DisplayCal vs Apple TRC for 2012 MBP

Home Forums General Discussion DisplayCal vs Apple TRC for 2012 MBP

Viewing 15 posts - 1 through 15 (of 45 total)
  • Author
    Posts
  • #17885

    Wire
    Participant
    • Offline

    Not sure if this question is General or Help/Support.

    Attached are plots for the calibration curves for two profiles for an Apple 2012 Macbook Pro.

    One is the Apple supplied profile for this Macbook (Color LCD)

    The other (MacBookPro9,1 #1…) is a DisplayCal / DTP-94 generated profile which targets the sRGB preset. Whitepoint was chosen “As mesuared”, otherwise all settings are DisplayCal defaults.

    The color balance and of these profiles look almost identical.

    But the Apple-supplied profile has more contrast in mid blacks, and you can see in the plot that the Apple profile has a J hook at the very bottom of the calibration curves while the DisplayCal profile does not.

    The DisplayCal profile makes everything look lighter at bottom end, and if I use it for image editing, it will invoke a distinct difference in results from the use of the Apple profile: I will tend to produce images with heavier blacks using the DisplayCal profile.

    Which one is truer to sRGB?

    The Apple TRC generally looks better in the shadows. But the DisplayCal profile doesn’t look wrong. And when I do simple mid-tone visual alignment tests, they look like they conform to G2.2.

    It’s tricky to tell with this Macbook Pro because the TN panel has a strong vertical axis contrast variability.

    I would just chock this up noise, except for the following.

    I found that an Apple supplied profile for a friend’s LG DCI-P3 capable IPS 4K display differs from the DiaplayCal produced prodile in the same way. In other words the Apple versions for this Macbook and the LG have similar TRCs, with shadows pulled down, while the DisplayCal versions with same instrument are lighter in same way. So there’s repeatability across devices.

    And I recall that when I used this same DTP-94 with Coloreyes Display Pro, it tended to produce a similar difference in black tonality with this same Macbook.

    However, when I use this same instrument and config on a old Dell 2209WA IPS, for which there is no Apple version to compare, I get shadow TRC results that look more like Apple’s profiles. IOW, side-by-side, the DisplayCal results for the Dell agree more with Apple TRC for the Macbook and the LG!

    ???

    Does Apple do something special with shadows for some devices or is there something funny about my puck or DisplayCal?

    Any insights appreciated.

    Attachments:
    You must be logged in to view attached files.
    #17898

    Florian Höch
    Administrator
    • Offline

    The Apple profile is not accurate if you have created it using Apple’s visual calibration tool. You can verify it on the “verification” tab in DisplayCAL.

    #17899

    Wire
    Participant
    • Offline

    Did I make a mistake?

    The profile I attached is an Apple-supplied profile. Not one made with the visual calibrator. It is what Apple thinks properly aligns this unit’s panel.

    #17900

    Florian Höch
    Administrator
    • Offline

    The profile I attached is an Apple-supplied profile

    Same applies. These profiles have an “idealized” TRC (= the actual sRGB curve) that is not based on measurements, and doesn’t match the calibration (the Apple profile calibration is likely aiming for gamma 2.2). Originally this mismatch was intended by the sRGB standard, but this is not how sRGB is usually used in a color managed environment.

    #17903

    Wire
    Participant
    • Offline

    Is it fair to say that Apple tailors the built-in alignment to suit the attractiveness of its displays—say punchy contrast—at the expense of standards accuracy?

    I am aware per Poynton 1886 that the industry has sort of punted on EOTF over the long haul, and the best presentation curve depends on the viewing environment. But let’s simplify by assuming that the sRGB TRC is acceptable for me regardless of environment.

    It seems this is a bit of a question of which to trust and why.

    If the answer to my first question, above is ‘yes’, then end of conversation. But if you are not sure, can you suggest an approach by which I can further examine this discrepancy, or is it an open/closed case of Apple just suiting itself?

    Can DisplayCal Verify tell me the Apple profile is divergent from the standard if it depends on the same instrument as was used to create one of the profiles in question?

    I’m pressing on this because the Macbook DisplayCal profile is lighter than the Dell 2209WA DisplayCal profile, and as I said before, a friend’s Apple compatible LG with Apple-supplied profile is a better visual match to the Macbook Pro Apple-supplied profile than the DisplayCal profile I generated for the LG using the same setup as for the Macbook. IOW, DisplayCal produces a lighter rendering than Apple.

    I have no axe to grind, I’m just looking for a way to determine which to trust re sRGB target?

    And as the whole modern ICC thing is largely an Apple creation and they made this game, I wouldn’t just assume they don’t care or haven’t thought it though. Looking across their products for decades they always seemed to care more about this than most other makes.

    So say Apple is pulling it’s built-in  TRC towards dark viewing environment assumption / video?

    In this sense I don’t know if I’m asking about DisplayCal or Apple

    #17906

    Wire
    Participant
    • Offline

    Not that it’s directly pertinent to my questions at hand, but for reference, here is where I first came across the argument for what became BT.1886, and  I assume WRT your comment about the subjectivity of sRGB output you are referring to this argument…

    http://poynton.ca/notes/PU-PR-IS/index.html

    (If the PDF is not readable in your browser—the fonts are messed up in mine—download it and use another viewer)

    Again while this is tangential to questions at hand, and it considers Rec.709, not sRGB,,  it puts the question of output intent into perspective as a subjective concern that got lost in the industry.

    So if Apple does it’s own thing, it seems like given the long-term state of the industry they would not be wrong to do so 🙂

    Hope this is not over complicating my questions. I’m trying to connect dots

    #17907

    Wire
    Participant
    • Offline

    Note my previous misuse of the term EOTF. In the above paper, the author is extremely careful with terminology and refers to and electro-optical conversion function EOCF for decode. My mistake

    #17908

    Wire
    Participant
    • Offline

    In case it got lost in all the factors I mentioned, the reason I even asked this question is because the appearance of the DisplayCal target sRGB result for some desktop Dell 99% sRGB IPS panels I have visually disagrees from the same alignment for these explicitly Apple-compatible MBP and LG displays. IOW, the Apple-supplied alignments for the MBP and LG agree more with the DisplayCal results for these Dells than than the DisplayCal results for the MBP and LG. This is very surprising. I would expect all the DisplayCal results to generally agree in TRC with the Apple-supplied alignments as the outlier. But weirdly it’s across the DisplayCal results that I find the discrepancy.  This oddness led me to wonder if Apple does something with their supported displays that DisplayCal does not account for?

    Everything else is academic.

    #17909

    Wire
    Participant
    • Offline

    Here are pictures showing what I’m seeing

    The MBP in the foreground with the lighter shadows is aligned with DisplayCal target sRGB, and with the darker shadows is the Apple-supplied alignment.

    The Dell in the background is aligned with DisplayCal target sRGB

    Why is the Apple-supplied MBP profile a better match to the DisplayCal Dell than the DisplayCal  generated MBP?

    Attachments:
    You must be logged in to view attached files.
    #17916

    Vincent
    Participant
    • Offline

    Is it fair to say that Apple tailors the built-in alignment to suit the attractiveness of its displays—say punchy contrast—at the expense of standards accuracy?

    No. Apple does the same as other display vendors: provide a generic profile with native RGB coordinates (maybe from EDID) and TRC curves that minimize rounding erors in color managed applications when no user calibration is applied.

    They do not care abou accuracy, they care about you not being able to notice it with default setup.

    It seems this is a bit of a question of which to trust and why.

    […]

    I’m pressing on this because the Macbook DisplayCal profile is lighter than the Dell 2209WA DisplayCal profile, and as I said before, a friend’s Apple compatible LG with Apple-supplied profile is a better visual match to the Macbook Pro Apple-supplied profile than the DisplayCal profile I generated for the LG using the same setup as for the Macbook. IOW, DisplayCal produces a lighter rendering than Apple.

    […]

    +

    In case it got lost in all the factors I mentioned, the reason I even asked this question is because the appearance of the DisplayCal target sRGB result for some desktop Dell 99% sRGB IPS panels I have visually disagrees from the same alignment for these explicitly Apple-compatible MBP and LG displays. IOW, the Apple-supplied alignments for the MBP and LG agree more with the DisplayCal results for these Dells than than the DisplayCal results for the MBP and LG. This is very surprising. I would expect all the DisplayCal results to generally agree in TRC with the Apple-supplied alignments as the outlier. But weirdly it’s across the DisplayCal results that I find the discrepancy.  This oddness led me to wonder if Apple does something with their supported displays that DisplayCal does not account for?

    The answer is in your first message:

    The other (MacBookPro9,1 #1…) is a DisplayCal / DTP-94 generated profile which targets the sRGB preset. Whitepoint was chosen “As mesuared”, otherwise all settings are DisplayCal defaults.

    […]

    And I recall that when I used this same DTP-94 with Coloreyes Display Pro, it tended to produce a similar difference in black tonality with this same Macbook.

    You are NOT using a very accurate measurement device for current backlights, hence you cannot expect visual match.
    It is not DisplayCAL’s fault or LG or Dell or your macbook.

    So:

    -rent a spectrophotometer, and if it is an Xrite one (10nm) make sure it is using High resolution mode (3.3nm), then make a suitable correction for your DTP94 and each display type you are going to measure with it (LG P3, macbook 2012, old CCFL dell)

    or

    -get/buy/rent/borrow an i1DIsplayPro/ColorMunkiDisplay/other i1d3 and use bundled spectral corrections for WLED PFS P3 displays (LG P3) or sRGB-like WLEDs (macbook from 2012? maybe there is a community made CCSS for it). Dell 2209WA should use standard sRGB-like CCFL correction.

    In long term 2nd option is cheaper … and it’s very fast.

    • This reply was modified 4 years, 10 months ago by Vincent.
    • This reply was modified 4 years, 10 months ago by Vincent.

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

    #17920

    Wire
    Participant
    • Offline

    One last bit of data:

    Here are Colorsync Utility 3D plots which compare the gamuts of the Apple-supplied vs the DisplayCal DTP-94 generated profiles.

    One image is the overlay of the DCal DTP-94 measurements (colored) overlaid the Apple-supplied (whte) for the 2012 Macbook Pro LED sRGB.

    The other is the Apple-supplied for the LG Ultrafine DCI-P3 (colored) overlaid on the DCal DTP-94 measurements.

    I ended up with two diff plot styles by chance.

    I tried a DCal verification report of 2012 MPB with a syth comparison with the Apple profile and get <1–3 dE  across the board. (attached)

    And when I soft-proof the Apple and the DCal DTP94 in Photoshop against Linbloom’s 16 million colors, I get excellent correspondence.

    So it is possible for the DTP-94 to be sufficient to measure this kind of correspondence with Apple spec gamuts for these two different techs, and to produce a visually excellent color-balanced calibration, but get the shadow tonality calibration curve wrong due to DTP-94 inaccuracy with LED?

    Attachments:
    You must be logged in to view attached files.
    #17927

    Vincent
    Participant
    • Offline

    Since DTP-94 may not be able to measure properly sRGB LEDs or P3 leds this kind of comparisons are useless.
    Also gamut boundaries may be close between those profiles you compare, but grey axis in brightness (speed for increasing L* varying input) and color (a*b*) could be off, very off, and won’t be noticed in that plot … that’s why you use DisplayCAL profile verification.

    BTW you and use DisplayCAL profile info to generate these display profile L*a*b* plots against some arbitray ICCv2 profile that you use as reference: VRML, interactive, rotations… etc

    • This reply was modified 4 years, 10 months ago by Vincent.
    #17929

    Wire
    Participant
    • Offline

    Thanks for the help guys.

    #17938

    Wire
    Participant
    • Offline

    I found this thread where you guys talk about compromises of low display dynamic range and colorimeter sensitivity trade-offs for getting TRCs right for a specific gamma.

    Black shadows and gamma calibration

    Black shadows and gamma calibration

    My reading of the discussion is that for a given power law TRC that there are contrast ratios that can’t fit onto the curve. And, during calibration, because gray balance is very important and at display black balance can be achieved only by adding light to get gray right, this creates a situation where a TRC fit to the desired gamma and gray balance have to be traded off. Is this correct?

    It looks like DisplayCal has two controls pertaining to this: “Black output offset” and “Black point correction”.  Do these settings directly affect calibration, or do these pertain on;y to how a CMS does translation from an image color space to the display color space?

    Can you tell me what visual effect I can expect from running Black output offset at 0 vs 100?

    DisplayCal on Mac reports that the Apple CMS doesn’t support Black point correction, but is supported in Adobe CMS.  Can enabling this help improve color correction in Photoshop, even if MacOS built-ins like Preview won’t show the results? Put another way, if my Photoshop working  space is sRGB, can using Black point correction for my system display profile help me produce more accurate display referred images, with the caveat that using MacOS built-ins may not produce the same results?

    #17948

    Wire
    Participant
    • Offline

    Te answer appears to be not the DTP-94, but that Apple does something non-standard with the graphics for the MPB.

    Spend a good day reviewing my setup, learning how to do DisplayCal verify, and CMS assumptions of my tools. Reset displays and compared with two more old IPS Dells in good working order. 5 displays agree, and agree with my memory of the LG but I will have to revisit.

    Everything I see through the DisplayCal alignments looks great and behaves as I would expect WRT changes in target gamma. Verify says everything checks out.

    The Apple-spplied  Macbook Profile has a TRC that does something unconventional in the deep shadows. They clamp near blacks in a way that makes the display look punchier with miniscule loss of lowest detail, and without blowing the rest of the curve or the color.

    Editing with Apple’s tools, won’t let the user easily fiddle with the shape of any curve in those tones, and editing to taste on the unit has the effect of producing images with a tad more shadow detail on export than a pure standard match, plus it lets the device work as both a web editor (mid / high tone sRGB) and a movie viewer (slightly darker 709 2.4 shadows). Soft-proofing the Apple profile appears to be identical to sRGB gamma. So as far as Apple is concerned, with their cal they win!

    I treid doing a DisplayCal Simulate report to compare the Apple calibration against a DisplayCal sRGB target cal, but I got confused.

    Nothing at the moment otherwise suggests my DTP-94 has issues and 5 displays all agree and look great according to everything I can send them.

    Thanks for a super interesting and great program! And thanks again for the help along the way

Viewing 15 posts - 1 through 15 (of 45 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS