Hi again everyone,
Okay, so what about Sunday, right? Sunday was awesome. I was a little sleep-deprived (remember the house was both crowded and busy), but I got to the store nice and early, got my computer set up, got one of the comfy theater chairs to sit in, and generally had a nice time of it. The room was not quite as crowded as on Saturday, which also helped, and both Value Electronics and the various presenters had the kinks out of their presentations, so everything went a lot more smoothly.
Before I forget: thanks more than I can say to Robert, Wendy, Jeff, Lianne, and Katie Zohn all of Value Electronics for putting this whole thing together, for putting me up, for putting up with me, and for making the weekend one of the most enjoyable I've had in a long time. Thanks in abundance to each and every old friend I got to see there and each new one made; I've been out of the loop for way too long, and I so missed this beautiful hobby of ours. Thanks in equal abundance to the manufacturers for working hard to step up to the plate and deliver the best TVs they judged possible in such an uncertain and roiling market. Very special thanks to the Sony and Samsung reps who where willing to sit down with me one-on-one, answer my questions so clearly and honestly, and even listen to my obviously half-informed attempts at advice for future improvement. I remain convinced that Sony really can make something happen here with their local dimming technology if they just make sure they use fast LEDs and make sure those LEDs never turn all the way off. That along with some software improvements in smoothing the transitions between various zones could punch the ball right through the goal posts. Oh--and Kevin Miller is right on: please, Sony: give calibrators full and fully functioning CMS controls and blue-only, red-only, and green-only screen capabilities. For the amount of money on offer, all these things are owed.
Sunday
People started showing up a little after two, and by 3:00 the place was hopping. Value Electronics got the video and audio feeds all straightened out by 3:30, and we were collectively off to the races again. Joe Kane gave his presentation again, but he included further information that was not in the Saturday talk, so you can bet I paid close attention to everything he said. His example from photography to show why a high bitrate matters was, IMO, inspired. Dr. Weber's presentation was fantastic, and it made very clear why and how each of the three display technologies encounters hurdles and why some such hurdles are surmountable while others are not. My biggest take-away from his talk, though he never mentioned this, is the extent to which LG's OLED TV this year is able to do what it does by including some already-known strategies from LED-backlit LCD technology, a technology in which LG has been successful for a long time now. Samsung, by contrast, really tried to do something new with their OLED, and I admire what they've done very much, but I'm dead certain the price difference will kill them. Given the bad economy both globally and in the U.S., manufacturers simply must control costs, and the white OLEDs pushed through a screen of color filters do that much better than three distinct colors of OLEDs pumping the light out emissively. It's that simple. Mainstream America no longer has that 90s-style delusion of being wealthy when we're not, so you can't sell us stuff as if we were. I think LG read the market a lot better than everyone else this year. They're even slated to release a much larger 4k version of this OLED TV in a few months, and that's the one that Ken Ross was gunning for yesterday. Not to steal my own thunder for the paragraphs below, but I agree with him. Were I buying this year and I had to buy something specifically from the wall other than the F8500 plasma, I'd grab that LG and not look back for a moment.
The manufacturers' presentations were largely the same, though a bit more informative, actually. The calibrators, I'm happy to report, streamlined their presentations considerably and really did a stand-up job yesterday. The actual audience test portion was a lot of fun and very informative, though it was hard to move around much with the place so crowded. I almost think for next year Robert & co. should have the audience queue up in some way and then walk clockwise (or whatever) around the room, thereby enabling every participant to get some proper quality time with each panel. After that it could become a free-for-all where folks just go where they wish and focus on whichever TVs most interest them. This way the room traffic could be rendered much more efficient.
Both before and during the test both Ken Ross and I were looking closely at the two OLEDs and I was looking hard at the XBR950, too (I don't remember if Ken had his eye on that one or not). The first and biggest thing we noticed is that the LG OLED did a much better job of handling off-axis viewing. When you go even just a little bit off axis on the Samsung OLED, the color balance and the gamma both get seriously borked. I mean, the gray ramps they had up there looked red and far from even in their transition from dark to bright. But as soon as we got both vertically and horizontally on axis, all that went away and Samsung's ramps actually were better than LG's. Really, though--you had to be dead center, which I should think is an obvious problem for anyone interested in buying an OLED this year. IIRC, one of the big selling points of OLED was supposed to be to absence of viewing angle problems, wasn't it? Well, that's not what we have with either of the ones on offer. LG's TV has definite viewing angle limitations, too, but they're nowhere near as extreme as Samsung's. For that reason alone I determined that, were I buying this year, I'd much prefer the LG OLED over Samsung's (regardless of price). The LG did have a very annoying problem to it, though: it had some of the most aggressive ABL I've ever seen. Seriously, it made recent plasma ABL that I've seen look half-hearted by comparison. It was sufficiently visually annoying that I resolved to ask them about it if Ken didn't (but he did). Both OLEDs handled black levels and perceived color saturation admirably, of course; here even the F8500 plasma could not hope to keep up. But on color tracking neither of them did so well, and the LG in particular went way off target in the color saturation charts. In real world content these flaws did not jump out at us immediately, but they were evident and sometimes very annoying. The industry in general, with the notable and admirable exception of Sony, has not paid anything approaching the attention to color accuracy that they have to black levels/contrast ratios, and I say that needs to change, pronto. If they want this kind of money from people, they owe their customers color reproduction technology that is not still derived from what they were using back in the 20th century. As best I can tell, there is just no defensible reason for color tracking to be as embarrassingly weak as it is. Like I said, the one exception to this is Sony, who has done strong work here for a long time, and that was evident yesterday. Sony allowed calibrators NO CMS controls of any kind on the XBR950 and only two-point grayscale controls. Armed only with that, the calibrators nonetheless got that thing to deliver the best, most accurate color I saw on any TV on the wall, and even the gamma was crazy near perfect. I mean, really--what excuse do Samsung and LG have for falling down so badly on this when they provide full CMS controls, blue-only screen capabilities, and TWENTY-POINT grayscale controls??? And even with all of that, the Sony still had better colors??? Look, that's just wrong. I say both LG and Samsung need to hit the drawing board on this.
But lest anyone think I have a hard-on for the XBR, I still say both OLEDs ended up delivering a much better picture, and black levels were NOT the reason why. The problem was the flawed implementation of local dimming on the Sony. I don't know if it was due to a decision by the calibrators or due to the way the panel works, but the 950 had its local dimming function set so that on a black screen the LEDs would shut all the way off, which I have said for years is just a disastrously bad idea. You don't just get uncorrectable blooming from that; you also get what I call "ghost images" from the relatively slow LEDs failing to keep up with motion-driven transitions in the LCDs. You watch the Star Destroyer first approach you in the Empire Strikes Back; There's a moon behind it. As that moon moves down the screen, the LEDs will be too slow in turning off, so you'll see a "ghost" of that moon chasing after the moon the LCDs show you. Anyone with a locally dimmed TV that turns LEDs completely off can get a copy of the DVD (or BluRay--I used a DVD) and test this. I tested it for almost a month before making my determination that the problem is real, that the slow LEDs are the culprits, and that stopping them from turning completely off is the only fix that really works. This strategy has the additional very considerable benefits of getting rid of all floating black level issues (assuming the TV has a reasonable number of discrete zones), getting rid of information loss ("black crush") in dark areas of the picture, and allowing manufacturers to get away with offering local dimming without using 300+ zones in order to make the picture acceptable. MY 46" UNB8500 has only 140 zones, and I never, ever have any of these problems, and stopping the LEDs from turning all the way off is the reason why. Sharp, I was delighted to see, apparently listened to all my screaming into the void of cyberspace and designed their Elites so the LEDs get very dark, but never turn fully off. And theirs are the best locally dimmed TVs out there. Their color problems are well known and very regrettable, but their implementation of local dimming technology has no peers. Sony, Samsung, LG, I beg you: learn from Sharp if you want to get this right. And while you're at it, study what they did with viewing angles, because they really nailed that one, too.
One last point on this before I forget it: Robert has publicly credited me with inventing the zone-count test. Strictly speaking, I don't think that's true. Wayyyyyyy back when, I bought one of the very first Samsung 81F locally dimmed TVs offered in the U.S. It had just 64 zones and a host of associated problems. (I have no idea what Vizio thinks it is doing trying to sell a "locally dimmed" TV with EIGHT zones and even trying to get that thing into the shootout!) I was, at the time, just beginning to learn to calibrate, and AVS was where I did it (back then they had much better troll control). It was between a bunch of us that we came up with the strategy of maxing out brightness, going into Dynamic mode, using a perfectly black desktop image from our computers, and then slowly dragging the white mouse across the screen to count the zones as they lit up and darkened in response to the mouse moving across them. Accordingly, I cannot take all the credit for inventing that test; I was part of its invention, but so were several other AVS members, and IIRC it was one of them, not I, who first made the suggestion. On the other hand, I can and do take full credit for figuring out that stopping the LEDs from shutting all the way off was the only way to effectively minimize the otherwise-troubling artifacts of local dimming. No one else came up with that one; that's my original contribution, and I stand by it. It took me a very long time--weeks and weeks of amateur calibration work--to figure it out, but it totally, totally works. Do you get perfect 0 black levels? No. Do I care? No. Because you still do get black levels that beat even the 9G and 9.5G Kuros and the VTs. Show me any other TV short of OLED that can make such a boast. And all the while, you quite get rid of these ridicuously annoying floating black levels (both locally and globally across the screen), black crush, LED ghosts, and all the rest. It's worth it, and the price paid in black levels is absurdly small. Using my C5 meter profiled to my i1Pro in a perfectly dark room, I got consistently measured MLLs of 0.00x fL and below. I actually often can get it so dark with the help of my iScan Duo that even the profiled Chroma 5 can't read any light, though I can just see it in a perfectly dark room. Anyone who finds those black levels insufficient in an LCD TV has gone monomaniacal and needs to just sell the thing and get an OLED, because even plasma will never make you happy, man.
So Samsung, LG, and especially Sony, I hope you guys are reading this. If so, please perform the tests yourselves and see what I'm talking about; get a Sharp Elite and see how it handles dark and mixed scenes. If need be, find yourselves an old Samsung UNB8500 like mine and test it. The Elites wisely won't let you shut the LEDs completely off, but the UNB8500s will, so they're useful for comparing the different results. If ever you have any, any questions at all about this stuff, please, please, please PM me either at highdefjunkies (my internet home) or at hometheatershack (my internet "second home," if you will). I would love to answer any and all questions the manufacturers' representatives and engineers may have for me and just to bounce ideas around for how to optimize it. I firmly believe that if local dimming is going to work, this is the way to make it work, and also this is the only way to make it economical, because you can get away with having just 200-ish zones on a 1080p screen (for 4k, the number does climb, I'm sure).
At last I come to my final paragraph (may all the angels of Heaven be praised with great praise). So which TV did I vote for as the winner last night? The plasma. Sorry LG and Samsung, but you guys have got to get both color reproduction and viewing angles firmly under control (and the ABL for LG). Then and only then will your OLEDs get the nod from me. On motion resolution, I say both companies also have work to do, but not nearly as much. COLOR ACCURACY and VIEWING ANGLES; fix them or forever come in second and third at best in my book. So between all the TVs from this year who gets my vote? LG hands down. The Samsung OLED was awesome on so many points, but its viewing angle problem totally killed it for me. Add the insane price difference to the mix and I say I'd have to be on heavy unprescribed meds to look at anything other than the LG OLED this year. After the two OLEDs the only serious new-this-year competitor to my mind was the XBR950, warts and all. Really, though, that F8500 plasma was still tops. Its viewing angles were superior, it wasn't curved (just dumb), and its colors were right there. All the other LCDs were, well, forgettable.
Yours,
David
Hah! Got 'er in just one post. So there you have it, folks.
