[argyllcms] Re: Calibrator correction for Eye One Display 2 & GB-LED Backlight LCD Monitor?

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Thu, 05 Sep 2013 12:51:23 +1000

Al Robertson wrote:


> The NEC has become dim over time and I'm looking to replace it soon with an
> Asus ProArt PA249Q which, like the better known Dell U2413, has the LG
> LM240WU9-SLA1 AH-IPS panel. But I have one problem.

> This relatively new panel has a GB-LED backlight (improved gamut), but from
> what I've read, the curves are sufficiently different that I would need to
> either go buy a new calibrator (X-Rite wanted to sell me the i1Display Pro -
> $229), or get a new correction matrix to successfully use my existing
> colorimeter.

The i1Display Pro is a pretty solid instrument, and well worth
the investment.

> Does anyone have a corrections file for the Eye One Display 2 and a GB-LED
> backlight display? Or perhaps be thinking about either making one, or
> including GB-LED support into argyllCMS/dispcalGUI as a new mode?

The variability of the instrument is such that I'd be a bit suspicious
of a correction matrix made for a different instance of the i1d2.

> If not, any ideas on how to make or fake one at home?

The best way of course is to use a spectrometer (ie. i1pro or ColorMunki 
as a reference on your display against your i1d2.

> One idea.The monitor claims  to come factory calibrated within 3 delta E
> (fingers crossed.) Is there at least a way to make a correction file that
> uses the factory calibration as a trusted baseline so I can always get back
> to that even if the colors drift over time (or if I want to customize the
> calibration)?

OK, here's an outline of how you might go about this. I'm
leaving out some details, and I haven't actually tried it,
so you may well need to troubleshoot the procedure:

Create a reference test set of (say) a 3^3 or 4^3 cube set using
targen, ie. something like:

    targen -v -d3 -e0 -B0 -m3 -f0 test

Copy this to a reference set:

    copy test.ti1 ref.ti1

Set your new display to (say) sRGB mode, and measure the
test set with your instrument.

    dispread -v -yl test

Create a reference set using the sRGB.icm file, that we
assume the new display is calibrated to:

    fakeread -v sRGB.icm ref

A problem with this is that the real readings will be
returned to absolute values (cd/m^2), whereas the fakeread
values are normalised to 100, so we will have to make sure
that the fakeread values get scaled the same way.

Open the test.ti3 file in a text editor and copy the
LUMINANCE_XYZ_CDM2 line, ie. for the sake of an example:

  LUMINANCE_XYZ_CDM2 "46.509281 49.445099 45.615122"

into the ref.ti3 file in a similar position. This should
cause the values to be scaled to match the instrument values.

Create a ccmx correction matrix from the ref and test results:

    ccxxmake -v -f ref.ti3,test.ti3 correction.ccmx

You can then use the correction matrix with your instrument
on that display, ie.

    dispread -yl -X correction.ccmx etc.

Graeme Gill.

Other related posts: