Eizo Monitor Calibration

I have an Eizo CS2731 that I bought last autumn and calibrated with Colornavigator, which in the process created an icc profile and placed in Windows 10. It's worked fine since then, but following recommendations I finally got around to recalibrating yesterday and posed some questions:

Firstly, when I calibrated the AdobeRGB profile I noticed in Windows File Explorer that the modified date of the other CS2731 profiles (such as sRGB) all changed their 'modified' date to the current date, so does this mean they don't have to be calibrated separately'?

Secondly, I looked at the profiles created last year and yesterday in ICC Profile Inspector and the numbers appear identical; I expecting to see some small changes, but I guess I'm missing something here?

one might assume that all changes are written into hw luts in monitors... nothing is expected to be changed in icc/icm files for OS

I think it's possible that when you update any profile, ColorNavigator updates all the Windows profiles for calbrations contained in the monitor, even if you haven't changed them. (I've just checked: every time I select a profile in ColorNavigator or select it on the On Screen buttons, ColorNavigator seems to rewrite the profile file in Windows and the date gets updated.)

With Eizo and other monitors using calibration by means of LUT internal to the monitor, there shouldn't be any need for LUT data in the graphics card. I see that Eizo profiles do contain vcgt information ("Video Card Gamma Table", which is where video card LUTs are stored), and there's 1554 bytes in them, but that's presumably a null (linear) transformation loaded into the graphics card to ensure the card doesn't do anything other than a linear mapping.

As for the numbers: if you create a profile with the same settings as one you had before, the LUT mapping loaded into the internal LUTs in the monitor might change to reflect slight changes (ageing) of the monitor. After creating the calibration (LUT data), the software then creates the profile, which is a measurement of the monitor after calibration. Unless the calibration is absolutely accurate (or any error exactly the same as last time) one would expect a slight change in the profile from the last time, so whatever you're missing, I'm missing too!

In fact: if you run the validation after calibration/profiling, you end up with figures which are (with my monitor) not identical to last time. So I really don't know why the profile itself wouldn't be slightly different.

You say you compared the numbers with profiles created last year: did you take a copy? The reason I ask: each time I recalibrate, ColorNavigator overwrites the profile files, so I don't have earlier versions of the profiles.

I've just recalibrated my Eizo monitor, and beforehand saved the profiles from the previous calibration.

For calibration to standard colour spaces (sRGB and Adobe RGB) the new profiles are the same as the old ones, except for header information. If they are identical, I would have expected the reported R, G and B values to be the ones for the standard (i.e. sRGB or Adobe RGB) but they're slightly different. I'm not quite sure what's going on here: values not identical to the standard but the same every time, which sounds suspicious if they were measured values.

Calibrating to native colour space, the new profile is not the same as the old one. The R, G and B values now appear to be the measured ones for the monitor's native space, which in the case of my CS2420 are a somewhat wider than Adobe RGB. This is what I'd expect: slightly different values after each calibration.

Sorry to be while getting back. To answer your question - yes, I'd saved copies of the earlier profiles. Like you, I'm not quite sure what's going on, but your native result is interesting. Whatever the answer, I have to say I can't fault the result, the match of screen to print is excellent.

I may be able to add a new twist soon, as I've treated myself to some retail therapy and ordered a CG2700S, with built in calibration, so I'll report back on whether the behaviour is similar.