MSI PS321QR Calibration Report

Home Forums General Discussion MSI PS321QR Calibration Report

Viewing 11 posts - 1 through 11 (of 11 total)
  • Author
    Posts
  • #26901

    Ian de Vera
    Participant
    • Offline

    Hi guys,

    Just got my hands on one of the newest wide gamut monitors from MSI. The Creator PS321QR.

    Here’s my calibration report using X-rite I1 Display pro (finally got one!).

    I used the Hp dream color correction as recommended by Vincent in one of the threads here. Results seem good (see attached measurement).

    I just have one question, how would I know if this is the proper correction? Given that there is no correction available yet for this relatively new monitor.

    All i know is that it is a AUO panel, white LED backlight, wide gamut.

    Also, if I use single curve + matrix, I never get a good measurement report. Have to use XYZ LUT + matrix. Is this normal?

    Thanks in advance.

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

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

    #26905

    Vincent
    Participant
    • Offline

    Grey shows neutral errors, although some color managed apps with dither may compensate it. Try redoing it with calibration speed to low.

    Regarding correction, plot the one you are using with “i” button next to correction. That is spectral power distribution (SPD) of an AdobeRGB+P3 WLED PFS.

    What is yours?

    -check in a review that plots SPD that match that graph. DisplayCAL graph is 1nm resolution, 10nm resolution plot form a review can blur all the spikes (and measured coordinates too)

    -check if somebody uploads a CCSS correction for your model. Plot PSD, see if it matchs generic 1nm. Same resolution issues as above

    -buy/rent an spectrophotometer, for Xrite ones make sure that you use 3nm “high res mode” to avoid “blur” SPD in red channel spikes. Again, check if it matchs.

    If it matches, the smaller resolution (1nm) the better

    #26912

    Ian de Vera
    Participant
    • Offline

    Okay vincent, will try to do low calibration.

    one thing that really bothers me is why I can never get a good calibration with single curve + matrix +BP comp. My XYZ LUT calibrations have good measurements but the SC+matrix always have bad values no matter what correction I use. (see attached measurements)

    I can simply stick with XYZ LUT + matrix except that it sucks that my blacks are crushed on mac OS X due to it’s bug with adobe CMM.

    Attachments:
    You must be logged in to view attached files.
    #26915

    Vincent
    Participant
    • Offline

    Matrix + 1 TRC  validated against display = MATCH means that display has a very “ideal” behavior so it can be described in a simple way = GOOD

    It you can get a match that way means monitor does not behave so well, so you have to map in a cube (RGB “in” ) what are actual color coordinates (CIE XYZ “out”) and vice versa and try to interpolate between those nodes in that 3D cube. The worse display is the highter the numbers of nodes needed to “precict” display behavior in an accurate way.

    You can even think of that matrix + 1 TRC as the simplest cube, and a a 5k patches XYZLUT profile as a 17 node per side cube mapping display irregulaties.

    Unfortunately while working on limited precision on color management that extremely detailed cube causes some kind of rounding errors that may be visible as banding on color managed gradients. Hence all good monitors with HW calibration (Eizo + color navigator for example) try to use the simpler profile to avoid those issues and since monitor hava  very ideal behavior that simple description is actually a match to display behavior.

    • This reply was modified 3 years, 4 months ago by Vincent.
    #26917

    Ian de Vera
    Participant
    • Offline

    I see,

    Thanks for explaining it.

    So I am getting it right?

    The problem is in my hardware? That the monitor does not behave ideally hence I can’t get a good correction with Single Curve + matrix?  And at the end of the day I should stick with XYZLUT+matrix for color critical work?

    Thanks so much for patiently answering my questions vincent

    #26918

    Vincent
    Participant
    • Offline

    Yes. Keep matrix one as default so Apple CMM does not make weird things, switch to the other for PS/LR.. even for video in MPV app if you can manage to configure it.

    #26919

    Ian de Vera
    Participant
    • Offline

    Thank you vincent I will do this

    will have two ICC profiles, the single curve +matrix for default.

    and ill just switch to the XYZ LUT + matrix when editing in PS/LR

    Thanks for the advice on getting X-rite. My next monitor upgrade will definitely be an Eizo, will just have to figure out how to get one.

    CHeers fromt he Philippines

    #26940

    Ian de Vera
    Participant
    • Offline

    I did a calibration report and it does not look good.

    Is this bad or acceptable/expected for these type of displays?

    Attachments:
    You must be logged in to view attached files.
    #26950

    Vincent
    Participant
    • Offline

    I would say that uniformity is expected for these low cost AdobeRGB widegamuts. Regarding aceptable or not I would say no buy YMMV, it is not just a single cellgrid with 3.xdC, there are a lot of them, but IDNK if you place things there. On white screen (or transitions to white screen) like in movies/tv shows or games it shpuld be very noticeable.

    Report says NOT OK, buy try those situations described above, or check if an A4/A3 white canvas in some photo/design app of your choice shows ugly uniformity issues => evaluate it visually. I assume that you did an uniformity report because you already saw these color tints…

    Unfortunately widegamut uniformity is QC and QC is to discard panels or going to manufacturers with better QC and that means money (Eizo CS or NEC pa or better)… or use bad panels and then apply an extremely agressive uniformity compensation that destroys contrast (1000 -> 600:1) like Benq does in SW line with C suffix.

    • This reply was modified 3 years, 4 months ago by Vincent.
    #26952

    Ian de Vera
    Participant
    • Offline

    I would say that uniformity is expected for these low cost AdobeRGB widegamuts. Regarding aceptable or not I would say no buy YMMV, it is not just a single cellgrid with 3.xdC, there are a lot of them, but IDNK if you place things there. On white screen (or transitions to white screen) like in movies/tv shows or games it shpuld be very noticeable.

    I see. Guess it is expected then for a monitor this price. Yes I noticed the cells with very bad results are relatively the areas I rarely place things unto. This is a relatively large monitor and I usually just use the central portions

    Report says NOT OK, buy try those situations described above, or check if an A4/A3 white canvas in some photo/design app of your choice shows ugly uniformity issues => evaluate it visually. I assume that you did an uniformity report because you already saw these color tints…

    Hmm. actually I just did a uniformity test out of curiosity. On actualy usage it is barely noticeable except on the extreme corners. Which even on movie viewing I don’t actually notice unless I force myself to notice it.

    Unfortunately widegamut uniformity is QC and QC is to discard panels or going to manufacturers with better QC and that means money (Eizo CS or NEC pa or better)… or use bad panels and then apply an extremely agressive uniformity compensation that destroys contrast (1000 -> 600:1) like Benq does in SW line with C suffix.

    Oh I see. I’ve been looking on the price range of Eizo monitors. It’s actually not that bad. Only problem is availability. I risked buying the i1display pro from adorama USA and had it shipped here to the Philippines because the item is relatively inexpensive that I could just forget about it if I were to have problems. With the Eizo monitor I don’t think I could do the same. Maybe when I am able to go to US and personally purchase it.

    Until then, I think I am relatively happy editing my photos with this one. Just being over obsessed with the numbers I guess.

    thank Vincent. I’ve learned a lot about colors and using displaycal.

    #27238

    Ian de Vera
    Participant
    • Offline

    Just wanted to update this thread.

    So for some reason I decided to try out upgrading from Mojave to Catalina.

    Lo and behold! In catalina I get good results using single curve + matrix profile!

    Life is good and I will be keeping this monitor 🙂

Viewing 11 posts - 1 through 11 (of 11 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS