Whitepoint on Different Screens

[Jacob Pritchard]

Hi there, professional photographer here of 10+ years. I’m not super technical, but trying to step my game up, as far as monitor calibration and have been experimenting and learning DisplayCal over the last several days.

I’m able to calibrate an Eizo and Macbook in a way that appears to be accurate/working well.

One issue confusing me: When I set “whitepoint” to “as measured” under the calibration tab, and then do a verification after calibrating, the measured whitepoint on my Eizo is very close to 6500K. However, what seems unexpected to me is that my Macbook is showing a measured whitepoint of around 7200K.

So, thinking that it’s a good idea to have both monitors calibrated to the same whitepoint, I tried changing whitepoint to “color temperature” and entering 6500K for the Macbook, and then ran another calibration. After that, however, the macbook appears to have a fair orange/magenta tint. Even after allowing monitor to run with this profile for some time so my eyes can adjust to the new whitepoint, this calibration still seems “off” compared to my Eizo, or to the “as measured” version.

Am I best off just sticking to “as measured” for both displays? Or is it possible I’m missing some other element during my calibration process?

• This topic was modified 5 years ago by Jacob Pritchard.

[Vincent]

1. Your measurement device could be innacurate, some models are (or could be configured to be) accurate, others are not.
   We do not know what do you use or how you configured it.
   Some reliable & accurate colorimeters like i1DisplayPro need a “reference” backlight for measurement ( *.ccss files).
   -LED sRGB macbooks or LED sRGB Eizos (like EW or CS230) should use “WLED” correction.
   -LED widegamut CS Eizos or CGs older or equal to CG277 should use a GB-LED correction (RGphoshor).
   -LED P3 macbooks should use a community made P3 correction or use WLED PFS phoshor correction (“Panasonic VVX17P051J00”)
   -newer LED widegamut CGs like CG2420/CG2730/CG279X/CG318X/CG319X should use a variant from WLED PFS phoshor correction with bigger gamut (“HP_DreamColor_Z24x_NewPanel”, found in HP Z32x spoftware but Florian uploaded that file in some thread a days ago), or a community correction (there is one for CG318) or if not found use the same as macbooks (“Panasonic VVX17P051J00”)
   More common spectrophotometers like the ones from Xrite are very limited when measuring very narrow spectral power distributions like those in WLED PFS phosphor variants. Use ArgyllCMS 3nm mode for them.
2. It is common for a laptop to do not have D65 white from factory calibration, it does not matter if they have a lit apple on the back or not. Manufacturers do not care about it. What you may expect from a “semi-premium/premium” models is a generous sRGB coverage and neutral grey with no color cast and more or less gamma 2.2 and a white point more or less white (close to daylight locus and its infinite number of whites from warm-yellow to cold-blue, but “white”)
   White point calibration in laptops (with few exceptions) or lesser/locked displays (like iMacs) is done in graphic card, in some HW tables called LUT which store grey+gamma+white corrections. It is done at the expense of contrast but a good IPS laptop shpuld be able to attain D65 with little contrast loss.
   These LUT tables “may” have limited 8bit precision causing that going up or down by 1 bit value in correction gives you a slight green/magenta cast in white. If that happens you cannot improve it with that hardware, it is limited (AFAIK all intel iGPUs but maybe newer ones have improved that)
3. White point is not defined by correlated color temperature (CCT). This is a common mistake tha you should avoid. CCT or its variants give you info about yellow-blue tint… but you do not know how it behaves in green-magenta axis. 6500K CCT white can be pink. There is a thread here (~1 week ago maybe) with a link to Andrew Rodney book explanation for white points. Take a look.

So, first of all:
-make sure you own an accurate device
-redo you white measurements using a proper correction if you didn’t that way
-one you know the actual behavior of your display you can plan an action like “do nothing”, make your macbook share Eizo’s numerical white,  make your Eizo share your macbook numerical white (if macbook was very limited) or make a visual approach as a solution of compromise (a not numerical approach, DisplayCAL for macbook  and your Eizo’ Color Navigator have toools for this ).

Calibrite Display Pro HL on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Jacob Pritchard]

Thank you so much for the thoughtful and thorough response!

For what it’s worth, I have been configuring using the “default” and “laptop” stock setting (and then sometimes tweaking a single factor like the whitepoint.

I’ll have to dive deeper into the “correction” options. I had been running at “Auto (none)”. Any tips for a newbie on determining exact correction for my monitor? I’m not sure if my model would be  LED sRGB, LED WideGammut or something else, and a preliminary google search, as well as look at the manufacturers specifications did not illuminate things for me. Clicking on the globe icon next to correction *does* bring up user contributed corrections for my macbook display, but all indicate they were done with instruments that are not the ones I own. There are no corrections in this section for my Eizo monitor.

One other follow up questions…

Any straightforward way to determine if I own an “accurate device”? I have a Spyder4, which is admittedly several years old. Are there any calibration/reference tests that can be run to determine that it is accurate?

Thanks again for the very illuminating response.

[Vincent]

Thank you so much for the thoughtful and thorough response!

For what it’s worth, I have been configuring using the “default” and “laptop” stock setting (and then sometimes tweaking a single factor like the whitepoint.

This is what you may want to modify, nothing else with the exception of colorimeter correction or spectrophotometer resolution

I’ll have to dive deeper into the “correction” options. I had been running at “Auto (none)”. Any tips for a newbie on determining exact correction for my monitor? I’m not sure if my model would be  LED sRGB, LED WideGammut or something else, and a preliminary google search, as well as look at the manufacturers specifications did not illuminate things for me. Clicking on the globe icon next to correction *does* bring up user contributed corrections for my macbook display, but all indicate they were done with instruments that are not the ones I own. There are no corrections in this section for my Eizo monitor.

The list with possible choices is above, in my 1st message.
In theory Spyder4&5 could use CCSS from community.

One other follow up questions…

Any straightforward way to determine if I own an “accurate device”? I have a Spyder4, which is admittedly several years old. Are there any calibration/reference tests that can be run to determine that it is accurate?

Thanks again for the very illuminating response.

Spyder4/5 were not very accurace even when you opened its box for 1st time. I would not call them accurate at all.

Spyder bundled corrections (upper right in DisplayCAL) in your situation are limited to WLED for sRGB LED screens and Widegamut LED for other P3/AdobeRGB displays. They are not very good and are not close to what these widegamut are… but it is what you have.

Try to get an i1DisplayPro. Colormunki Display is accurate but it is not going to work with Color Navigation for the Eizo.

[Vincent]

… or try the visual approach. It should be explained in DisplayCAL FAQ/documentation and in ColorNavigator manual
It depends on which display you choose as “reference” because it is better or because it is difficult to modify its whitepoint (so you modify the other one to match).

[Author]

Posts