[argyllcms] Re: AW: Re: Comparing sRGB and AdobeRGB
- From: "Wire ~" <wire@xxxxxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Mon, 15 Feb 2021 09:07:50 -0800
On Mon, Feb 15, 2021 at 05:05 Hermann-Josef Röser <
posts@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
Hello Graeme, Simone Karin, and Daniel
Thanks a lot for the detailed explanations.
The colour management should be working correctly. If I set the monitor to
sRGB the WINDOWs colour management tab shows the sRGB EIZO ICC-Profile. If
I
set the monitor to AdobeRGB, the Adobe-ICC-profile is set there
automatically.
As far as I can tell, the behaviour is as described by Graeme in his last
two paragraphs. I did the comparison with PhotoShop, which is to all my
knowledge, colour managed. I looked at the images also with imgaeJ, which
definitely is not colour managed. Here on the monitor set to sRGB the
AdobeRGB image looks pale, the sRGB image alright. If I switch the monitor
to AdobeRGB, the AdobeRGB version looks alright and the sRGB version is too
saturated.
Best wishes
Hermann-Josef
Hi H-j,
You originally wrote:
"I was astonished that also the sRGB-version shows more vivid colours in the
AdobeRGB-setting, i.e. the wide gamut is also pf advantage if rendered into
sRGB, which surprised me."
So you are saying you have a color-managed system with all the parts
apparently working as desired, and you notice that an sRGB image looks more
vivid when rendered as such in Adobe RGB mode, which seems surprising
because if the display alignment is of high quality the sRGB image is
expected to render the same in either mode.
If I follow you, I myself have noticed a similar effect but ignored it. I
have not studied it carefully, and so I just presume that I am subject of
some barely conscious perceptual bias. My system can't swap profiles on the
fly so given that my awareness involves color memory, the simplest
explanation has been I'm seeing things (so to speak). So if I understand
you, I cannot corroborate your experience but I know why you are asking.
The first thing that popped into mind was whether display uses perceptual
rendering intent, and with regards to finding a perceptual fit maybe there
are rendering artifacts? I am unable to express this as a statement bc I
don't understand how perceptual rendering intent works. What Graeme
describes as display standard operating procedure (SOP) sounds like RelCol
to me.
So while I can only amplify your question, for me the next thing I want to
know is how does perceptual intent work, and do display graphics libraries
use it?
Per another commenter's point about how much of the system is color-managed
and on one hand how much does OS observe in its built-in programs versus on
the other hand how does it support keeping programs informed about device
profile status, and do those programs pay attention to this information,
well that's a huge can'o'worms with a criminal poverty of documentation.
It's so bad that even Apple doesn't know what's going on in their own
programs, a la the XYZ LUT display profile rendering bug that DisplayCal
has to work-around regarding MacOS built-in apps, and Apple has always
offered the best CM integration.
Your setup adds in display driver support for swapping system profile, so
yikes, the nightmare continues.
Great thread topic, and I don't know if what you describe is for realz, and
CM at its best is a quagmire :)
/wire
Other related posts:
