I have done more than one test comparing the i1Pro rev D and 2 different models of i1Display2 (one “normal” and the other one branded by NEC). I can tell that units with a proper correction matrix (correct calibration) of the i1Display2 can work better than the i1Pro, because the i1Pro, in my case, is always lacking in dark-area measurements. I also have problems with temperature. If I let worm-up the i1Pro like somebody suggests I get even worst results in the dark areas. And if I need to read a number of patches greater than 50 I will start to get inconsistent results between the first and the last reads (letting the instrument worm up and using it without a worm up, it seems that the temperature will increase more during the readings. In my opinion i1Prois neither very consistent or very precise. As I reported in previous posts I tried i1Pro rev D with more than one software and I found ArgyllCMS to be the most complete, because it allows you to use features like adaptive measurements, drift compensation, high-resolution readings that can be useful to get the best from the instrument, but I was never total satisfied with the results using it. I tried more than one i1Pro rev D and I now own one of the last units with the i1Profiler (i1Publish Pro bundle), so it is not a problem that I noticed in one particular unit.
Trying and testing different spectros and colorimeters, is that colorimeters are usually better for monitor calibration. I am not saying that spectros are not good, simply that they are not the best. The problem of colorimeter is that they need to “perfectly fit” the monitor to give good results. With an i1Display2 and an i1Pro you can use ArgyllCMS to create a correction matrix for the colorimeter using the spectrophotometer. What I noticed using a Nec SpectraView monitor is that the NEC i1Display2 required no matrix (the matrix I got was very close to an identity matrix); using the i1Display2 the matrix was making some “noticeable” corrections, and so the original “calibration” of the i1Dispaly2 was not good for the wide-gamut monitor that I was using, but after the correction matrix the colorimeter showed a good behavior.
I would not say that the i1Pro REV 2 is substantially better than the i1Display2 for monitor calibration. In my tests I found the i1Display2 to behave similarly to the Discus with a proper correction matrix in the darker areas. Anyway I really think that the latest colorimeter units from X-Rite can be substantially better than the i1Pro REV D in monitor calibration, because as you can read in the link that I have already posted, the new colorimeter show very different features, for example the i1Dsiplay3 seems to have an interesting temperature stabilization and other features that you may want to check if you are interested in a monitor-only device.
as for i1Pro drift:
1. the HDTV calibration I have ask you to recalibrate it every few minutes, however most monitor profile software doesn't allow for re-cal mid-profiling
2. if you use the US NEC PA series SV II it doesn't bother trying to measure the entire scale it just measures a bright white point and relies on the sets 14 bit 3D LUT and great linearity for the rest so dark measurement don't matter at all and drift no so much either
anyway comparing some probes:
NEC PA241W in native gamut mode
According to factory settings it should read (at least when it was new):
"( )" are simplistic raw deltas in xyY from original factory setting claims
With i1 Pro:
R .679,.309 (.001,.003)
G .214,.690 (.014,.004)
B .152,.056 (.000,.002)
With NEC i1D2:
R .683,.306 (.005,.006)
G .214,.691 (.014,.003)
B .152,.057 (.000,.003)
With i1 Display Pro:
R .679,.308 (.001,.004)
G .207,.694 (.007,.000)
B .152,.054 (.000,.000)
Now with the monitor in sRGB emulation mode (the measured values don't match D65 sRGB with the probe used to calibrate it since it hadn't been re-calibrated for many weeks at this point in time so ignore that, that's not the point here):
factory settings claim, based upon how it was internally altered during the calibration, that it should read (and keep in mind the monitor is more than a year old now and has had heavy usage):
B .150 (take it as .152 though since the PA can not actually hit below .152),.057
NEC i1D2 before being warmed up:
WP .319,.335,94.8 (.012,.008)
R .645,.327 (.007,.005)
G .303,.600 (.012,.002)
B .152,.063 (.000,.006)
NEC i1D2 after sitting on the screen for 20 mins and then being continuous read from for about 20-25min (the latter seemed to help it warm up much more than the former):
WP .316,.331,94.45 (.009,.004) [delta to pre-warmup .003,.004]
R .644,.327 (.006,.005) [delta to pre-warmup .001,.000]
G .3025,.600 (.0115,.002) [.0005,.000]
B .152,.062 (.000,.005) [.000,.0001]
So warming up an i1D2 makes little difference when it comes to measuring the primaries but a noticeable difference when it comes to reading the white point.
i1 Pro (no warm-up):
WP .315,.3325,95.4 (.008,.0055)
R .640,.329,20.43 (.002,.004)
G .302,.600,68.8 (.011,.002)
B .152,.062,7.1 (.000,.005)
The i1 Pro readings are nearly the same as the NEC i1 D2 readings other than for the red primary). When both pucks and monitor were all new I thought the i1 Pro and i1D2 readings were even yet a bit closer than the two are to each other now.
i1 Display Pro with no warm-up:
WP .311,.329,92.6 (.004,.002)
R .640,.329,19.8 (.002,.004)
G .296,.603,67.4 (.005,.001)
B .152,.059,6.72 (.000,.002)
i1 Display Pro after 10+ minutes of continuous read warm-up:
WP .3095,.327,92.9 (.0025,.000)
R .639,.329,19.72 (.001,.004)
G .2955,.6015,66.96 (.0045,.0005)
B .152,.058,6.73 (.000,.001)
So the i1 Display Pro read things as being darker than the other two with the i1 Pro have the brightess take on things. The i1 Display Pro readings were closer to factory measurement expectations (although after so much usage who knows what the ideal is).
The i1 Display Pro seems to read best after about 10 min of continuously reading to warm it up after a real lot of non-stop usage for a long time it might start drifting a bit too far, not sure, again warm-up seems to really only affect reading the white point and not the primary locations.
For extra kicks here are DTP94b readings without matrix compensation (or warm up):
Color Eyes branded old copy:
WP .331,.333,92 (.024,.006)
R .642,.330 (.004,.003)
G .318,.589 (.027,.013)
B .155,.063 (.003,.006)
newer copy from and said to have been calibrated by Spectracal:
WP .329,.331,90.5 (.022,.004)
R .642,.331 (.004,.002)
G .315,.592 (.024,.010)
B .153,.062 (.001,.005)
Without compensation they don't actually read the R or B with any greater differences than the other probes either compensated or spectro but they are badly off for reading the green primary and white point on a wide gamut monitor without compensation as known. The DTP94b read the white point darker than any of the other probes, especially the spectracal calibrated one (maybe their filters block too much green light?), although the Color Eyes batched one was closer to the i1 Display Pro brightness readings.
Without a $30,000 reference measuring the screen as it is now, it is a guess what exactly to make of all of it. If you trust the primaries won't have changed too much and that the factory measurements were extremely accurate and that even the white point has only drifter a tiny bit then the i1 Display Pro seems to beat out the expensive i1 Pro and the custom calibrated to the NEC PA241W screen NEC i1D2 measuring all the primaries no farther off at best and often closer and the WP closer. Of course that who knows if the assumption is correct.
The i1 Display Pro should be more accurate and precise than the i1 Pro for dark tones and it doesn't need the constant recalibration during usage of the i1 Pro or DTP94b.
When I calibrated an HTDV with the i1 Display Pro none of the probes agreed. The i1 Pro said it had a bit too much R,G and bit too little B on the gray scale. The i1D2 said it had too much G and too little B. The DTP94b both said it had too much R and too little B (with one saying the differences were to a bit greater degree than the other). Doing a tiny bit of fiddling in photoshop it seemed like adding B and/or subtracting R,G,R+G didn't necessarily seem to help a match to a real life color checker chart (I think I was viewing it under somewhat D65 conditions, I hope) so maybe the i1 D Pro had the best read? A little sketchy and not a really rigorous test by any means at all though.