[argyllcms] Re: General Questions
- From: "Michael Nagel" <michanagel@xxxxxxxxx>
- To: <argyllcms@xxxxxxxxxxxxx>
- Date: Sun, 23 Sep 2012 23:10:59 -0700
Graeme,
thank you for your response. I am currently using dispcalGUI with Argyll, I
hope I can ask questions here that refer to the usage of Argyll with
dispcalGUI.
Have a few follow up questions:
(1) dispcalGUI does not report the max and average fit error (or any other
report) when creating a correction matrix. Can ccxxmake (I am assuming
command line usage) which you referred to report on an EXISTING correction
matrix (by reading it's values and reporting) or does it only report after
it has just freshly created a new correction matrix ?
(2) understood, thank you
(3) To my knowledge, there are different chromatic adaptation methods such
as Bradford, Cat02 (from CIECAM02), von Kries etc. I was just asking b/c in
other software it is possible to change the chromatic adaptation method and
see if you get better results for a particular display.
But, I always had excellent results with Bradford (CAT02 was better on some
displays with specific options), so I was just asking if I missed a feature
of your fantastic software.
(4) yes, this is a little bit unclear to me, I'm just trying to understand
the options and possibilities of the ADVANCED GAMUT MAPPING menu (in
dispcalGUI):
By default in dispcalGUI, under the Advanced Gamut Mapping Options menu, the
source profile is set to sRGB.icm. I am not sure if it actually maps to sRGB
if I do NOT make any selections in any of the options in the Advanced menu
(Gamut mapping for perceptual, Gamut mapping for saturation, etc.)... ?
What would I select (or not select) if I want to calibrate my wide gamut
display to the widest gamut possible (not restrict to sRGB) ?
If I do want to map my display to sRGB (or any other color space), what
would be good, default selections for the options "Gamut mapping for
perceptual intent", " Gamut mapping for saturation intent" and "Default
rendering intent for display device profile" ? I am unsure what to select
here...
Thank you for all your work.
- Mike Nagel
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx]
On Behalf Of Graeme Gill
Sent: Sunday, September 23, 2012 8:01 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: General Questions
Michael Nagel wrote:
Hi,
> (1) Is there a way to see a visual report of a correction matrix
> (similar to a profile verification report) ?
ccxxmake reports the max and average fit error. Would it really be useful to
separate out this aspect as a verification function ?
> (2) Can this program write into the hardware LUT's of an Eizo CG275W ?
No, there is no current capability to write to specific display hardware.
I'd be pleased to add this, but no-one has seen fit to provide me with the
displays to RE and test, and I don't currently have the budget to buy dozens
of high end displays myself, and give the resulting code away.
> (3) How can I change the adaptation method (Bradford, Cat02 etc.) ?
Which adaptation do you mean ? What benefits do you see in changing this ?
The ICC white point adaptation is hard coded as Bradford in icc/icc.c,
icmChromAdaptMatrix(). cLUT profiles B2A gamut mapping tables will use
CIECAM02 white point adaptation whenever a gamut mapping algorithm using
CIECAM02 is selected.
> In all my verification reports I see that the adaptation method was always
Bradford.
Yes. It seems to be the accepted best simple (ie. 3x3 matrix) white point
adaptation algorithm.
> (4) Can somebody advise on the ADVANCED GAMUT MAPPING OPTIONS ? What
> would be standard default values to map to a gamut ?
I'm not sure what you mean. When creating profiles with gamut mapping, there
is a choice of pre-defined gamut mappings - see
<http://www.argyllcms.com/doc/collink.html#i>
for instance. If you want to alter the gamut mapping details beyond this,
then you would need to alter the source code, ie. see the gammapweights
tables in gamut/gammap.c. Beyond that, you'd need to code your own
algorithms.
Graeme Gill.
Other related posts:
