AndN11 pisze: ↑17 lipca 2019, 14:38 - śr
Tom01 pisze: ↑17 lipca 2019, 10:05 - śr
...Metodę sprawdzenia miał Pan podaną kilkukrotnie. Dlaczego nie skorzysta Pan z niej?...
OK. skorzystałem z ArgyllCMS/DisplayCAL. Ostateczne wnioski..., pozostaję przy trybie fabrycznym Adobe RGB i profilu dołączonym z płyty CD
Równomierność podświetlenia wygląda tak:
tryb Adobe RGB.jpg
Raised brightness in corners for white is usualy a smoking gun of some kind of uniformity compensation. Uniformity compensation is done ALWAYS at the expense of contrast. The worse it is, the more contrast that display has to loose to be aceptable.
While 600:1 with uniformity compensation ON and a D50 white point for printing is aceptable, loosing to much contrast in a D65 general purpose (photo, video, multimedia) because uniformity compensation is not acceptable.
A respectable uniformity report should have attached reference white point and contrast. Here is missing. If you do not want to believe me, look in "prad.de" reviews.
So your uniformity report is incomplete.
It is possible of course that you get good contrast and good uniformity, congratulations... but ... is it common? Given Benq reputation and the reposts of former SW2700/SW271/SW320 owners across internet your situation is extremely unlikely.
I mean, you may be lucky but the chances that a ramdom EU customer who buys a Benq SW and gets acceptable results are
VERY VERY LOW.
For example, this SW271( more than 1000 euros) is trash:
https://cdn.photographylife.com/wp-cont ... 60x764.png
Very bad uniformity. Very common to get this bad uniformity with a Benq SW.
AndN11 pisze: ↑17 lipca 2019, 14:38 - śr
Nie bardzo wiem jak wstawić raport z walidacji (verification). Plik ma rozszerzenie .htm
Może jest lepszy sposób, spakowany zip: Fabryczny_aRGB_Measurement Report 3.8.3 — BenQ SW2700 @ 0, 0, 2560x1440 — 2019-07-17 13 31 22.zip
Średnia ΔE 0.25, Max ΔE 0.8
As I explained before Benq's SW2700PT is not a RGBLED monitor. It is a QLED from AUO, hence spectral correction you chose is incorrect. There are CCSS community contributions with such backlight but unfortunately they need some "user cooking" because the person that created them made them in a wrong way (CCSS stores emulated gamuts like sRGB or AdoberGB, not native gamut, no it is useless as is, without "cooking").
Anyway... this would be related to wrong whietpoint and can be corrected choosing the right one, so it is not a monitor problem, just user configuration problem.
This is a graphics card (GPU) calibration. Grey is corrected using LUTs in graphics card like in common monitors:
(extracted from your report if you click "additional statistics")
Calibration red tone values
>= 95%
96.1%
(246/256)
Calibration green tone values
>= 95%
96.1%
(246/256)
Calibration blue tone values
>= 95%
95.3%
(244/256)
Calibration grayscale tone values
>= 95%
95.3%
(244/256)
It has a very good neutral grey (range a*b* <1),
good!...
but it is a GPU calibration. That means that depending on which GPU is inside your computer you'll get banding. For example laptops and intel integrated iGPUs, and some nvidias.
There are visual ways to test banding caused by GPU Calibration, like seeing lagom gradient png (
http://www.lagom.nl/lcd-test/img/gradient-h.png) in Microsoft Paint.
With a discrete desktop AMD GPU all should render smooth with ZERO banding (10+bit LUTs and temporal dithering, like in HW calibration)
AndN11 pisze: ↑17 lipca 2019, 14:38 - śr
Po kalibracji programem PME wyniki są gorsze:
Średnia ΔE 1.88, Max ΔE 4.5 (trzy próbki koloru na 50 maja delta E powyżej 4.0 )
We do not care about mean and max errors. You should not care about them. Do a DisplayCAL valiation and check grey like you did in you graphics card calibration, like the report you did. Is grey ok? That is the important thing, if PME is able or is not able to correct grey issues in uncalibrated CAL1 or CAL2 modes.
If it is not, if it shows a* or b* tints far for zero or huge oscilations between positive and negative, calibration is faulty...but there is hope. If you own a AMD Radeon, run a DisplayCAL calibration on top of PME calibration. All grey issues should go off. With other cards banding is granted (intel) or is likely to appear (DVI + nvidia)
AndN11 pisze: ↑17 lipca 2019, 14:38 - śr
Pogarsza się też równomierność. Pięć pól przy prawej krawędzi jest na czerwono.
Uniformity compensation zone mapping is disabled in some OSD modes. It is common in these Benq, or Dells etc. Dell's uniformity compensation goes to OFF when yo use CAL1/CAL2... you cannot use it.
NEC PAs and Eizo CS and CG allows you to use uniformity compensation in every OSD mode, including OSD modes with HW calibration.
NEC PAs allow you to choose how much contrast you loose white correcting uniformity which is very good feature.
So, you see here... you get what you pay.
AndN11 pisze: ↑17 lipca 2019, 14:38 - śr
Może jeszcze coś tam źle ustawiłem, może tak jest i będzie, a może po dalszych aktualizacjach PME coś się zmieni.
Na pewno jeszcze wrócę do kalibracji sprzętowej jak przejdę z DVI na DisplayPort.
Coś kojarzę, ze ostatnie kalibracje robiłem wybierając opcje matrix, a nie 3D LUT, trzeba będzie się jeszcze temu przyjrzeć.
If you have a nvidia GPU is discouraged to use DVI if you want to calibrate using GPU: banding. Use Displayport.
AMDs do not suffer this issue over DVI.
