- This topic has 7 replies, 3 voices, and was last updated 1 year, 9 months ago by
.
-
Posts
-
Hi,
I have a problem when calibrating monitors with my Eizo EX3. When calibrating and verifying afterwards I always get a white balance offset dE from 10-15. The measured and true white point seems to have a big difference. For example when I set the white point for a calibration to 0.3157 0.3113 with the visual white point editor, calibrate and verify afterwards, the verified white point is 0.3125 0.3004.
When I just calibrate to D65 there is always a heavy green tint after calibration, which seems to be consistent with the y-axis offset that I can measure. It is consistent over every display I tried calibrating so far.
The “Display White LED” preset is used.Since I bought the EX3 used I don’t know if there is a hardware defect or if it is a user misconfiguration. I hope you can help me with my problem.
Thanks!Spyders are infamous for being inaccurate. You should buy a new probe from the i1 Display Pro family:
- X-Rite I1 Display Pro or Pro Plus;
- Calibrite ColorChecker Display Pro/Plus/HL
Stay away from the non-Pro/Plus versions, which don’t support hardware calibration through such software as Eizo ColorNavigator and NEC SpectraView II.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.I know they are not the best tool, but for sure they are not dE 10-15 bad for white balance.
What’s odd that it only fails to correct the white balance, the rest of the colours seem to be accurate with an average dE below 1 after calibration. I’ve also calibrated two monitors on my PC, which looked quite different out of the box but were matched nicely after calibration.
So if the spyder 5 isn’t accurate, why does it seem to calibrate monitors white balance wrong with high repeatability? Isn’t that an indication that there is a misconfiguration?In the attachement there is a report, maybe that helps to diagnose the problem better.
Assumed WP distance => distance to daylight curve
Profile WP distance => distance to recorded WP at profiling stageSince you aim to an alternative visual white point high errors to daylight curve are expected… you told app to behave that way because you did not trust Spyder 5 accuracy (and that’s too it’s very innacurate)
Profile WP distance is small ~2dE, not ideal but all seems to work as intended.BTW do not use very fast calibration speed, your display is not so good and you got high range a*b*, color tint in greys. Aim for medium speed.
Also do not use XYZ LUT profiles unless you know why to use them, aim yo single curve matrix profile + black point compensation.Thank you, so I read the verification result wrong. The manually dialed in white balance is in matched, but “verify” tool just uses the D65 white point.
But there is still something that makes me suspicious that the white balance measurements are off. I’ve measured my laptop (Thinkpad Yoga 370), and uncalibrated I got a white point of 0.2982 x 0.2884. After calibration the display looks a lot greener.
Notebookcheck has a review of that laptop where they measured 0,301 x 0.3322, and even mentioned “Out-of-the-box, the display did have a slight green tint, a phenomenon encountered quite regularly with IPS displays that have not yet been calibrated.” Their colour profile adds a slight pink/red tint to the white. So whats the matter with that?The LG Monitors stand in a dark corner with very few sunlight. Do I have to adjust for that in white balance?
Whats the problem with XYZ LUT profiles?
Sorry for my noob questions, I am here to learn more.
Thank you, so I read the verification result wrong. The manually dialed in white balance is in matched, but “verify” tool just uses the D65 white point.
Uses closest daylight white, D65 is one of them: dE to daylight locus, the closest one, D55 D50, D65 D75…
But there is still something that makes me suspicious that the white balance measurements are off. I’ve measured my laptop (Thinkpad Yoga 370), and uncalibrated I got a white point of 0.2982 x 0.2884. After calibration the display looks a lot greener.
Notebookcheck has a review of that laptop where they measured 0,301 x 0.3322, and even mentioned “Out-of-the-box, the display did have a slight green tint, a phenomenon encountered quite regularly with IPS displays that have not yet been calibrated.” Their colour profile adds a slight pink/red tint to the white. So whats the matter with that?The matter is the innacuracy of Spyders and the very reason to do not buy them.
Edit: Thats why you use a visually selected whitepoint, because you cannot trust WP readings on that spyder.
The LG Monitors stand in a dark corner with very few sunlight. Do I have to adjust for that in white balance?
No, just brightness to whatever needs you have
Whats the problem with XYZ LUT profiles?
Sorry for my noob questions, I am here to learn more.
Less compatibility and stores actual TRC in 3 indepenent channels… and that may cause rounding errors color managed on displays that after VCGT calibration is applied “visually” look grey neutral on a grey gradient NON color managed.
As a general rule for *well behaved displays* after VCGT calibration:
-for color managed apps using OS ICC profile: single curve matrix + BPC
-for building accurate LUT3D: use XYZLUT,Thats why you use a visually selected whitepoint, because you cannot trust WP readings on that spyder.
Instead of using the visual white point editor and dial it in by eye, would you recommend to measure the white point of my laptop with the notebookcheck profile and transfer that to my other monitors?
Maybe in the future I’ll buy an xrite i1Display pro, but since I don’t do colour sensitive work it is more or less just about getting a consistent image across my monitors.Thats why you use a visually selected whitepoint, because you cannot trust WP readings on that spyder.
Instead of using the visual white point editor and dial it in by eye, would you recommend to measure the white point of my laptop with the notebookcheck profile and transfer that to my other monitors?
No, because spectral power distribution of the LG and the laptop may be different. Even if the two are sRGB only (supposed to be White LED, blue led + yellow phosphor), nowadays its common to see manufacturers using WLED PFS + gamut limitation in formware beacuse that backlight os cheap and reliable.
