[argyllcms] Re: 1st post (Spyder 3 Express + Dell 2711)
- From: Sam Berry <samkberry@xxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Thu, 9 Dec 2010 12:25:43 +0000
On 9 December 2010 11:14, Knut Inge <knutinh@xxxxxxxxx> wrote:
> On Thu, Dec 9, 2010 at 8:34 AM, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:
>
>> Knut Inge wrote:
>>
>>> Is the LUT included in the ICC profile? If not, how can I make any
>>> sensible
>>> comparision between my ICC profile and downloaded ICC profiles if all of
>>> them assume some non-included LUT that does per-channel gain?
>>>
>>
>> Which LUT are you referring to ? The calibration VideoLUT or
>> matrix/shaper shaper LUTs, or the cLUT device LUTs or the
>> cLUT multi-dimensional LUT or the cLUT PCS LUTs ?
>>
>> Graeme Gill.
>>
>> I thought I was confused earlier...
>
> I am referring to the info quoted below. I thought that the ICC/profiling
> acted as an absolute and relevant characterization of the display in itself.
> Judging from Sam Berrys response, it is not totally so, you first need to do
> calibration. Since he can see from my ICC profile that I did not do
> calibration, I am assuming that it is some "stuff" that will go into my
> graphics card, altering the apparent display characteristics and included in
> the loop while profiling the display. Is that the VideoLUT?
>
The stuff that goes into your card is indeed the VideoLUT, and these are
usually contained within the ICC. However.....
An ICC profile is only fully valid for one set of display settings, and
these settings are not contained within the ICC, and so the ICC is far from
an absolute description, video LUT or not.
>
> On Wed, Dec 8, 2010 at 10:19 PM, Sam Berry <samkberry@xxxxxxxxxxxxxx>
> wrote:
>
>> Your dispcalgui results, however, suggest that you have not calibrated the
>> display to D65 first though. The results are not directly comparable unless
>> they are all relative to the same white point.
>
>
> New and updated table of profiles. Now I am doing profile+calibrate, and I
> included dgm's measurement using the eye one display2
>
> Spyder 3 express
> SN: 030197xx
> Instrument Version: sw 1.2/hw 4CL
> Software version: 4.0.1
>
> All profiles obtained with out-of-the-box display settings.
>
> "Spyder software" is my measurement using the software that came with my
> Spyder 3 express
> "DispCalGui" is my measurement using the same calibrator but the GUI for
> ArgyllCMS - NOW USING CALIBRATION + PROFILING
> "TFTCentral" is the "Dell U2711 Standard Profile" from the review at
> http://www.tftcentral.co.uk/icc_profiles/Dell%20U2711%20-%20Collection.zip
> "Eye one D2" is an Eye One Display2 measurement kindly provided by DMG
> "Nec271w driver" is from the Nec driver available at
>
> http://www.necdisplay.com/cms/documents/Drivers/PA271W.zip
>
> Using ICC Profile Inspector
> Spyder software DispCalGui TFTcentral Eye one D2 Nec271w driver
> Illuminant:
> X 0.96420 0.96420 0.96420 0.96420 0.96420
> Y 1.00000 1.00000 1.00000 1.00000 1.00000
> Z 0.82491 0.82491 0.82491 0.82491 0.82489
>
> luminance:
> 181
> MediaWhitePoint Tag
> X 0.94728 0.94891 0.96420 0.93262 0.95137
> Y 1.00000 1.00000 1.00000 1.00000 1.00000
> Z 1.09006 1.09258 0.82491 0.95439 1.08815
> x 0.311878 0.311989 0.345702 0.323040 0.313000
> y 0.329235 0.328786 0.358537 0.346379 0.328999
>
> MediaBlackPoint Tag
> X 0.00250 0.00154 0.00092
> Y 0.00264 0.00148 0.00104
> Z 0.00288 0.00194 0.00189
> x 0.311721 0.310484 0.238961
> y 0.329177 0.298387 0.270130
>
> RedColorant Tag
> X 0.65756 0.66727 0.53729 0.60466 0.60602
> Y 0.30211 0.30634 0.21552 0.26689 0.27550
> Z 0.00958 0.00929 0.00783 0.00929 0.00694
> x 0.678421 0.678879 0.706366 0.686458 0.682102
> y 0.311695 0.311670 0.283340 0.302995 0.310087
>
> GreenColorant Tag
> X 0.17140 0.16295 0.28874 0.22639 0.21756
> Y 0.65845 0.65848 0.73769 0.68779 0.67461
> Z 0.10768 0.11018 0.07498 0.06554 0.08890
> x 0.182821 0.174912 0.262155 0.231076 0.221758
> y 0.702324 0.706819 0.669769 0.702027 0.687627
>
> BlueColorant Tag
> X 0.13524 0.13397 0.13817 0.13315 0.14064
> Y 0.03944 0.03519 0.04680 0.04532 0.04996
> Z 0.70763 0.70543 0.74210 0.75006 0.72908
> x 0.153279 0.153180 0.149039 0.143399 0.152923
> y 0.044701 0.040236 0.050482 0.048808 0.054323
>
> rTRC
> gamma = 2.2 256 points gamma = 2.2 256 points gamma = 2.2
> gTRC
> gamma = 2.2 256 points gamma = 2.2 256 points gamma = 2.2
> bTRC
> gamma = 2.2 256 points gamma = 2.2 256 points gamma = 2.2
>
> Now, "Spyder software" and "DispCalcGui" seems reasonably consistent,
> eliminating one potential source of errors. So perhaps my probe is bad, or
> perhaps my display really needs a lot more gain in the reds to do D65?
>
> Thing is, when I look at my "calibration curves", red is 255->255 while
> blue and green are 255->203. In other words, calibration chose to put
> massive gain into the red channel. How do I compare the sanity of this to
> other sources if the calibration curve is not included in the ICC?
>
You largely can't, as it all depends on the original display settings. The
actual numbers mean little, unfortunately. I still suspect that your Spyder
is pretty far out though. The green primary error is very similar to the one
my first sypder gave, and made skintones look really obviously wrong. I
suggest you do a profile only of the factory sRGB and see what white point
that gives, although it's still not really a definitive test.
Sam
Other related posts:
