monitor profiling weirdness

hey guys,

i recently started working on my monitor profiles since i bought a new graphics card and i'm about to start editing and colour grading a documentary in november. usually i'm only working (photography) on my calibrated screen, but for editing and grading i added an older (cheap) screen as well as a crt preview monitor to the mix. i wanted to work in rec709 colour space and to have (at least some) colour consistency between these three monitors. i started reading and experimenting and while i made good progress at first, i'm kinda stuck and confused right now.

my problem is that i seem to get weird (very greenish) results when trying to calibrate any of my screens to D65 (or any other white point than its native). it does not matter, whether i use dispcalgui/i1profiler to manually set white point (RGB controls on monitors), or use the native white point of the monitor and let the profile do the rest. i end up with a green result. so green that it hurts my eyes and i'm pretty sure it cannot be a matter of getting used to a new white point.

if any of you has experienced similar problems with adapting, and you think this is a case of user error, please stop reading here and just let me know (but please keep in mind that my case is pretty severe - the weird green appearance persists even after going to work and coming back - and the screens look green in any lighting condition and compared to any other device i have at hand - iphone, macbook pro, macbook air,..). otherwise please continue reading the whole story:

my setup:
main screen: eizo sx2762w (1 year old)
second screen: acer al1906 (very old)
video monitor: sony triniton broadcast monitor (probably even older)
colorimeter: x-rite i1 display 2 (3 years old)
graphics card: geforce gtx760
os: mac osx 10.9 mavericks
software for calibration: dispcalgui+agryllcms and i1profiler lion edition

a little background - here's how colour management started for me: i had bought my i1 display when i was still on my old acer screen. i remember playing around with different settings. iirc i wanted to profile to D65 but couldn't get it to look right (i was put off by the strange green cast i perceived - thought it was the acer's fault..), so i decided to go with native white point and lived happy ever since. last year came my eizo screen and another round of playing with profiling. i again ended up with native white point (because of the green tint). basically all my screens (even laptops) are profiled with native white point. i decided to live with the slight difference in colour as i almost never use them side by side...

a few days ago things changed (the above mentioned three monitor setup) and i decided to do something about colour consistency. i installed dispcalgui and started experimenting. i tried it all: screens calibrated to D65, eizo at native and trying to match acer, the other way round,... i never managed to get an acceptable result. either one, or all of my monitors where greenish. of course i know that i will never reach a perfect match between my eizo and a years old and cheap acer - but i was hoping to be able to get close enough. after some more experiments i decided to live with some inconsistencies but still at least i wanted to have monitors calibrated to D65 (monitor space for photo work and rec709 for video work). and that's where i'm stuck right now. i even cannot get my eizo to look acceptable in D65, neither with dispcalgui nor with i1profiler - that green again. btw, this is your second chance to tell me i would have to wait for my eyes to adapt, but if you still think something else might be wrong continue reading - i tried my best to evaluate the situation and will present some facts now:


my eizo looks about right as is (with standard profile from osx and any profile calibrated at native white point) at any possible white balance setting i can choose (of course it starts looking warm below 5500). 6500 looks very pleasing. as soon as i try to calibrate to D65 or any other setting than native things start to get weird.

finding 1: take a look at the curves that i1profiler shows at the end of the calibration process.

monitor set to 6500 kelvin - calibration set to native
monitor still at 6500 - calibration set to D65:
finding 2:
this one correlates somewhat with finding 1 - and, at least to me, proves that something is wrong...

here's the calibrated reading from within dispcalgui using the profiles from above:

monitor set to 6500 kelvin - calibration set to native:

Current calibration response:
Black level = 0.1795 cd/m^2
50%   level = 27.62 cd/m^2
White level = 125.03 cd/m^2
Aprox. gamma = 2.18
Contrast ratio = 697:1
White chromaticity coordinates 0.3714, 0.3326
White    Correlated Color Temperature = 3898K, DE 2K to locus = 19.4
White Correlated Daylight Temperature = 4000K, DE 2K to locus = 20.5
White        Visual Color Temperature = 4472K, DE 2K to locus = 18.4
White     Visual Daylight Temperature = 4471K, DE 2K to locus = 19.9
The instrument can be removed from the screen.



monitor still at 6500 - calibration set to D65:

Current calibration response:
Black level = 0.1795 cd/m^2
50%   level = 22.22 cd/m^2
White level = 101.59 cd/m^2
Aprox. gamma = 2.19
Contrast ratio = 566:1
White chromaticity coordinates 0.3130, 0.3293
White    Correlated Color Temperature = 6485K, DE 2K to locus =  4.6
White Correlated Daylight Temperature = 6486K, DE 2K to locus =  0.0
White        Visual Color Temperature = 6322K, DE 2K to locus =  4.4
White     Visual Daylight Temperature = 6486K, DE 2K to locus =  0.0


i'm pretty sure that something's wrong here, and i' guessing that either, my eyes (unlikely), screen (unlikely - i'm having the same problem with the old acer - tests above yield same results, although the courves are not as straight, which was expected) or my colorimeter are completely off target. clearly the i1 display 2 is my prime suspect, but what i find a bit weird is that at my screens' native white points it seems to do a great job...!?

unfortunately i do not have a second colorimeter at hand (yet) to proove or falsify my suspicions. i thought i'd ask here before going out and buying a new one - you know, this wouldn't be the first time that it's actually me who's off target. what do you think?

sorry for the long post!
thanks a bunch,
joerg

The first suspect would be a common procedural error during calibration across the different monitors. But from your description so far, I'm not seeing evidence of that. So to me the next most obvious problem seems likely a bad i1 Display 2 colorimeter.

I wouldn't necessarily buy a new colorimeter immediately, unless you can find a shop that will let you return it if you end up not needing it. Try to borrow from somebody instead, if possible. But in my opinion you should try a 2nd colorimeter (or spectro) before spending any more time on troubleshooting. You're describing a consistent problem across 3 different monitors and 2 different calibration software apps. You need to rule the colorimeter in or out as the source of the problem; if you can't trust it, then anything else you try to sort out with your process is wasting time. These devices do go bad, and it's even possible that yours was defective from the beginning.

Also, simplify your process. Just try to get the Eizo monitor calibrated first without spending time on the others. The Eizo is newer and likely in better shape for colour critical work, and if you can't get it sorted then something fundamental is wrong. The other monitors may have problems of their own, and throwing them all in the mix is not helping you correct the procedural problem at this point.

A 3rd option for the source of the issue, is that something with your Mac OS (or some installed video "helper" app) is haywire and is interfering in a major way with the calibration process. But the info you showed looks to me more consistent with colorimeter readings that are way out.

thanks for your reply, royce. i was hoping you would be reading my thread and i'm happy that my conclusions so far don't seem too be to far off.

just tried calibrating the eizo on windows 7 - again with a greenish result.

hopefully i'll be able to borrow another meter tomorrow. i'll post an update asap.
thanks again,
joerg

It would suck to have to buy a new calibration device, but if you value your time at all that's probably the cheapest route. Otherwise from your description there's going to be some serious debugging to figure out what's going on, if it's not the colorimeter...

mistery solved: as we both suspected, the colorimeter was causing the problem. i went and bought a new i1 display pro today and everything works as expected - no more green tint!

thanks again royce, for helping me out here and confirming my suspicion!

Good news. Pity the previous device died but at least the new i1 Display Pro is even better.