What you wrote was an assumption without data ("if it were in this app, you'd be calibrating at the top left corner of your screen, which then wouldn't be ideal.")
Now for the data! Here's the dE between 9 measurement values taken dead center and in the upper left-hand corner of my PA. It's invisible!
You were unable to answer my question, but I could do so with measurement data.
I'm sorry the colorimetric data may ruin your day.
Andrew, I really don't get it man, why do you want to "ruin my day with colorimetric data?" I even acknowledged earlier that the issue was not one of calibration, but a simple annoyance of the OS. You made an assumption that what I meant by "measuring in the upper left hand corner, which wouldn't be ideal" meant that I thought the upper left corner of my display would be less accurate by design. What I actually meant was that I wouldn't want to be forced into measuring any specific part of the screen because of some stupid OS display issue. I'd rather choose where I can measure. Also, many monitors that are not like your PA
do deviate in a significant amount from the middle to the edges. Even by Eizo's own admissions, without their DUE compensation, there can be not-insignificant deviations. Your monitor also has uniformity compensation. I don't know what OP is using. Also here's another good example-- I have an (older) professional sony mastering monitor that has a backlight issue at one of the edges. It's so far over that it's easily ignorable. However, if I were forced to calibrate THAT area, it would screw up the rest of my monitor to the point of being unusable. By your logic I should throw it out and spend $30,000 on a new one?
I completely agree with your assessment of the situation that is relavent to this posting-- that this slight movement of the measurement area is insignificant. But I don't see why you would venture to stop the discussion of the OS display issue, which IS AN ACTUAL ISSUE, even if it's a minor one with no bearing on the calibration of that display. I realize you literally wrote the book(s) on this stuff, and I've even led other members your way when they've asked for outside printer profiling help. I just don't understand the attack here. By saying "it would be less than ideal" to measure the top left corner of MY monitor, I'm not exactly spreading harmful disinformation. And proving that YOUR monitor is very uniform edge-to-edge is also not a particularily useful guide for anybody not possessing a monitor of such quality. All monitors with built in calibration devices measure at the edges of their displays. But these are also expensive monitors, and they're not the only thing out there that can get a job done. Bigger displays can have worse uniformity issues depending on technology-- is it not worth calibrating those displays? I wouldn't say so, but I'd like to choose the point at which I can measure. Sometimes it's even a matter of actual physical space-- If I'm doing a simultaneous measure for a colorimeter correction with a spectro, there might literally not be enough room to have both devices at the corner of the screen at the same time. Sure there are ways around that too. As for SpectraView, I don't believe you have the option to choose where you can place the measurement window, so in this instance it truely absolutely does not matter for the point of calibration of this display, but there is a "not never" situation in which you might want to change the position of the measurement window, and for most of my monitors, i'd like to use the center (not DEAD, idc if it's off by a few millimeters or even inches in any direction) because that's the place my eyes rest the majority of the time.
If you really wanted to have this discussion it would make sense to start a different topic about it, I still don't understand the vitriol for continuing to help solve a window issue that doesn't relate to color management even though it was posted in the color management section of the forum.
Here are some attachments about across monitor uniformity, one is from a test of an eizo flexscan (the one in color) which does not have any kind of uniformity compensation, the other is from Eizo's website discussing their compensation, with no key for what the different shading values mean in any numerical sense. Not my data, but I've looked at a lot of uniformity graphs in my day, and for monitors that don't have compensation they often have enough variance across the panel that differences are noticable.
Just anecdotally, I used to use the NEC PAxxxxQ monitor that you have, Andrew, daily, and when you displayed a black and white image on it full screen, there were noticeable shifts from left to right in green/red. Maybe you won the panel lottery, maybe the shifts I noticed wouldn't be measurable. The red on that display is PFS phosphor, a very spiky narrow wavelength, and wreaked havoc with the perceived neutrality in my eyes. Since spectraview doesn't allow colorimeter corrections, I had to switch to using displaycal and calibrating throught he graphics card instead of the hardware calibration in order for it not to drive me crazy with a pinkish cast when calibrated to D65.
