wide open. Right
there, they show themselves to be less than rigorous thinkers. It's well
known (and matches my personal tests) that the 5D out-resolves many
lenses, and I assume the same to be true of the 200D. So they have done
a test of the C vs. N 50/1.4 lenses, not a sensor test, and
misunderstood what they are doing.
Worse, because of the different sensor sizes and use of a test that
normalizes resolution to LPPH ratings, they have given the N lens a
handicap by enlarging its image more. Or have they handicapped the C
lens by including the weaker outer image area. Who knows, but whatever,
they haven't really tested any one component of either lens/sensor
system against its counterpart.
I'm not sure myself how meaningful the f8 tests are, or the tests on
dpreview. On dpreveiw, tests with the same lens show
resolution/sharpness differences between the 5D and the 1Ds II. Yet in
my own tests of macro lenses, I can easily see differences in
resolution/sharpness between very high quality lenses at their optimum
apertures and in the center of their fields. This leads me to believe
that this is one of those complex systems bandwidth thingies that AG
talks about where the bandwidth of either component cannot be exactly
determined from system bandwidth, but system performance is increased by
improvements in either component.
There's a lot of interesting and potentially useful info in this and
other on-line tests, but you have to keep your wits about you when
reading them to avoid being lead astray.
As to the print comparison, I'm sure Bob and others are right in saying
that effective sharpening was almost certainly very different for the
two prints. It's easy to forget that resolution is not simply sharpness.
The visual sharpness that humans perceive is some difficult to
understand and quantify combination of resolution and local contrast at
edges.
I'd bet that 16x20 prints from the two cameras processed to optimal
sharpness (Rather than some minimum, what's the point of that?), by the
same competent person for both, would be indistinguishable from each
other as to sharpness/detail without a loupe, and maybe even then.
Moose
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|