Re: camera evaluation brought up image processing/color space questions
Re: camera evaluation brought up image processing/color space questions
- Subject: Re: camera evaluation brought up image processing/color space questions
- From: email@hidden
- Date: Fri, 10 Nov 2000 13:51:49 EST
In a message dated 11/10/00 11:21:14 AM, email@hidden writes:
>
Hello, Does anyone here have experience using the new Jenoptik Eyelike
>
MF
>
Scan digital back? We just had one demonstrated in our studio and I am
>
would like some feedback on its software image processing. Specifically
>
the
>
effect it has on the gamut of the capture. Even if no one has used this
>
camera, perhaps some of the resident color gurus have an opinion to share
>
on my observations. Photographing a kodak Q13 with no curves adjustment
>
(but a grey balance) reveals that the camera default is a 1.8 gamma curve.
>
This is the same as PhaseOne's default, although PhaseOne offers a 1.0
>
option if you truly desire raw data. This is not a problem as it is a good
>
starting point and in fact exactly what we desire for our archival master
>
images.
So far so good...
What does raise an alarm is that captures opened in Photoshop in
>
Adobe RGB 1998 are grossly over saturated. ColorMatch is a much better
>
match. Am I correct in seeing this as an indication that the data has been
>
processed for a relatively small gamut color space?
That would be the implication...
If so, is the capture
>
gamut of this device limited?
Capture devices have no gamut in the strict sense, but you can say that the
capture capabilities of hte device are being unnecessarily limited to a
convenient colorspace gamut...
Assuming an accurate camera profile is
>
created, could the capture data be safely transferred to a wider gamut
>
space?
Only if they haven't keyholed the process to an sRGB or ColorMatch space
somewhere that you do not have control of...
In my understanding you can't move data from a small gamut space
>
to
>
a large gamut space and magically replace what might have been compressed.
If you have broad gamut info moved to a smaller space by perceptual
compression it is possible to regain the more saturated colors by applying a
wider gamut profile to the image which will stretch them out again; this is
not the same as converting to the wider profile, which would simply place the
colors into the larger space at their current reduced gamut. However if the
file was clipped to the smaller gamut (colorimetric conversion, as opposed to
perceptual) then it will have lost the distinctions between in- and out- of
gamut colors, and no reversal is possible. This all assumes high bit files to
avoid banding and other such artifacting...
But simply opening the file in AdobeRGB is the equivalent of this, and you
find that this offers an oversaturated result, so there doesn't appear to be
a simple fix to your problem, and you will need to hope that it is possible
to profile the device without such gamut limits.
C. David Tobie
Design Cooperative
email@hidden