[argyllcms] Re: Display Calibration Hardware Capabilities

  • From: "Alan Goldhammer" <agoldhammer@xxxxxxxxxxx>
  • To: <argyllcms@xxxxxxxxxxxxx>
  • Date: Thu, 12 May 2011 08:25:31 -0400


Thanks for the test evaluation information.  Part of the difficulty with the
ColorMunki is that X-Rite does not publish any detailed specifications about
the relative performance of this instrument on their website; they do have
data for the i1
Thus, we don't know what the inter-instrument variability is, or what QA/QC
procedures X-Rite use on CM prior to releasing specific units for sale.  I
have both a CM and the NEC branded i1 display that is configured for high
gamut monitors.  Until I started learning the ins and outs of the ArgyllCMS
tools I have been using NEC SpectraVision II to calibrate my monitor.  I get
comparable calibration results with both instruments.  I also have an i1 pro
coming in next week (it has been on back order) and will also look to see
how this performs.  I have done some extensive paper profiling (Epson 3880)
on two gloss papers (Museo Silver Rag and Ilford Gold Fiber Silk) and get
better profiles with ArgyllCMS than with CM software as judged by both the
gamut output and visual inspection of a test print (I used 980 patches and
maybe going to higher patch values can refine these results).

I'm not disparaging Ethan's test results since he only had one CM to test.
Looking at his test methodology, it was quite sound.  I don't think CM users
should despair, it is a decent enough instrument.  Whether the quality
control at X-Rite is sufficient to assure minimal variability between
instruments is unknown at present.


-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx]
On Behalf Of Graeme Gill
Sent: Wednesday, May 11, 2011 11:00 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Display Calibration Hardware Capabilities

Rishi Sanyal wrote:
> Anyone in Seattle with a ColorMunki or i1 Pro? I'm not convinced the 
> Munki is that bad b/c it generates reasonable profiles across my Dell
> U2711 as well as my older Dell 2005FPW. Couldn't say the same for my
> i1 Display 2, which just failed miserably on my U2711.

FWIW, I've just run a simple test with the i1Pro's and the ColorMunki I
have. I created a simple test chart with 5% device values + white, and ran
it on my (dim) CRT.

        dispread -w -v -V disptest2_XXX

The Y levels are around 0.1 cd/m^2.

Cross checked the results with

        verify -n -v disptest2_XX1.ti3 disptest2_XX2.ti3

(the -n is to remove absolute luminance level calibration differences)

So: (All raw Delta E)

        i1 Rev A vs. i1 Rev D:
                Total errors:     peak = 1.566833, avg = 0.570788
        i1 Rev A vs. Munki:
                Total errors:     peak = 0.926322, avg = 0.315342
        i1 Rev D vs. Munki:
                Total errors:     peak = 0.908153, avg = 0.530078

So I'm not seeing much discrepancy between the Eye-One and the ColorMunki on
low level readings.

[I did note that black drift on the Eye-One's is a bit of a problem
  in measuring such low levels. You really do have to let the instrument
  acclimatise for a while.]

Graeme Gill.

Other related posts: