Large Shifts in 0-10% RGB Color balance iMac

Home Forums General Discussion Large Shifts in 0-10% RGB Color balance iMac

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #23850

    swordinhand
    Participant
    • Offline

    I measured my whitepoint and used that for the chromaticity coordinates. Seemed to yield a better overall score. When I do this, I’m reading the native white point of the display, no?

    Is that why the “Measured vs. assumed target whitepoint ΔE*00” is so high? It’s assume Rec709, in this case?

    The bigger question is why are the 0-10% points in “Correlated Color Temperature” and “RGB Grey Balance” so skewed. Is this the nature of this display or is there a way to make this tighter?

    Otherwise do my numbers seem good?

    Attached is my report.

    Attachments:
    You must be logged in to view attached files.
    #23863

    Vincent
    Participant
    • Offline

    I measured my whitepoint and used that for the chromaticity coordinates. Seemed to yield a better overall score. When I do this, I’m reading the native white point of the display, no?

    Is that why the “Measured vs. assumed target whitepoint ΔE*00” is so high? It’s assume Rec709, in this case?

    Because MacOS color management engine is full of bugs and a way to avoid them ist to do not correct white (by default configuration)

    The bigger question is why are the 0-10% points in “Correlated Color Temperature” and “RGB Grey Balance” so skewed. Is this the nature of this display or is there a way to make this tighter?

     

    Check “combined Δa*00 and Δb*00 range” (RGB+ gray balance) to se it in a*b*. It looks like 0.5 a* oscilation, seems fine. Try to see a blakc to white in a non color managed enviroment (difficult under macos) to see if there are issues in grey. If they are not caused by very limited GPU fro calibration, a slower speed may improve it, but yours seems fine on paper, range ~1.27, OK

    #23868

    swordinhand
    Participant
    • Offline

    Check “combined Δa*00 and Δb*00 range” (RGB+ gray balance) to se it in a*b*. It looks like 0.5 a* oscilation, seems fine. Try to see a blakc to white in a non color managed enviroment (difficult under macos) to see if there are issues in grey. If they are not caused by very limited GPU fro calibration, a slower speed may improve it, but yours seems fine on paper, range ~1.27, OK

    That’s the stat that was bothering me most and also the crazy RGB balance below 15%, but I guess it’s so low down on an LCD you can’t really see it. However, when I turn black output offset to 0% it smooths out. However, in Resolve it makes the blacks look super washed out. So visually 100% black output offset looks more correct to my eye.

    Also, I’ve noticed that when I run the verification through resolve the results are slightly different and mostly a little worse, but this is subtle. When I run through resolve, should I be running a simulation and use the sim as the display profile, checked? What about the viewer LUT in resolve? It seems this gets disabled. Or should I set this to the LUT I’m using for the Device link profile? Or is checking through Resolve not the best way to check / accurate?

    I guess I just want to know that what I’m looking at is +/-10% correct, in Resolve. Meaning, I don’t need reference grade, but close enough that I can trust what I’m seeing and use common sense with scopes.

    Just to clarify this is my workflow:

    Profiled and created p3d65 2.6 gamma 3D LUT with XYZ LUT + Matrix, using Absolute Colorimetric with White Point Scaling and black output offset 100%. Installed Profile. In Resolve loaded created LUT into “3D Color Viewer Lookup Table”. Disabled “Use Mac Display Color Profiles for Viewers” in Resolve preferences. Using P3D65 as In / Out / Timeline color space in RCM workflow.

    Essentially, I grade with these p3d65 settings and then, when I’m done, turn my output color space to whatever I’m outputting (sRGB, Rec709, etc.) I’ve tested this with sRGB onto another computer that is calibrated to sRGB and it looks identical to the p3d65. Which means that my color management is working, no?

    I’ve attached the three reports I referenced above (black output offset 100%; black output offset 0%; through resolve black output offset 100%)

    Thank you for any insight!

    Attachments:
    You must be logged in to view attached files.
    #23873

    Vincent
    Participant
    • Offline

    You use Resolve as you did, as DisplayCAL FAQ says. You verify in DisplayCAL using a simulation profile and a device link which is a LUT3D equivalent

    Near black is difficult to correct. Your display shows an unusual low contrast for an IPS display, ~500:1, it should be twice that value. Also use P3 mac WLED PFS CCSS correction bundled in DisplayCAL.

    #23876

    swordinhand
    Participant
    • Offline

    You’re right! I looked into the contrast ratio and Apple says around 1000:1. I measured the uncalibrated display on both my iMac and an older MacBook Pro and low and behold, the MacBook is around 900:1 with the iMac around 500:1. So there must be something wrong with my screen. I also notice that the screen has cloudy or greenish white haze in some of the corners. Which definitely seems defective and might be altering the overall level. Does this sound right or is something wrong with my settings?

    I’m still under warranty, so I should just swap it out.

    The CCSS you’re referring to is this one right?: Spectral: LCD PFS Phosphor WLED IPS, 99% P3 (MacBook Pro Retina 2016)

    There’s another one called: Spectral: LCD PFS Phosphor WLED IPS (Apple 27″ iMac (Early 2019) (i1 Pro 2)) but it uses a Refresh (generic) mode.  which is for OLED or Plasma, right? It also seems to yield strange results. I downloaded this from internet.

    Finally there’s this one: Spectral LCD RG Phosphor LED (Apple iMac and iMac Pro (i1 Pro)) This one seems like it might be good too.

    Thanks for all of your help!

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

    i1Basic Pro 2 on Amazon  
    Disclosure: As an Amazon Associate I earn from qualifying purchases.

    #23893

    Vincent
    Participant
    • Offline

    The CCSS you’re referring to is this one right?: Spectral: LCD PFS Phosphor WLED IPS, 99% P3 (MacBook Pro Retina 2016)

    This one (bundled with DisplayCAL). The other ones may be from user, I’ve not checked them but user made CCSS are usually from i1pro/munkiphoto readings with limited resolution (10/3nm) with may be unsuitable for WLED PFS backlight and their narrow spectrao power distribution in red wavelengths

    Regarding the extremely limited contrast, return all display setup & profiles to default, disable all image enhancements from apple: true tone, auto bightness, night shift… all off, then try again. The possible causes are too many and there are a lot of unknowns to give you detailed help. I would even try apple support, it should not behaved that way.

    Also raise brightness in Imac to 50% or 160-200cd/m3 and measure contrast at that setting, maybe iMac is not able to keep “contrast window” if you go under some % setting of brightness. You’ll have to find that % value by yourself

    • This reply was modified 5 months, 4 weeks ago by Vincent.
    #23899

    swordinhand
    Participant
    • Offline

    Good to know about CCSS.

    I set up the monitor as you said, regarding contrast ratio and tried different brightness settings. As the display got over 200cd the contrast ratio started to go up. At max brightness I achieved almost 1200:1. But in my dimly lit room, it’s blinding haha, 500cd white. Also, the black level is now 0.47 cd, which I know is a limitation of LCD tech, but what I don’t understand is why my old 2012 15# MacBook Pro and my 2016 13″ MacBook Pro both achieve around 1000:1 at 80cd and 120cd, respectively? It still feels like a defect, especially given the cloudy corners. Is there a database of people who have measured their iMac 5Ks on this site or somewhere, where I can see what others are getting?

    #23900

    Vincent
    Participant
    • Offline

    Can you plot it like in prad.de reviews? Try to report it to apple support attaching that 2D plot (contrast and brightness vs % brightness slider), maybe there is something that you misconfigured. Also check that there is not light leak when you place colorimeter (but laptop behaves ok… so it seems unlikely)

    Usually (unless some kind of uniformity compensation or something like that) window contrast remains static at the same value like 900-1000:1 range across brightness range.

    #23915

    swordinhand
    Participant
    • Offline

    I will try and plot the contrast vs brightness. But the monitor seems defective to me. Given that my two MacBook pros are hovering around 1000, this can’t be correct for the iMac. You would think that the iMac would be superior to the MacBooks, right? But it would be nice if others might confirm these suspicions. Or if there was a user database somewhere???

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS