At the "per-pixel" level, my ancient E-1 and the 6D are a pretty close
match. Sharpness in the E-1 is variable depending on the color and
brightness level, where it's more uniform with the 6D. That has less
to do with the sensor and more to do with the converter algorithm. The
Adobe algorithm I'm using is doing a 3-pixel merge on the 6D (nearest
neighbor, creating a phantom pixel at the intersection), where the E-1
is a 4-pixel merge with replacement RGB values for the monochrome
12-bit value. The E-1 is forced to the 4-pixel merge because of the
two-color green sensels. The 6D sensels are sensitivity balanced for
the clip point, where the E-1 sensels are sensitivity balanced for the
midtone. The 6D is far more uniform in sharpness from shadow to
highlight than previous generations of Canon DSLRs and actually is a
pretty good match to the E-1 in shadows where CMOS sensors have tended
to produce mush in the shadows. (The E-1 tends to produce mush in the
highlights).
I mention all that because it's really hard to do a direct comparison
of these two cameras in regards to resolution because there are so
many other variables that affect the resolution unless you are
comparing with a fixed monochrome line-bar lens-resolution test chart
like DPReview uses. You can have a camera that nails that test but
fails miserably in real-world images. This is why images from the
Foveon sensor out perform everything else out there in the "real
world" in regards to uniform resolution regardless of color and
brightness.
So, back to the E-1 vs the 6D. The pixel-pitch is very close to the
same. That means that a 4/3 crop from the 6D will be a very close
match to a native 4/3 image from the E-1. And in real-world shooting,
it does some of the time, subject to the color/brightness variables
mentioned above. This is a situation where pixel-pitch is a nice guide
or reference point, but means less in real-world shooting. Just
increasing the surface area (6D) should theoretically be equal to a
multi-frame pano stitch from the E-1. Sometimes it is, sometimes it
isn't.
Increasing pixel density should theoretically improve resolution. Per
lab tests with monochrome test charts, that would be true. But let's
look at the 7MP Panasonic L1 and compare it to the 10MP Olympus E-3.
In real-world shooting, the L1 out resolves the E-3 almost every time.
(is it possible that this sensor in the L1 is the hidden gem of the
4/3 world?).
To the subject of diffraction, I have personally found that pretty
much without exception that the diffraction algorithm is off by a full
stop when shooting digital. This is because of the AA filter and the
converter itself. I have also found that with few exceptions, the
diffraction algorithm is EFFECTIVELY off by two stops with digital
because of sharpening tools available. The diffraction Airy Disk is
not a fixed Gaussian Blur, but a wavelength-sensitive reduction in
contrast at the pixel level.
AG Schnozz
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|