Colours changing before calibration

Home Forums Help and Support Colours changing before calibration

Viewing 13 posts - 1 through 13 (of 13 total)
  • Author
    Posts
  • #33689

    fleuret50
    Participant
    • Offline

    Hi,

    Firstly I must say that I’m new in screen calibration. I recently bought a second-hand Datacolor Spyder4 Pro, and I’m trying to calibrate an iiyama ProLite X2474HS-B1 monitor connected to a MacBook Pro 15,2 (2019) running macOS Catalina and using DiplayCAL 3.8.9.3 and ArgyllCMS 2.3.0. I’m particularly interested in high-contrast B&W photography.

    To do so, I’ve used the following parameters:

    • Settings: Default (Gamma 2.2)
    • Instrument mode: LCD (generic)
    • Correction: Spectral LCD White LED family (AC, LG, Samsung)
    • Whitepoint to 6500 K
    • White level to 120 cd/m2
    • Tone curve: Gamma 2.2
    • Calibration speed: High
    • Profile quality: High
    • Testchart: Auto-optimized with 175 patches
    • Calibration was done at night, with no light coming at the monitor. All dynamic parameters were turned off in macOS preferences.

    I ended up with a profile that, once loaded for my monitor colour settings in macOS, displays a green tint all over the screen. In particular, when looking at a B&W photography, the grey seems “greenish”.

    What is really intriguing me is that, with the profile loaded and when I start a new calibration, there is a noticeable change in the colours displayed BEFORE the calibration process begins. You can see it (I hope) on the two images attached: 1) with the profile loaded; and 2) with the profile loaded and the “Interactive display adjustment” running. On the second image, the black and white seems not greenly tinted. Indeed, that change also happens when running the Verification process.

    Does anyone knows why there is a change in the colours displayed when running the calibration process? Doesn’t it affect the calibration process itself?

    I also attach the profile generated and the verification file. Any help would be appreciated!

    Attachments:
    You must be logged in to view attached files.
    #33712

    Vincent
    Participant
    • Offline

    You have a chain or errors due to your hardware.
    First of all all spyders are innacurate, hence whatever you measure, these coordinates may be far from actual ones.
    Second your OS’ color management engine (macOS) is faulty, you have to stick with default profile types (idealized 1TRC matrix type with fake infinite contrast) because your OS cannot use other types. DIsplayCAL should have warned you fist time.

    So
    -get an i1d3 colorimerter
    -use default profile settings if you use macOS

    Regading calibration process
    1- fix white, VCGT (grey calibration ) cleaned
    2-automatic iterative process to fix grey (VCGT “semi” cleaned, actual grey being tested is subject to VCGT changes)
    3-calibration process ended, calibration loaded (VCGT) to GPU LUTs, then measure calibratd profile to build ICC profile.
    All apps that use GPU calibration must do the same.

    #33728

    CALmelion
    Participant
    • Offline

    Hi,

    I have gone down the same Spyder path – and am still on it. The price of an i1d3 was just too high (>2x), not knowing what you’ll actually get out of it.

    Firstly, some change in color tint and temperature is normal and known to be difficult to get used to. But the greens the Spyder4 was giving me as well is just not natural. Without DisplayCAL we’d be out of luck, but there’s some optimizations to be done in DisplayCAL (which i’m still figuring out myself).

    What seems to help a little for me is to turn on “Show advanced options”, then set “Black point correction” to 0% in Calibration, and use a “Single curve + matrix” under Profiling. I believe “black output offset” under gamma is best left at 100%.

    Another thing you could do is “Import colorimeter corrections from other display profiling software…” under Tools > Correction and use a more fitting Instrument mode.

    Hope it gives you something to go on.. just trial and error.

    #33783

    fleuret50
    Participant
    • Offline

    First of all, thank you both for your replies!

    You have a chain or errors due to your hardware.
    First of all all spyders are innacurate, hence whatever you measure, these coordinates may be far from actual ones.

    Well, I saw that after having bought my Spyder4… Don’t you think I can get a pretty good calibration? After all, people succeed in using this harware, right?

    Second your OS’ color management engine (macOS) is faulty, you have to stick with default profile types (idealized 1TRC matrix type with fake infinite contrast) because your OS cannot use other types. DIsplayCAL should have warned you fist time.

    In fact, DisplayCAL warned me, so I clicked “Yes” and another profile was created using “single curve + matrix” and “Black point compensation”, however it ended up the same (green tint in grey colors).

    So
    -get an i1d3 colorimerter
    -use default profile settings if you use macOS

    The first time I ran calibration, default settings were used, same ending.

    Regading calibration process
    1- fix white, VCGT (grey calibration ) cleaned
    2-automatic iterative process to fix grey (VCGT “semi” cleaned, actual grey being tested is subject to VCGT changes)
    3-calibration process ended, calibration loaded (VCGT) to GPU LUTs, then measure calibratd profile to build ICC profile.
    All apps that use GPU calibration must do the same.

    I am sorry, but I don’t understand here what you suggest for me to do. Isn’t VCGT part of creating a 3D LUT, which macOS can’t use natively?

    #33784

    fleuret50
    Participant
    • Offline

    What seems to help a little for me is to turn on “Show advanced options”, then set “Black point correction” to 0% in Calibration, and use a “Single curve + matrix” under Profiling. I believe “black output offset” under gamma is best left at 100%.

    Thanks for your suggestion. I tried it, but the results weren’t better.

    Another thing you could do is “Import colorimeter corrections from other display profiling software…” under Tools > Correction and use a more fitting Instrument mode.

    I tried using the different options in DisplayCAL to get a .ccss file, without success. I even tried to run the oeminst executable from ArgyllCMS on the installator for Sypder4 software and only get a “spyd4cal.bin”, which I cannot figure out how to use. I didn’t find any .ccss file on Google for my monitor or Spyder4…

    #33785

    Vincent
    Participant
    • Offline

    First of all, thank you both for your replies!

    You have a chain or errors due to your hardware.
    First of all all spyders are innacurate, hence whatever you measure, these coordinates may be far from actual ones.

    Well, I saw that after having bought my Spyder4… Don’t you think I can get a pretty good calibration? After all, people succeed in using this harware, right?

    No. Spyder4 & 5 are garbage

    Second your OS’ color management engine (macOS) is faulty, you have to stick with default profile types (idealized 1TRC matrix type with fake infinite contrast) because your OS cannot use other types. DIsplayCAL should have warned you fist time.

    In fact, DisplayCAL warned me, so I clicked “Yes” and another profile was created using “single curve + matrix” and “Black point compensation”, however it ended up the same (green tint in grey colors).

    That means that error is caused by calibration curves once loaded into GPU LUT.

    So
    -get an i1d3 colorimerter
    -use default profile settings if you use macOS

    The first time I ran calibration, default settings were used, same ending.

    Read above

    Regading calibration process
    1- fix white, VCGT (grey calibration ) cleaned
    2-automatic iterative process to fix grey (VCGT “semi” cleaned, actual grey being tested is subject to VCGT changes)
    3-calibration process ended, calibration loaded (VCGT) to GPU LUTs, then measure calibratd profile to build ICC profile.
    All apps that use GPU calibration must do the same.

    I am sorry, but I don’t understand here what you suggest for me to do. Isn’t VCGT part of creating a 3D LUT, which macOS can’t use natively?

    Get an i1d3 or better device, or run with slowest speed (iterative steps for grey calibration will take longer, 12->24->48->96 step as you change speed from faster to slower). If even at slowest speed + default profile settings for macOS you get tint in greys, get an i1d3 like everybody who wants accuracy.

    In the paragraph you quoted I have explained why you see color change in each step of calibration process (it varies as something is load or unload to GPU LUT)

    #33795

    CALmelion
    Participant
    • Offline

    It’s funny the things they write on their own site FAQ.

    Q: My previous Spyders gave very different measurements of screen temperature, one Spyder to another, on the same display. If I move to SpyderX, will it solve this issue?

    A: The consistency between multiple SpyderX units on the same display (inter-instrument agreement), or the same SpyderX when placed on the same display multiple times (intra-instrument agreement) is much more precise than with previous calibrators. Rather than having variations of as much as a thousand degrees Kelvin between calibrators, as was common for Spyder3 or 4, they now measure nearly ten times more accurately; accuracy between calibrations on the same screen is similarly improved.

    Q: I’ve never been quite satisfied with the color matching between my various displays after calibration. Will SpyderX solve this issue?

    A: SpyderX should certainly improve screen-to-screen matching between displays. Its more human eye-like way of viewing the screen, improved control of white balance, color gamut, and shadow detail means different screens should look much more similar to the eye after SpyderX calibration.

    Q: I’ve been happy with my current Spyder. Is there any reason I should upgrade to SpyderX?

    A: More accurate white balance and shadow detail, and the blazing speed of calibration with SpyderX are the top reasons.

    ..

    No need to read between the lines, they basically just come out and say it.

    I distinctly remember a bit about (green) tint as well – writing it off as the user getting used to a new normal. This was before the Spyder5 even existed, but i guess they’ve removed that.

    Note: The SpyderX came after the 5, the X is not a placeholder as is a common misconception.

    • This reply was modified 4 months ago by CALmelion.

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

    #33797

    CALmelion
    Participant
    • Offline

    What seems to help a little for me is to turn on “Show advanced options”, then set “Black point correction” to 0% in Calibration, and use a “Single curve + matrix” under Profiling. I believe “black output offset” under gamma is best left at 100%.

    Thanks for your suggestion. I tried it, but the results weren’t better.

    Sorry. 🙁 It appears that on MacOS the Single curve was already the option chosen.

    I guess the only other thing to try is the slowest calibration speed.

    And i wonder if enabling Black level drift compensation could possibly make any difference?

    #33798

    Vincent
    Participant
    • Offline

    SpyderX uses a filter type akin to i1d3, non degrading. This is a good upgrade, but SpyderX is not a good product because it cannot be upgraded in a “remote way” to new backlight types, while i1d3 can.

    Spyder 4 or 5 CCSS upgadeability is useless since as arequirement to make CCSS work, spectral sensivity curves of actual device and the ones supplied in firmware (sor Argyll or any other software can use them) must match, and they don’t (reasons go from filter drgrading to poor inter instrument agreement and huge filter mismatch with CIE observer).

    There is no reason to upgrade a XXXXX device to any spyder. Datacolor devices should be avoided, bad ROI, not future proof…

    #33803

    CALmelion
    Participant
    • Offline

    Certainly wasn’t advocating for another Spyder. As I’m sure they must have known about these inaccuracies, yet kept selling them for years. In that regard they are no better than cigarette companies.

    These answers just show how bad things truly were even according to them. Variations of as much as a 1000K, i mean isn’t some kind of accuracy there the most basic requirement? How does that not make it utterly useless at fulfilling its purpose. And how could that possibly go unnoticed for years. Hint: they definitely knew.

    #33960

    fleuret50
    Participant
    • Offline

    Hello guys,

    Thank you for your replies. I’m giving up on the Spyder4. I’m assuming Vincent is right and the filters of Spyder faded away…

    I’ll buy a second-hand X-Rite i1d3 and see what happens. Keep in touch!

    #34058

    fleuret50
    Participant
    • Offline

    Hi,

    I bought an i1d3 and calibrated my two screens (internal from the MacBook and external) with DisplayCAL and, indeed, results are very good! No green tint after calibration!

    So, thank you really much for your advices. Definitely, Spyder instruments are now out of date. Maybe they have been since their release…

    Anyway, the next step will be printing my photographs !

    #34060

    CALmelion
    Participant
    • Offline

    Very interesting, glad to hear it.

Viewing 13 posts - 1 through 13 (of 13 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS