Moosie wrote:
>I'm not in the least bit interested in mimicking any film with digital.
>Digital is, as you have pointed out, different from film. I would say
>not better nor worse, in any ultimate sense, depending on what one
>values, just different. I'm interested in how any sensor system aids or
>gets in the way of creating the image I see in my mind when viewing the
>subject.
Which begs the question of do you REALLY know what your digital sensor
system is actually doing? Also, there is an assumption that the IT8 target
is the end-all, be-all of color calibration. This is only somewhat true.
For example, on DPR, they use Gretag MacBeth ColorChecker charts. Does
anybody actually think that those few squares will correctly identify how a
sensor responds to different lighting conditions and materials? Shall I
bring up the touchy subject of the Leica M8 and it's ability to render some
black fabrics a nice beautiful magenta? The pigments or dyes used on
reflective test charts are NOT representative of actual fabrics or colors in
nature. The blue-sky squares on the Gretag MacBeth is a color
representation "in print" ie, reflective, of something that is actually not
only translucent but contains wavelengths outside the standard 380-700nm
range of commonly accepted range of visual accuity. So, the various test
charts, pigment or dye-based, are limited to ONLY the range of reflected
wavelengths of the chart. The black square on the Gretag MacBeth ColorCheck
is perfectly neutral and black in the M8 tests. Without realworld testing,
we really have no clue what is going on--this is probably how Leica and
Kodak managed to totally screw up something that should have been a "duh"
item.
And I wonder how many manufacters' camera testers actually test for this or
is the scope of testing limited to usabilty issues? Are camera companies
still stuck in the old ages of testing the cameras or are have they hired
the necessary talent from the film manufacturers to properly test
color/brightness responses in nature? (when Konica and Minolta merged, the
Konica film engineers built all new color profiles which were less accurate
on the charts, but not only more accurate in nature, but more pleasing to
the eye). Kodak and Fuji took years of fine-tune testing before they ever
released a color film or even a film update. Even then, they didn't always
get it right, but they took the effort over a great period of time than an
abbreviated test schedule with a half-dozen prototype sensors in a computer
lab. OBVIOUSLY Leica never real-world tested the M8's sensor. Didn't
happen, guaranteed! I honestly believe that Leica wasn't unusual in
this--others were just a bit luckier. The trade-secret issues with digital
cameras and sensors are so closely guarded that proper testing with mules
just isn't done.
I asked the question to get people starting to think about this. Please
don't be too quick in blowing off the question with the attitude of "Digital
is different, this doesn't apply to me." Oh, really? Would you care to
tell me which digital cameras match each other in color responses? And not
just on an IT8 target, but with realworld conditions? Do the fabrics
actually match? Can you shoot the 5D and a 1DsMk3 side by side and get
exactly the same results?
Purple and yellow flowers anyone?
What if you change light-sources? Did you know that RGB LED light-sources
used in stage lighting will render fabric colors totally differently than
normal stage lighting which has a huge IR content? Would you like to
through a bit of florescent lighting into the equation? Which florescent
tubes are you talking about? An ICC Profile ain't going to help you there!
We know that Fuji Reala responds to mixed lighting unlike any other film.
Does your digital camera handle mixed lighting like Reala or more like
Kodachrome?
Can we quantify in layman's terms what happens to green vegetation in a
photograph? Again, the Gretag Macbeth chart will not necessarily be an
accurate representation of the UV and IR reflectances in leaves. No two
films responded to this the same way--that's why we had preferences of
Kodachrome over Fujichrome or some people preferred the way Ektachrome
captured the scenes. I found that shooting the northern forests of
Michigan, I couldn't use Kodachrome or Ektachrome because it rendered tree
bark a grayish white color, wheras the Fujichromes would render tree bark
brownish. Same tree, same lighting condition, same white-balance, different
colors of bark. Can you tell me how your digital camera is going to render
the tree bark?
(short answer is the typical DC will capture a high blue-content which is
"corrected" through mammoth WB swings, rendering the surrounding vegetation
too warm)
Day-Glo is another adventure in futility. Day-Glo pigments and dyes absorb
some wavelengths and reemit the light on other wavelengths. The human eye
sees Day-Glo totally different than film or sensors. Have you photographed
a florescent yellow-green safety vest or traffic sign? Can you tell me how
your camera/film will respond to it? The short answer is that no two films
or sensor types will come up with the same color in capture. Some will turn
the yellow-green fabric orange, while others will be yellow and others will
be green. In some cases, it is even possible for it to have a substantial
blue content to it. These are NOT white-balance issues and also this is
absolutely nothing you can do in an ICC Provile to fix it. Either it got
captured or it didn't. This is not a RAW vs JPEG issue, as this is the
electro-optical or chemical-optical interface.
>But then I'm the guy who uses icc profiles on films to negate the color
>and tonal distribution differences between them, so what would I know?
Again, this is partially applicable. You have created ICC profiles for how
the films responded to the specific pigments/dyes used in the reflective
chart under the specific lighting conditions subjected to it during the test
exposure. You'll get closer, but you've neglected the real-world conditions
affecting the initial optical-chemical interferface.
> On the other hand, David's Kodachrome project and your and others'
> comments about B&W conversion make an interesting thought project.
That B&W conversion project is rather exciting, but is frought with issues
totally related to the above subject. Black and White films respond to the
near UV and near IR wavelengths in totally different ways than color films
and/or digital sensors. We can mimick the results in some cases, but it
still doesn't make it the same. I have one digital camera that will render
a distinctly Tri-X look to the in-camera B&W images, but the same identical
shot with my other digital camera yields a completely different look. Yes,
we can profile the tonal response curves quite a bit, but this doesn't
address how the film and sensors respond to the initial colors and UV/IR
componants. This is why everybody had a favorite B&W film. Tonal curves were
only part of the "picture", the ability to lift or lower certain items (like
skintones) are the distinct signature characteristics.
Again, I ask you--what film does your digital camera mimick the most?
Ken Norton
ken@xxxxxxxxxxx
http://www.zone-10.com
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|