> I took a long series of shots at different ISOs of the same complex, multi
> colored subject, same lens, same light, on a
> tripod, with E-M5 and GX7. There just isn't a noticeable difference in image
> brightness with the same exposure on both
> bodies. I believe Ctein is right, and the DXOMark ISOs are an artificial
> construct from the far end of the histogram and
> don't reflect the middle of the curve.
Which is what I've been trying to say, but evidently rather clumsily.
More clumsy to follow...
> But it's not really so quick. I would be looking at at least a few hours to
> get an idea. Different light and subjects,
> not just ISO, make a difference.
When studying calibrated targets under matched lighting (strobe in my
case), I find that most of the cameras tested match very closely.
Close enough to almost be irrelevant. But when comparing images in
real life, some cameras just handle things differently. In some
lighting conditions (golden hour), the DMC-L1 goes nooculear in the
reds. The E-1 held them without going into orbit. Yet, the calibrated
targets showed no meaningful difference. My Zone-10 test fixture would
definitely show differences, though. Why is that? I believe it has to
do with the fact that colors in real life aren't made up of just a
half dozen pigments or less.
Another factor, which Ctein and Moose brought up, is the natural
"gamma" of the sensor and where the manufacturer chooses to place the
mid-tone. I maintained, from the very beginning, that the E-1 sensor's
gamma matched the natural world better and didn't need as much bit
bending as the same-era CMOS sensors from Canon. You can make anything
"match" just by moving bits around, but something has to give
somewhere. It's like comparing colorspaces--in order to get a wider
colorspace you've got to steal bits from somewhere. All fine and
everything unless you really need those bits where they were. "Bench
Bob" here on the list (Bob, are you still here?) shoots with
medium-format, Leicas and Canons. It's fascinating that with his body
of work that I (and others) can identify the types of cameras used in
most of his shots because of the color/contrast/saturation. Not that
the Canon images are "bad", but in comparison, for the stuff Bob
shoots around the Sierras, the CCD-based cameras just created a more
pleasing image. Why is that?
This Curves/ISO/Midpoint/Gamma is something I and others have been
speaking about for 10 years now, but have been relegated to the cookoo
camp because the concept of "Raw is not Raw" flies in the face of what
we want to believe. The fact is that for the most part "Raw IS Raw",
and the Raw Converters require a different curves (or gamma)
adjustment for each type of camera. Adobe (and others) is doing all
this behind the scenes and under the hood so we don't see it. But it
is there. Trying to convince somebody that curves adjustments are
happening by the raw converter when your sliders are all set at "0" is
a fool's errand. You just want to ask people "what do you think the
"profiles" provided by the camera manufacturers actually do?
It will be VERY interesting to see how this new CMOS sensor by
Phase-One does. I wonder if SONY had to really work hard to get the
CMOS sensor to match the natural response of a CCD sensor.
--
Ken Norton
ken@xxxxxxxxxxx
http://www.zone-10.com
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|