Originally Posted by millerwill
Well, Jeff Meier (UMR) was at my house last night doing his magicand it was a most enjoyable and informative evening (from 5 pm to about 10). I had met Jeff last year when he was at my friend Les Alderson's (Htwaits on the Forum) house calibrating his 60 Kuro plasma. At that time I had a RS1 but was anticipating the RS20, and was so impressed by Jeff that I decided I wanted him to do a calibration after I got one. And it was a pleasure also to have Les here last night to lend an extra pair of eyes and witness Jeff's results. (FWIW, I might note that though Jeff can come across as rather abrupt' in his posts here on the Forum, in person he is extremely agreeable and gentle in dealing with persons much less knowledgeable that he is in the HT game.)
He first addressed grayscale. I have been doing this (about half a dozen times by now) with my I1/LT and HCFR, and gotten pretty comfortable with this methodology, though was under no illusion about it being highly accurate. Jeff very kindly noted that I had done pretty good' within the limits of these fairly elementary tools (and my inexperience!); in fact it measured (with his very precise tools) fairly reasonable down to ~20 IRE, but fell apart below that, color temp ballooning from ~6500K to ~9000K at 10 IRE, and gamma, which followed ~ 2.2 down to that point, also blew up (to over 2.4). Fairly quickly he was able to dial in grayscale to give R, G, B absolutely spot on down to 10 IRE; and one could really see the difference in grayscale patterns, i.e., that the color tinges in the low IRE grayscale were eliminated. He also gave me tips on how I can profitably use my I1/LT to measure the results of his grayscale settings, to generate approximate correction factors for the I1/LT, to use on my own to dress up' his values as the lamp changes with time, etc.
He then moved on to CMS; and here it was extremely fortunately that Jeff has a RS20 himself and is thus extremely familiar with it (though he has so much experience that he is familiar with many displays). Jeff's methodology, as he has described here on the Forum, is to use his state-of-the-art instrumentation together with his very experienced eye to dial in color (as well as grayscale above). Jeff has his own software for grayscale and CMS, and it was amazing to see how rapidly he can generate measurements of these. (He trains his I1/Pro to his super accurate (and super expensive) spectrophotometer and then uses the former to run through the measurements. And though I've read about how slow the I1/Pro is, with his software it is incredibly fast.) He noted, as have others, that the THX Picture Mode is very close to Rec 709 once the grayscale has been set accurately (and Color and Tint also adjusted somewhat). In User 1 he then dialed in his own CMS, which is a slightly expanded gamut. His philosophy is to dial the CMS to intermediate color points, not the corners' of the CIE triangle; he has his own custom points that he has devised which sample typical scenes. This pushes the corners of the CIE triangle out slightly, but the result is an extremely rich, yet completely natural and pleasing color palate, esp for movies, as evidenced on a number of dvd and BD samplings.
So I am extremely pleased with the experience and the result. The grayscale is unbelievably precise, and I think Jeff's benchmark (and tips) will make it easier for me to keep it close. The CMS he provided is a very rich and striking palate, with the (properly tweaked) THX mode providing Rec709 in case it is preferred for some material.
As an aside, he also measured the light output and CR of my pj (which took just a minute or two with his spectrophotometer), providing much more accurate values than I have been able to get with my CA813 light meter: on my 126 diag HP screen, the pj at min throw, Normal lamp mode, 780 hrs on the lamp, and in a pretty good room (no batcave, but no external light and black material on ceiling and side walls out ~8 ft from the screen wall), his values (after the full calibration) are 17.7 ftL, o/f CR = 29,000:1 with Iris = 0, and 15 ftL, o/f CR = 40,000:1 with Iris = -7, with ANSI CR ~ 250:1. I have the RS20 mounted on a stand, about dead center of the screen both horizontally and vertically (i.e., perfect for the HP screen), but he noted (as have others) that ANSI CR is considerably improved if one can mount the pj higher, say, and use lens shift. He said that in this config (and in his much more ideal room) he gets ANSI CR of ~ 450:1 with his pj.
PS. We also looked carefully for any banding, in some of Jeff's test scenes that are extremely revealing in this regard, and there was absolutely NONE that one could discern.