> I don't understand how the Kodak sensors don't lend themselves to
> "Expose to the Right". There's no metering involved other than running
> the histogram to or near the right hand edge of the graph. Are you
> saying that those Kodak sensors can't handle a brightness level near
> maximum? Or that your histogram lies, or... what is it?
>
> I find the statement suspect.
The Kodak sensors as used in the Olympus cameras utilize two different
green pixels. They have slightly different color and brightness
curves. Where most cameras mix the pixels of one each color at each
intersection, the Olympus cameras require four pixels to be mixed.
Unfortunately, the four pixels do not reach saturation at the same
point.
This causes major color shifts in the top stop of the range. If you
made sure that absolutely none of the four color pixels reached
clipping then there would be no shifts, but the histograms shown in
the camera as well as in a RAW converter are derived and not literal.
On the Zone-10 website, is a really good illustration of this. On the
left side, look for "Product Reviews - Olympus E-1". Click on that and
then on the "Olympus E-1 Dynamic Range Test". Go to page 2.
The test photograph is 3 stops over-exposed. I did this to make sure
that I was well into the world of clipping. The second image on the
page is the same image converted with highlight recovery and exposure
compensation to bring it back down to a normalized image.
Inotherwords, this image is pulled 3 stops with maximum bit recovery.
On the Kodak test target, look at column 13--Cyan. H13 is starting to
lose color, but G13 and F13 are gray. In colunn 17 is red. H17 went
pink, G17 and F13 went yellow. Column 18--green lost all color in G
and F. But the Blues held one step farther. Meanwhile, the other two
derived colors of magenta and yellow both topped out at I14 and I15.
As you can see by the grayscale, highlight recovery pulled back four
steps from 10 to 6.
On Page 3 of that article you can see what happens when you push the
exposure back up. Now, granted the image is quite noisy, but for
crying out loud, this is SEVEN (7) stops of boost!!!! The second one
is six.
So, what I'm saying is that when it comes to color integrity, the
Kodak-Olympus sensors don't handle the top very well and anytime you
use highlight recovery you will get color shifts. It's actually kinda
like Kodachrome in that manner.
Granted, the E-1 isn't the cleanest sensor around and Olympus adds
dithering noise across the entire image, so the acceptability of a
severally pushed image might be beyond what you can stomach, but when
it comes to color integrity, I'll gladly push an E-1 one stop than
pull one stop and have to fight the highlights. Because of the added
dithering noise, there is not as much noise reduction to be gained
through pulling the exposure.
These controlled tests, as illustrated in this article also have
proven themselves out through empirical testing and usage.
The E-1 is specifically a different animal than other Olympus cameras
that followed. Even the E-300 and E-500 are a bit different because
the times, they were a changing. The E-1 was introduced as a "Film
Replacement". All other cameras following were "Digital Replacement".
The E-1 was really the only Olympus camera designed to be used like a
film camera.
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/
|