[argyllcms] Re: ColorMunki measurement drift

  • From: Juergen Lilien <jp.lilien@xxxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Tue, 28 Sep 2010 06:45:43 +0200

Graeme Gill wrote:

> -Iw is specifically about trying to stop changes in display brightness
> during measurement affect the calibration or profile. It certainly doesn't
> attempt to second guess the instruments accuracy. So it won't have any
> effect if the instruments absolute calibration were to drift.

Thanks for clarifying!

> As I said before, I wouldn't get too obsessed with white point errors,
> especially when measured in degrees. It's visual difference that counts,
> and that needs to be measured in delta E.

Please don't get the impression that I have such great faith in pure
(K) numbers. But I found that when I do not set the white point during
the interactive calibration to be within 0,3 DE, the resulting LUT
curves will not end in 255,255,255 and as a result I loose more tone
values, and that is what I want to avoid in the first place.

A complete calibration and profiling in hires+adaptive mode takes
nearly an hour, and the drifting seems to degrade the results
(bigger changes to the LUT curves than actually needed).

I'm still struggling to get a visual neutral grey gradient, after
calibration, although its perfectly neutral without calibration
and the profile verification looks perfectly fine.
So I'm really more after the visual results, I'm only observing
the numbers to find an explanation for the visual shortcomings.

Regards, Juergen

Other related posts: