Calibration seems too magenta

Home Forums Help and Support Calibration seems too magenta

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #20499

    DenK
    Participant
    • Offline

    Hi there. After calibration, my display seems to be too reddish / magenta. This is, of course, especially notable in skin tones.

    I’m calibrating a LG 32UD99-W which is connected via Blackmagic Intensity 4K. From what I can tell, it already becomes too red  at the veryfirst step of calibrating, when I manually adjust the colors on the display. But DisplayCal tells me that I’m pretty damn close to Delta 0. But still, I can’t help but think that it’s just too much magenta. Maybe I chose some wrong settings for the calibration (Lut for Resolve).

    Is there any way I can verify my calibration? I could also upload the report if anyone might be willing to take the time to look at it.

    Thanks a lot!

    #20500

    Florian Höch
    Administrator
    • Offline

    Hi,

    are you using the correction “Spectral: LCD PFS Phosphor WLED IPS, 94% P3 (Panasonic VVX…)”?

    #20510

    DenK
    Participant
    • Offline

    Hi Florian, thanks for your quick reply. I did indeed not use the correction. I’m profiling again at the moment with the correction and will let you know when it’s done.

    #20520

    DenK
    Participant
    • Offline

    I did a completely new calibration and profiling with the correction applied. It seems to be a bit better, but I’m certain now that there’s still a red / magenta tint. I really can’t wrap my head around that at this point.

    Also, I just calibrated an LG 49WL95C-W, and it shows a lot cooler, more neutral looking picture. This one is connected directly on my GTX1080ti, while the other one is connected via Blackmagic Intensity 4K. Btw, I’m using the XRite i1 Display Pro.

    Any other ideas would be greatly appreciated!

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

    #20536

    Florian Höch
    Administrator
    • Offline

    See the entry about visual matching in the FAQ.

    #35611

    Kirk
    Participant
    • Offline

    Old post, but this might help someone if not the original poster, and I still have some lingering speculations that are unresolved myself.

    I adjusted my 32ud99 RGB using the DisplayCAL interactive whitepoint feature. The first time through I was just testing to make sure my ColorMunki  and DisplaCAL setup still worked, and forgot that I had the  second monitor with the interactive panel pointed right at the 32ud99. The second monitor significantly had a bright blue background. I ended up with red turned all the way up, which produced a very pinkish tint. Having already read this post I feared the worse. The second time through I pointed the second monitor away from the 32du99, used a black background and set the brightness and contrast so I could just barely see it in the pitch dark. This allowed me to lower the red while still achieving the target 65k, but the weird thing which may be significant here, the interactive display showed the same whitepoint result while lowering the red from 100 all the way back down to 50. At 49 the red in the interactive panel lowered and the whitepoint changed. I raised the red back up to 50, and the result looks very good. The two things I take from this which may apply to the post here, it’s easy to get the wrong initial RGB hardware setting from ambient light, and once the red is to high, it’s difficult to see that it is too high, because even when the ambient light issue is fixed, DisplayCAL shows the target whitepoint is still good.

    I wonder if this is a common scenario with calibration and I should be satisfied with the result, or does this indicate the ‘Spectral: LCD PFS Phosphor WLED IPS’ correction may not be ideal? Reading in this forum and elsewhere that the 32ud99 backlight is a mystery leads me to wonder about which correction to use. PCMonitors speculates the 32ud99 uses gb-led/gb-r, but I can find no other speculation about this, and LG simply says it is wled, which apparently does not accurately define the backlight since there are more than one variant of wled. In the absence of more info, I find no real reason to doubt using ‘Spectral: LCD PFS Phosphor WLED IPS’, but just in case someone does have more information on this after three years I thought I’d through this out there.

    Thanks,

    Kirk

    #35612

    Vincent
    Participant
    • Offline

    I belive that you have a colormunki display, the colorimeter.

    i1d3 “distributed” colorimeter correction with no direct reference device on your screen to correct colorimeter, relies on two things
    -an accurate spectralpower distributiion sample, a CCSS
    -accurate sensor spectral sensivity stored in colorimeter firmware.

    For 95% P3 displays like yours with almost P3 green and P3 red the known backlights that gives you such primaries are: “Spectral: LCD PFS Phosphor WLED IPS, 94% P3 (Panasonic VVX…)” and some QLEDs.
    Measure display primaries without correction. White reading may be innacurate but you’ll get a hint of primaries.

    2nd factor relies on colorimeter QC and aging. i1d3 is not know to age fast, unless you damage it its filter response should not vary in a decade
    BUT… maybe Xrite is not applying the same QC to munki batches as they do on i1d3pro. Instead measuring filter sensivity curves each X number or per batch they use a more relaxed approach. There is no way to know it.

    Also your display seems to have a weird respnse of RGB gains.

    So… how to proceed?
    -use the correction suggested by Florian for 95% P3 displays
    -try to use some OSD preset that is close (<10dE) to D65
    -use full GPU white point correction instead RGB gains to ensure that there is no weird RGB gains behavior, just to test colorimeter accuracy
    If that results on a whiter look, there is an issue with monitor firmw or RGB gains and it’s hard to tell without using it, you’ll have to test. If you keep >90% unique grey values with full GPU white point correction maybe it is not a bad deal.

    if you get the same pinkinsh tint then may be 32ud99 uses other backlight (try to rent an spepctrophotometer and share CCSS with community) or that colorimeter firmware data does not mach actual colorimeter response (use visual whitepoint editor or rent an spectrophotometer to make a CCMX matrix) or your own visual system is not  very close to std observer (use visual whitepoint editor).
    If you use a xrite spectrophotometer use high res 3nm mode, since WLED PFS has very narrow spikes.

    • This reply was modified 1 year, 9 months ago by Vincent.
    #35617

    Kirk
    Participant
    • Offline

    Thanks Vincent. That’s quite a comprehensive reply! Yes I have a ColorMunki Display colorimeter. I look forward to working through this, although I might not be able to find the time right away.  I guess I will need to update DisplayCAL, ArgyllCMS and HCFR, which I am thinking may be the easiest way to measure the primary colors. Unless there is an easy way to do this in DisplayCAL now? With the slow speed of ColorMunki I hope I don’t have to profile everything to get the primaries.

    #35618

    Vincent
    Participant
    • Offline
    1. Argyll’s cmdline “spotread” over 255-value patches on MS Paint
    2. Maybe tools \ report \ un calibrated display report
    3. Measurement report without simulating a colorspace, choose absolute values, choose xyY or XYZ, look for measurement values for 255 primaries

    (primaries measurement)

    • This reply was modified 1 year, 9 months ago by Vincent.
    #35848

    Lorycrog
    Participant
    • Offline

    Ive messaged a couple folks, but cant seem to find a mod to find out when I can get my contribution confirmed. I submitted a contribution 2 weeks ago, but havent heard anything, and my contribution shows nothing 0.

    Hoping someone can check on it for me.

    

Viewing 10 posts - 1 through 10 (of 10 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS