Two monitors: calibrated to same settings, different colors

[Freek07]

This is weird to me: I have two monitors (HP LP3065 and Dell U2410), both ips (OK, dell is "eIPS"). Nevertheless, I calibrated both (Spyder3 + DisplayCAL / Argyll).

Measurement report shows both are "OK" (within reasonable dE ranges) but when I'm editing photos, DELL has noticeably less vibrance and is a bit colder despite HP having lower luminance. I know they should be both set to same luminance but it's tricky on old HP as it has too much drift.

What I'm wondering is: how can be colors so different when calibrator says both are withing defined ranges (and settings are same)? Any ideas on making them look more simmilar?

Here are the reports for both of them:

• http://www.apollonus.net/m/DELL.html
• http://www.apollonus.net/m/HP.html

 

[Simon Garrett]

No complete explanation, but a comment: I find that two monitors with different brightnesses also tend to look different in colour and contrast.

 

[bmoag]

Unless the monitors are set to the same luminance you can not compare them by eye. The eye sees what it expects or wants to see but calibration devices see what the monitor actually displays.

Calibration tells a color managed program what RGB value the monitor actually displays versus what it ideally should display. There is no way to do that by eyeball.

The eye/brain is more sensitive to changes in brightness than it can discriminate subtle changes in hue. Hence a brighter or darker version of the same RGB value is usually interpreted as a different hue.

I have two monitors with somewhat different gamuts calibrated to the same dimmish levels for color managed printing. When I preview/soft preview on either panel I hardly see any differences but there were seemingly more differences at stock brightness levels. I think that is an example of how color discrimination decreases with decreasing brightness.

Even if set to the same brightness there can be color mismatches between monitors depending on what part of the sRGB/RGB gamut the particular monitor is best at displaying. If two monitors are rated as showing 90% of the sRGB gamut it may not be the same 90% or there may emphases in different parts of the gamut.

If you are pursuing color managed printing one experiment would be to process the same raw image on each monitor independently and see what prints look like. An out of camera/not manipulated sRGB image should print the same even if it seems to look different on the monitors but that might also help you decide which monitor to preferentially use for image processing.

 

[Pictus]

Check

 

[Simon Garrett]

Unless the monitors are set to the same luminance you can not compare them by eye. The eye sees what it expects or wants to see but calibration devices see what the monitor actually displays.

Calibration tells a color managed program what RGB value the monitor actually displays versus what it ideally should display. There is no way to do that by eyeball.

The eye/brain is more sensitive to changes in brightness than it can discriminate subtle changes in hue. Hence a brighter or darker version of the same RGB value is usually interpreted as a different hue.

I have two monitors with somewhat different gamuts calibrated to the same dimmish levels for color managed printing. When I preview/soft preview on either panel I hardly see any differences but there were seemingly more differences at stock brightness levels. I think that is an example of how color discrimination decreases with decreasing brightness.

Even if set to the same brightness there can be color mismatches between monitors depending on what part of the sRGB/RGB gamut the particular monitor is best at displaying. If two monitors are rated as showing 90% of the sRGB gamut it may not be the same 90% or there may emphases in different parts of the gamut.

If you are pursuing color managed printing one experiment would be to process the same raw image on each monitor independently and see what prints look like. An out of camera/not manipulated sRGB image should print the same even if it seems to look different on the monitors but that might also help you decide which monitor to preferentially use for image processing.

+1 to all that (also see the video Pictus linked - thanks Pictus).

It may be impossible to match two dissimilar monitors exactly, and even if they are identical they need to be at the same brightness for the colour and contrast to look the same.

However, one can very often get pretty close provided the brightness is the same. I've had a variety of wide and standard gamut monitors over the years. I get new monitors one at a time, and I've rarely had two the same. However, at the same brightness (I normally use 100 or 120 cd/m2) then they normally look virtually identical for colours within both gamuts.

For example, if one is wide gamut (around Adobe RGB, say) and the other standard gamut (around sRGB) then colours within sRGB look the same on the two monitors (as sRGB gamut is a subset of the Adobe RGB gamut).

 

[sygnus21]

What I'm wondering is: how can be colors so different when calibrator says both are withing defined ranges (and settings are same)? Any ideas on making them look more simmilar?

This is not unusual as even the same model from the same vendor can display different colors due to manufacturing and other process properties. And here, you're talking about two different monitors and vendors altogether.

At the end of the day there are simply too many variables to get every monitor to calibrate to the exact same calibration spec. That is, as you've seen, calibrating a monitor using the same settings can result in different results due to a myriad of variables not the least which starts at the manufacturing processes.

Good luck.

 

[Simon Garrett]

What I'm wondering is: how can be colors so different when calibrator says both are withing defined ranges (and settings are same)? Any ideas on making them look more simmilar?

This is not unusual as even the same model from the same vendor can display different colors due to manufacturing and other process properties. And here, you're talking about two different monitors and vendors altogether.

Quite true if we're talking about monitors in their uncalibrated, unprofiled state.

But the purpose of profiling is to measure the actual performance of the monitor so that colour management software can take account of the difference between individual monitors.

At the end of the day there are simply too many variables to get every monitor to calibrate to the exact same calibration spec. That is, as you've seen, calibrating a monitor using the same settings can result in different results due to a myriad of variables not the least which starts at the manufacturing processes.

Good luck.

Exact, perfect matching may be impossible. As mentioned in the video Pictus links, differents in spectral power distribution may mean that different displays may not match for all xyY values. In practice, the differences between two calibrated and profiled monitors for in-gamut colours may be imperceptibly small.

 

[Phil Shaw]

Are you running both monitors at the same time? If so, and this is the difference that you are seeing, Windows will only load the profile for one of the monitors.

 

[Simon Garrett]

Are you running both monitors at the same time? If so, and this is the difference that you are seeing, Windows will only load the profile for one of the monitors.

Do you think so? Windows has been capable of loading separate profiles for each monitor at least since W7, and every video card I've used for the last 10 years has managed separate profiles for each port.

 

[sygnus21]

Exact, perfect matching may be impossible.

"Exact, perfect matching" IS impossible, and you know it.

Moving on.