Blacks / Shadows Lifted in MacBook Pro (Mid-2015) Calibration

Home Forums Help and Support Blacks / Shadows Lifted in MacBook Pro (Mid-2015) Calibration

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #25570

    Daniele Sestito
    Participant
    • Offline

    Hey everybody!

    I’m trying to calibrate my MacBook Pro (2015, Retina Display, 15in) with DisplayCal using my X-Rite i1Display Pro Plus. I’ve done the calibration three times since the first two results had off hues. The third one is the best and had solid numbers on the verification test (attached).

    However, the blacks are too milky and the shadows feel lifted in this latest calibration. In other words, they don’t appear as dark as I feel they should be. It’s pretty noticeable when viewing clips from some movies that I’ve seen many times before and know generally how they should look like.

    Here are my latest settings:

    • Mode: LCD
    • Correction: Spectral LCD White LED Family
    • Interactive Display Adjustment: On
    • Whitepoint: As measured
    • White level: As measured
    • Tone Curve: sRGB
    • Profile Quality: High
    • Amount of Patches: 3454

    I turned off the advanced options on this latest calibration since I didn’t feel like I was making use of those settings, besides the Whitepoint (normally D65) and White level (normally 120 for me).

    Let me know what you think, or if you need any more info from me. Thanks!

    Attachments:
    You must be logged in to view attached files.

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

    #25610

    Vincent
    Participant
    • Offline

     

    However, the blacks are too milky and the shadows feel lifted in this latest calibration. In other words, they don’t appear as dark as I feel they should be. It’s pretty noticeable when viewing clips from some movies that I’ve seen many times before and know generally how they should look like.

    • Mode: LCD
    • Correction: Spectral LCD White LED Family     <—————- Wrong if it is a P3 macbook (use P3 for macs), OK if sRGB laptop
    • Interactive Display Adjustment: On
    • Whitepoint: As measured
    • White level: As measured
    • Tone Curve: sRGB                                                               <—————- HERE is the culprit
    • Profile Quality: High                                                          <—————- & next: macOS color management engine is extremely faulty
    • Amount of Patches: 3454                                                so stay in “default setting” or UI artifacts may appear in UI elements

    I turned off the advanced options on this latest calibration since I didn’t feel like I was making use of those settings, besides the Whitepoint (normally D65) and White level (normally 120 for me).

    Let me know what you think, or if you need any more info from me. Thanks!

    Expained in quote

    #25611

    Marcin Czornyj Kaluza
    Participant
    • Offline

    – turn off auto brightness in MacOC prefs

    -use gamma 2,2 instead of sRGB

    -change profile type to single curve matrix

    #142758

    Marcello Frisina
    Participant
    • Offline

    Did you ever figure this out? I’ve tried calibrating my iMac 24” retina screen with the i1Display Pro and it’s giving me the same issue you’ve had—raised blacks. It also does this on an OLED monitor. I’ve noticed that pure black stays pure black, but any value that is just slightly above pure black jumps to be far too bright, killing contrast even at 2.2 gamma

    #142760

    Vincent
    Participant
    • Offline

    ArgyllCMS aims to neutral grey rather than gamma because actual gamma (near black any correction/calibration must choose a priority) can be recorded accurately on ICC profile’s TRC. Your problem is macOS and its limited profile supports that forces you to use idealized TRCs instead of actual ones.

    Also your claims unless are done non color managed are no proof of calibration lifting blacks, because CMM of macOS desktop may be playing a role here.

    • This reply was modified 1 year, 4 months ago by Vincent.
    #142764

    Marcello Frisina
    Participant
    • Offline

    I’m new to color management in general, so forgive me if I sound ignorant here–does what you’re saying mean that it’s impossible to get good and accurate gamma curves (Tone Response Curves?) in lower values while calibrating in MacOS? Is there some way to work around MacOS’s imperfections here?

    #142765

    Vincent
    Participant
    • Offline

    No.

    I’m saying that:

    -your claims about lifted blacks may or not may be caused by calibration. You need to measure it non color managed. Maybe there are no lifted blacks at all and your imac24″ has am actual TRC  that does not match default generic display profile from apple (that linear ramp in TRC), so with default profile you are seeings things “wrong”. Same applies to whatever OLED monitor you pluged to macOS, macOS is going to build a default profile with EDID color coordinates and likely that fake TRC with linear ramp.
    You have to measure to support your claims regarding calibration.
    -macOS CMM, the one used by desktop and Apple app is very limited regarding profile support, but works well if display matches defsult uncalibrated prodile. Adobe CMM is not.
    -argyllcms seems to aim to neutral black rather that target gamma near black, because your target gamma calibration will be undone by CMM, you just need that display ICC profile stores actual TRC. You could have gamma 1.8 on display and render AdobeRGB (g2.2) images accurately if its colors are in gamut while on color managed apps, you just need that display profile stores that 1.8.
    -given macOS limitations and the behavior of idealized profiles (matrix, perfectly additive and power-law single curve TRC), such an idealized profile suitable for macOS desktop may not track actual display TRC. Adobe apps have no such limitation because Adome CMM (ACE) is better, you can use other profile types that capture accurately display behavior.

    But first of all you need to find if there are lifted blacks. For example verifying custom profile and default profile on iMac.

    #142769

    Marcello Frisina
    Participant
    • Offline

    Okay I think I understand your meaning. I did take some measurements to make sure I wasn’t crazy when I was calibrating. Before profiling, instead of setting black point automatically, I was able to measure the black point instead. On my iMac at 120nits it came to about 0.1 I believe. But after calibration, the “achieved” was always measured at something like 0.4. I’ve returned my i1Display Pro (via Wacom Color Manager) and ordered a Calibrite Display Plus HL to replace it, hoping for better fidelity, so I can’t take an exact measurement right now, but hopefully those results I mentioned above are some indication of what’s going on. I’ll also say that I’ve been able to compare the iMac screen to both my Pro Display XDR on my MacBook Pro, as well as my hardware-calibrated Wacom Cintiq Pro. After much scrutinizing over low-luminance gradation patterns on all three monitors, it’s readily apparent that the iMac looks closest and most accurate to the other two in terms of gamma when on its factory default “iMac” icc. When I switch to the calibrated profile, whether done in DisplayCal or the Wacom Color Manager software, the black levels approaching full-black raise (while pure black stays the same). Is this just an Apple thing, and if so, is there any way to navigate the issue?

    #142770

    Ben
    Participant
    • Offline

    Try blackpoint compensation off in the the calibration section not the profile section.  Your black grey balance is so far off it has to add grey to balance it.   I would see what the lut looks like in the iMac ICC.   Its ramps must be linear I assume.   Is its gamma correct on the iMac icc ?

    #142776

    Vincent
    Participant
    • Offline

    Okay I think I understand your meaning. I did take some measurements to make sure I wasn’t crazy when I was calibrating. Before profiling, instead of setting black point automatically, I was able to measure the black point instead. On my iMac at 120nits it came to about 0.1 I believe. But after calibration, the “achieved” was always measured at something like 0.4.

    300:1

    I’ve returned my i1Display Pro (via Wacom Color Manager) and ordered a Calibrite Display Plus HL to replace it, hoping for better fidelity, so I can’t take an exact measurement right now, but hopefully those results I mentioned above are some indication of what’s going on.

    It’s highly unlikely that this was caused by colorimeter or ArgyllCMS by itself. More likely to be caused by some user misconfiguration.

    As Ben pointed, rather that start to “order things” online, check calibration curves first.

    I’ll also say that I’ve been able to compare the iMac screen to both my Pro Display XDR on my MacBook Pro, as well as my hardware-calibrated Wacom Cintiq Pro. After much scrutinizing over low-luminance gradation patterns on all three monitors, it’s readily apparent that the iMac looks closest and most accurate to the other two in terms of gamma when on its factory default “iMac” icc.

    This is unlikely too. XDR track profiles very well out of the box. Show us the HTML reports.

    When I switch to the calibrated profile, whether done in DisplayCal or the Wacom Color Manager software, the black levels approaching full-black raise (while pure black stays the same). Is this just an Apple thing, and if so, is there any way to navigate the issue?

    First of all, check calibration curves and chack that you are not adding some extra configuration. For example video levels, if you get 300-400:1 on a 1000:1 IPS withou activating uniformity compensation or something like that, 99% chances that you are appyling video levels (is it even possible on an iMac using its screen?) or applying a brutal WP correction through VCGT or messing with near black calibrartion by enforcing some extrem correction on calibration tab.

    #142782

    Ben
    Participant
    • Offline

    It could be running the report wrong.   I agree it is not the i1display pro.   It would be wrong on all the tests.   Low contrast indicates a video level problem or a LUT that is way higher 256,256,256  or is not linear at all in RGB 1 to 32.   The report would be good evidence.   It is hard to fix without calibration reports. HTML reports are preferred.   The ICC profile would give more clues to what is going on too.

    • This reply was modified 1 year, 4 months ago by Ben.
    #142861

    Marcello Frisina
    Participant
    • Offline

    Hi Ben and Vincent,

    Thanks for all your assistance in this matter. Thankfully I’ve found a solution, although my situation has changed a bit. For the sake of anyone else on MacOS who is dealing with the issues I was having, I’ll first state my solution simply, and then I’ll go into a bit more detail in how my situation has changed since last writing here.

    The simple solution for solving near-black luminances banding and elevating too high after a calibration: change the profile’s target gamma setting from 2.2 to sRGB (perhaps not if a photographer, but I am an animator, and color accuracy is most important to me as it will be seen on other sRGB screens, so not via Adobe RGB print, but rather the Display P3, sRGB, and Rec.709 color spaces)

    Anyway, since I last wrote here, I upgraded my computer (unrelated to my monitor issue, I just needed an upgrade) and went from the M1 iMac to an M4 Pro Mac mini. This, however, called for a new secondary display to replace the iMac’s, and I actually went with an OLED: the MSI 321URX. It achieves 100% sRGB color even after calibration, so while not quite as vibrant as other pro monitors, it’s a great cost-effective solution at seeing how an SDR animated scene will display on a high-end modern display. Anyway, even after getting the OLED and switching computers, I struggled initially with the same issues I had with my Mac’s display: lower luminance levels being raised far too high in comparison to literally any other screen I own.

    I then did some more research on sRGB gamma, versus 2.2 gamma, and while they are often considered quite similar, it is apparently in the low-end light levels that sRGB tracks differently. So I figured, what the hell, give it a shot… and voila!!! It now looks extremely accurate to my other, factory-calibrated Wacom Cintiq Pro. Despite the fact that the OLED has (theoretically) infinite contrast, and the Wacom is a measly 1000:1, they match perfectly both in color and bright/dark distribution of values, all without the OLED having to artificially raise its black levels. The best I can describe it is that they look exactly the same, but the OLED is prettier, due to its perfect black level of 0.

    Another part of my solution was investing in the Calibrite Display Plus HL, which unlocked Calibrite’s newer Profiler software, offering me more options to fine-tune my OLED display. For example, I was able to set a manual black level to “0” in profile settings. Before, in the Wacom Color Manager, this was limited to “0.3.” DisplayCal might have options such as this as well, but to be honest the program can feel so complex that, considering my time constraints, Calibrite Profiler was perfect. I imagine one might be able to dial things in just slightly better in something like DisplayCal, but CP is just so user friendly, I have to mention it here as an alternative option for anyone else struggling with this issue. On top of unlocking CP, the Display Plus HL has a better sensor than the i1Display Pro’s, so pretty soon, I hope to accurately calibrate the monitor’s HDR as well.

    Anyway, to calibrate the OLED monitor in Calibrite Profiler, I selected a target D65, 120nits, sRGB gamma, and a zero-luminance custom black point. I also selected OLED as the backlight type. After using those settings, I was able to dial in the RGB white balance and brightness using my on-screen-display, and sure enough, I got the result I wanted.

Viewing 12 posts - 1 through 12 (of 12 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS