Hermann-Josef,
You are on the right general track, your understanding is correct.
Here is more food for thought:
• Your creative intent when you scan is best handled by delivering the
image into a working space that's covered by your calibrated display. And
in your case it is. and will be if you convert to sRGB. Most of us assume
that the display is our primary window for determining the proper
rendering. It doesn't have to be, but you would have to reason through
another choice, which by common sense is only productive if you have the
reasons :)
My understanding of the way a scanner works is that there's no
critical intermediate store of sensor data, so nothing like camera raw to
preserve. I have also read there is no idea of a scanner-native space
beyond the internal choices used to convert the devices spectral response
traits into RGB data.
There very much is the idea of scanner profiling, which can help get
film response into a working space with good fidelity. This is a worthwhile
pursuit. It also maybe can be deferred if you run the scanner in a very
repeatable way. So the idea of a "scanner space" can emerge in the sense of
a list of repeatable capture settings that can guide future profile
creation and application to film captured the same way. You have to apply
the necessary discipline to make it work, and defer artistic rendering to a
distinct step after capture.
There is a possible trade-off — which is probably so incremental as to
be ignored — of sacrificing some scanner sensitivity that could make the
very most of a problematic image for the consistency and range requires to
capture all images.
• Re general presentation vs archiving: today you still cannot count on
color mgmt to cover for you. As an example, I am working w blogging SW that
renders images into thumbnails for various purposes of UI and presentation
and the auto generation doesn't always respect the image profile. OTOH
most devices will respect profiles, so it's always something. The safest
way to go is sRGB image state.
Adobe RGB only offers a green advantage over sRGB, which in practice
becomes cyan limit, because worldly greens are well covered by sRGB. My
actual world is more likely to include reds outside of sRGB.
Display P3 is a better compromise for any advantage of extended gamut
and balanced degradation if image space gets ignored by primitive
presentation. So if you accept the hazard of improper rendering, Adobe RGB
is a less desirable wider gamut. And if you want to avoid hazard then use
sRGB so whole thing is moot. Adobe RGB is a very awkward space.
In practice, most of the world fits into sRGB and the parts that don't
are not memory colors, so hazard of improper image state rendering far
outweigh hazards of gamut limits
HTH
/wire
On Sat, Dec 5, 2020 at 07:56 Hermann-Josef Röser <
posts@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
The linear scans are, of course. But the final product should be in
AdobeRGB or sRGB.
Hermann-Josef
---------- Forwarded message ----------
From: Mark Wilhelmsson <markanini@xxxxxxxxx>
To: <argyllcms@xxxxxxxxxxxxx>
Cc:
Bcc:
Date: Sat, 5 Dec 2020 16:36:37 +0100
Subject: [argyllcms] Re: sRGB or AdobeRGB ?
Ideally they should be stored in the scanners native color space.
On Sat, Dec 5, 2020 at 10:29 AM Hermann-Josef Röser <
posts@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
Hello,
currently I am digitizing my slide collection. I am using a calibrated
EIZO
monitor with AdobeRGB gamut thus render my scan into AdobeRGB colour
space.
The question now is, what to do if I use these images on another monitor
with only sRGB or a beamer? Do I have to batch-convert all the images from
AdobeRGB to sRGB?
Displaying AdobeRGB images on a colour-managed sRGB device I would not
expect a problem and expect no difference on image quality to using a
version converted to sRGB before displaying them. Is this correct or do I
misunderstand something here?
Displaying on a non-colour-managed device is a different issue, of course.
Many thanks in advance and best wishes
Hermann-Josef
