Getting bad results with calibrating MacBook Screen

Home Forums Help and Support Getting bad results with calibrating MacBook Screen

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #25634

    HB123
    Participant
    • Offline

    Hello, I am trying to calibrate my MacBook 2020 13” I am not doing any color specific work on this computer, I only use it for school assignments along with Netflix/YouTube/movies. My goal is to just balance the white point and get a 2.2 gamma so that movies look good.

    My problem is that I tried to use the default settings (single curve + matrix with Black Point Compensation) and I am getting very splotchy looking oranges/reds/browns.

    I then tried doing XYZ LUT + matrix without Black point compensation (which is what I have on my Windows PC) and this made my blacks too dark, as well as giving me grainy appearance in movies.

    What settings can I use to make my screen look correct? All I want is for Netflix to look good (I am watching it in safari, all color options like True Tone/night shift/automatic brightness are off during calibration)

    #25837

    Wire
    Participant
    • Offline

    Your Macbook is very well aligned from factory for your purposes. The screen should look great out of box and be colorimetrically decent.

    If it looks bad, check Displays  > Color pref to ensure the built-in display profile is selected, it’s named something like “Macbook LCD” or some such. If it gets deleted it will automatically be replaced

    Apple whitepoint is pretty consistently a bit cool at about 7100 and decently tracks the white locus (not too green or pink). This coolness appears to be design choice. Getting it closer to D65 would be desirable in a rigorous use case, but in a single-display daily-use setup you will never notice the difference.

    Apple tends to a slightly higher than spec gamma for contrast enhancement that sacrifices no detail and most people think looks better than more accurate response—what’s known as difference between ‘objective vs preferred’ contrast in film tonal response.  This helps bridge display between rendering concerns of desktop publishing (Adobe) and video. It can be that built-in display profile advertises a sRGB (Display P3) TRC but delivers a Rec.709. This is just a rendering intent and nothing to fuss about unless you do photo/video color grading.

    As to trying to force whitepoint thru DCal, don’t bother unless you have a very clear reason and you understand trade-off of potential loss of tonal precision.

    Your problem with messed up color after alignment suggests a basic misconfiguration or error in the process.

    XYZ LUT profiles are well-known to not be handled properly by MacOS itself and can cause effects like you describe. Do not use.

    Be careful to avoid confusion with bad profile sticking around, there are user and system colorsync folders; you can go to user or both. Don’t install in both (system folder)  unless you need to avoid heartache of dups

    If your instrument has a problem you would expect to see prob elsewhere but you don’t? Nonetheless this is something that has to be questioned…  A mismatch between instrument and display lamps isn’t likely to cause blotchiness you describe.

    There’s no reason why you shouldn’t be able to use DCal to create an excellent alignment for a 2020 macbook. But there’s no basic need to either.

    #26234

    Roger Breton
    Participant
    • Offline

    Hope this is the best place to post my question with regards to calibrating a 2019 15″ MacBook Pro LED screen.

    I am in a situation where the screen ought to match a printed inkjet proof. Needless to say, I suppose I need to select “D50 target white point”. My setup is as follows: MacBook Pro with Photoshop and a JUST-Normlicht (JNL) Colormatch 5000 desktop viewing booth (so JNL 20″ ProGraphic lamp).

    Making the display profile is pretty straightforward using the i1Pro2. Once the profile is made, I use a good-old Macbeth ColorChecker 24 patch chart for comparison. I have painstakingly measured each patch of my ColorChecker with the i1Pro2 and created a digital version of the chart in Photoshop with those exact CIE Lab values. So, I’m comparing a digital ColorChecker to a physical ColorChecker under my JNL viewing booth.

    Usually, when I do this exercise on, say, my NEC PA271W monitor, I get a satisfactory match. But, for some reason, on a MacBook Pro LED, the match is quite poor.

    I experimented with various target white point chromaticities to no avail.

    Please excuse the newbie kind of question but is it possible that the “shiny” screen finish of the screen, somehow throw the i1Pro2 measurements off?

    Yes, I deactivated every possible ‘smart’ options in the Mac before proceeding.

    #26247

    Wire
    Participant
    • Offline

    I will guess you are using a D50 environment, and Mac native doesn’t fit that so well?

    To be clear, you are using a spectro, not a colorimeter… Sometimes there is confusion for readers.

    When you say “quite poor” what does this mean to you?  Does it mean that when you measure the display face you get Lab that doesn’t agree with your color checker measurements after accounting for whitepoint and luminance? How would you describe the normalized difference between your measured ColorChecker and your display in dE? You can use Argyll spotread command to answer this. It reports dE Lab and ’94 as well as color temperature with “-T”

    Using the Macbook Pro’s default display profile should provide visually good results notwithstanding whitepoint and finer points of tonal response. The distinction is between a pleasing viewing experience and absolute accuracy.Yes, the Mac can be made more accurate with a custom display profile. Nothing wrong with that, unless you get the config messed up. Use DisplayCal in default mode until you have very good reasons for doing something else. If your Mac display just looks bad out of box…? Not likely. Once you start messing with with Displays Color  Preferences, you open the door to a misconfig. So double-check.If you plug Mac into NEC, what happens?If you want a close white match to a D50 viewing booth, getting the Mac to harmonize to some “white” reference substrate is quite possible using DisplayCal’s visual white-point editor.  I will lament for you a D50 display appearance in general day-to-day use. There’s nothing wrong with adapting a display to the surroundings. But sharing on the webz and video drags you inexorably towards D65. I understand there may be a preference for D50 as a matter of pro-forma protocol. But how does it work in general use? I don’t understand why people will tolerate a display that looks bad in order to feel like they’re tracking standards. Again, getting parts of a system to harmonize makes sense. D50 looks bad if there’s any other lightsource around.White is relative, so good to make yourself comfortable. I have found that in the long run D65 is the sweet spot — IMO.

    An image color space with a D50 white gets adapted to the display, just like an image space with a gamma gets adapted to the display. The genius of ICC is that if you can get your image data into a color space correctly, it will tend to render on a properly profiled display well; both white and gamma. So in a sense, display white-point / gamma only matter WRT adapting the display to the environment and for proofing.In general, the more differences there are between parts of an imaging pipeline, the more trouble it is to keep track of how things can go wrong. This is sRGB’s eternal gift to consumer color. But things still go wrong anyway.

    But it all comes down to what you mean by “quite poor”

    #26248

    Roger Breton
    Participant
    • Offline

    1 – D50 environment, yes, but MacBook Pro LED “native”  seems to have a mind of its own.
    2 – X-Rite i1Pro2 spectral device
    3 – “Poor” means regardless of how well the systems says the monitor is calibrated, its visual appearance does not agree with measured accuracy. IOW, even if the system ‘reports’ acceptable color differences (between expected and measured, as far as the generated monitor profile is concerned), matching of real world objects viewed under ‘D50-types of lighting’ is not acceptable. So even though colorimetric accuracy is achieved, the visual appearance does not correspond? It could be a case of Observer Metamerism but I’m not the only one seeing the mismatch?

    4 – I usually never question the use of D50-like lighting for  ‘screen to proof’ applications. I suppose it’s ‘possible’ to use something other than D50? In my humble experience (I don’t know how many monitors of various makes I calibrated/profiled in the past, including MacBooks and PowerBooks and Apple Displays), D50 always worked. It’s not perfect but always more acceptable that anything resembling D65-like target white point calibration.

    5 – Moving away from D50, in my application, would have profound implications and I’m not even sure how best to go about it. This is crazy that I’m even writing this out? It’s ‘heretic’!

    Thank you, Wire, for taking the time to articulate your vision.

Viewing 5 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS