2. Hopelessly bad profiles
Average deltaE in the 4 range
Can't speak to your first item but... I wouldn't call the profiles hopelessly bad, that said I wouldn't call them good either. When I tried it out, using the color checker patches I managed an average dE of under 1.0 and a max of under 1.4 (this is on an HP LP3065). If you haven't already done so, go into advanced mode and make sure you select the appropriate monitor type (in your case, CCFL wide gamut). Also, after the first profile is built, go back and measure the luminance again. If you're white point target is D65 and your monitor's native color temp is 5700ºK, and you set the luminance whilst at the monitor's native temp, the resulting profile will have a lower luminance than your target and this can significantly increase dE. After building the first profile, go back and measure the luminance again and manually adjust it if you have to, and then reprofile and revalidate. Your dE should go down.
Another note about dE. You can't compare dE values from Color Eyes Display Pro, Spectraview, Spyder3 Elite, BasICColor etc. to i1Profiler. dE validation results are only relevant to the actual test in the actual program you're running. If you were to change the test type to one of the different types of tests that i1Profiler has, you'll probably see a change in the results.
4. Totally buggy implementation of ADC
The software was unable to control a Dell U2711 properly. See the following thread for a description. This problem was also experienced by another forum user. I logged a support query with x-rite and they were able to reproduce problems 1 and 3. For problem 4, they told me to put the monitor in Standard mode, rather than Custom Color mode, which is not an acceptable solution. They attempted to be helpful, but it was clear that the software is very buggy. It's worse than beta software. I honestly don't know how it got past QA into general availability, because it's obvious that it has not been properly tested on 64 bit Windows 7.
Why they called it ADC is beyond me. It's not ADC, it's DDC/ci. ADC is a TLA that X-Rite's marketing department made up and yes, it has flaws. That said, if the 2711 firmware is anything like the 2410's firmware, the mere act of changing any of the RGB offsets or gains will immediately introduce a dE of 10+. I've seen the exact same behavior in the Asus PA246Q and I have a sneaking suspicion that the guts might be OEMed from the same company (both 10 bit panels with 12 bit monitor LUTs and identical bugs when changing the RGB gains or offsets).
So, it's possible that your high dE values are a result of the monitor and not X-Rite. I get average dE values of under 1.0 with peaks under 1.4 on an HP LP3065. Haven't tried it on the NEC and I don't plan to either (kind of silly).
The only reason I bought an i1Display Pro is because it's now supported by Argyll which is what I'm using it with. NEC, Eizo, and BasICColor have indicated plans to support it as well.
Eventually, in frustration, and despite the widely reported claims that third party software cannot use the retail version of the i1Display Pro, I installed ArgyllCMS and dispcalgui. And thank goodness for quality open source software.
i1D3 support in Argyll is the reason I bought it in the first place.
Dispcalgui functioned flawlessly, giving me a very high quality profile with a deltaE of 0.2. Without this software I would have been left with an expensive paperweight.
Again, you cannot use a software's validation results as a yardstick for the performance of the software. Color Eyes Display Pro uses a default validation test that runs something like 13 grey step wedges and then the RGB primaries. It always results in a very low dE. Change the validation test from its default to something that has more patches, and lots of patches with color, and you'll find that the dE goes up dramatically.
The proper way to evaluate the quality of a profile is to look at some images, see how they're reproducing.. look at a print and compare it to the image on screen, etc. A white to black greyscale gradient should ideally be smooth without any 'bands" and the tonality should be neutral all the way across without any color casts.
The software does produce useable results but Argyll is much better. My preferred solution though for non NEC monitors is still Color Eyes Display Pro with an Eye One Pro spectrophotometer. You sacrifice a bit of black point (has to be at least 0.35 cd/m^2) but the overall neutrality of the profiles it produces are the best I have seen so far (and I have a good collection... Spyder 3 Elite, BasICColor Display, i1Profiler, Eye One Match, and Monaco Profiler). That said, I'm still exploring Argyll.
My biggest complaint with i1Profiler with the i1Display is that it just won't reproduce shadows very well below 10, 10, 10. I have no idea what that's about but it's pretty bad. I find it extremely perplexing that the same piece of software which is capable of producing truly wonderful RGB printer profiles can't do the same in the monitor department. It's still a version 1 piece of software. Hopefully it will improve.
Cheers, Joe
