[argyllcms] Re: Spyder4: Difference in using argyllcms and original software

  • From: Werzi2001@xxxxxx
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Wed, 11 Nov 2015 23:01:27 +0100

Hi,

thanks for your reply. Just as a test i tried to rerun the calibration
using:
dispcal -v -y l -q m -t 6500 -g 2.2 -o TargetA

The result did change a bit but not in the way the original software
does. You can download the result profile here:
http://www.xup.in/dl,62667722/DisplayA.icc/

During the calibration i got a few errors of the following form:
Failed to meet target 1.200000 delta E, got worst case 0.072355

Might that be the reason for the difference?

Thanks again.

Best regards
Werzi2001

On 11.11.2015 11:49, Florian Höch wrote:

Hi,

the main differences between the profiles are the whitepoint and tone
curve (gamma) target.
Both Argyll + dispcalGUI used the native whitepoint of the display
(which Spyder4 reported as roughly 7350K daylight). The Spyder4 software
used a 6500K whitepoint target.
Argyll by default uses a 2.4 gamma target. dispcalGUI uses 2.2. I assume
the Spyder4 software uses the latter as well.

You can set a whitepoint target for Argyll as follows:

dispcal -yl -t6500 DisplayA

The equivalent in dispcalGUI would be to choose a 6500K target on the
"Calibration" tab (or just use the "Office & Web" preset which sets it
for you).

Am 11.11.2015 um 09:57 schrieb Werzi2001@xxxxxx:
Hi,

i am using the Spyder4 colorimeter and wanted to calibrate and profile
my laptop screen using ArgyllCMS und Linux. The following commands were
used to create the profile:
dispcal -v -y l -q h -o TargetA
targen -v -d3 -f836 DisplayA
dispread -v -y l -k TargetA.cal DisplayA
colprof -v -D"Display A" -qm -as DisplayA

The profiles.zip can be downloaded here:
http://www.xup.in/dl,17734653/profiles.zip/

The result profile (see spyder4_argyll_linux_cmd.icc in profiles.zip)
didn't really change much. It seems that the brightness just gets a
little darker. After that i tried using dispcalGUI to create a profile
but the result (spyder4_argyll_linux_gui.icc) changes even less than
using the CMD version. Just to have a comparison i then tried the
original software under Windows. The result (spyder4_windows.icm) seems
to be right. It increases the red colors and according to a print i did
of one of my images my laptop was missing red color so far. Using
dispcalGUI i compared the color curves of all three files. The results
of ArgyLLCMS seem to be much smoother but also there near to the linear
curve. The result of the original software is a little more rough but
also changes the colors in a way that i suppose to be more correct.

So my question is why there is such a difference. Is it the colorimeter
that simply returns incorrect values under Linux or are the results of
ArgyLLCMS generally not expected to be (nearly) equal?



Other related posts: