Stefan,
In a message dated 9/12/2005 1:52:40 P.M. Pacific Daylight Time,
email@hidden writes:
<<I really doubt there will be such a thing as an emissive olor
checker.
Why ? Well think yourself. As soon as you use emissive lighting
values to
calibrate a
camera chip, all you really do is linearize to a
spectral response curve of
the chip.>>
If the spectral sensitivities and linearity characteristics of the camera
are known, then this is the fundamental information required to construct an
optimum camera profile. The best way to get this data is with a monochromator.
The LED target is essentially an abridged (variable wavelength variable
bandpass) monochromator.
<<You do not collect a single bit of information about the
illuminant of the
actual photographed scene.>>
Right. To get the viewing conditions of the capture one either has to rely
on the image meta data (white point, luminance, etc) recorded with the image, or
analyze the image data itself. Usually the former is enough.
<<Without a spectral
response curve(can be done once in the
factory),
metameric information and overall whitepoint calibration of the
actual scene
you will get nowhere>>
It would be great if this information was made available, but its generally
not. So the emissive target provides an opportunity to create an
optimal camera/lens specific profile for the photographer's particular
setup.
<<why the spectral response has to be measured inside
the camera.
I´m pretty shure that the technical (spectral response) specs are well known to
the chip manufacturer>>
Each camera is somewhat different. And because of the optical properties of
a camera chamber (flare, f number, etc), measuring the detector is just not
enough. Most camera manufacturers, if they have in-camera measurements at all,
are loathe to share this information as they consider it to be highly
proprietary.
<<any other
system with interchangeable lenses is going to have
different values for
every other lens mounted>>
Right, but a well determined profile and processing flow can be
surprisingly robust. These errors can be rather small in the overall scheme of
things. In any case, the emissive target (and presumably smart profiling
software) could allow the photog to profile each camera/lens combination if he
feels the need.
Thanks,
Eric
Walowit
Tahoe