[argyllcms] Re: Bad measurements & white point calibration with Argyll 0.7b7 and GretagMacbeth i1Display 2
- From: Craig Ringer <craig@xxxxxxxxxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Thu, 10 Jan 2008 16:47:50 +0900
Craig Ringer wrote:
> Graeme Gill wrote:
>> Hal V. Engel wrote:
>>
>>> I think the issue is that LProf and ArgyllCMS, although they basically
>>> agree with each other, are giving significantly different results for
>>> white point =D65 then the vendors software with the same devices (IE.
>>> meter and display).
>> There is no verification that the vendors software actually achieved D65,
>> therefore it's of limited used as a reference. What's needed is
>> verification with something independent, say an Eye-One Pro
>> measurement.
>
> That's entirely true, there is indeed no proof the vendor tools managed
> to match D65. However, in this case the on-screen appearance of mid
> greys is a bright blue-purple when dispcal resports the white point to
> be near D65. We're talking serious blue here, this is not a minor colour
> cast. I know it's usually unwise to trust your own perception, but there
> are some points beyond which it becomes hard to argue.
I just figured out how to at least somewhat usefully explain this -
comparisons with the PANTONE swatch booklet. It's hardly ideal since I'm
trying to compare luminescent/backlit and reflective colour, but it
gives an impression of the situation. I assume you have a PANTONE
booklet or similar around?
A mid grey (RGB 117/117/117 in this case) viewed on the display when set
to its built-in sRGB mode looks somewhere around PANTONE 406M, 420M,
PANTONE 427M, PANTONE Cool Gray 1M or PANTONE Warm Gray 1M when viewed
in natural light (hooray for windows!). That is to say it looks vaguely
comparable with the mostly grey-ish colours in the swatch book. I'm not
going to try to guess closer than that because it's so hard to judge
matches between the display and printed swatches.
If set to the gain values that produce a D65 white point as reported by
dispcal the closest swatch is PANTONE 2705 . That's 9.4 violet, 3.1 pro
blue according to the swatch book.
I grabbed one of the Eizo S2410W panels from production and I'm getting
similar results to what I got with the Dell when testing with it just
now. The vendor software produced results where, after adjusting the
panel for a reported D65 white, midtone greys look grey and compare
resonably to the greys in the PANTONE swatch book. That proves nothing
at all, but is interesting.
Anyway, as I said I'll get you those readings so we can put some useful
numbers to all this.
--
Craig Ringer
Other related posts:
