SMPTE article - UHD Televisions Advance Faster Than Expected
Hot Button Discussion
UHD Televisions Advance
By Michael Goldman
About a year ago, broadcast industry analyst Pete Putman suggested to Newswatch that while the broadcast industry's overall transition to a 4K Ultra High Definition (UHD) ecosystem would be a long and winding process, a foundational element was already well under way--the inexorable march by consumer display manufacturers toward phasing out large-screen, high-definition televisions and replacing them with Ultra HD panels (with a resolution of 3840 x 2160). A year later, Putman says the display industry's transition has picked up considerable speed, to the point where he expects the production of large-screen 1080p HDTV's to largely cease in the next few years.
"The trend line is proceeding," says Putman. "Just take a look at the retail prices for televisions. I [recently] found a 50-in. Ultra HDTV for around $900, and I've seen a few others in the 55-in. range that were just above $1,000. We are slowly reaching price parity between UHD TV's with basic feature sets and 1080p TV's that are the same screen size with maybe a slightly larger feature set. So, right now, for consumers, it's becoming more of an either-or situation. They can spend a thousand or $1,200 dollars and buy a really nice, loaded 1080p set, or a nice UHD TV set. Some people right now are tending to buy the UHD set, seeing it as the wave of the future, and others are still buying the 1080p sets, because they don't yet know where they will get 4K content. But on the manufacturing side, the incremental costs to make panels with Ultra HD resolution, as opposed to panels the same size with 1080p resolution, is relatively small."
"So I see them really ramping up. I don't know exactly where the line of demarcation will be, but let's say for now it will be 55-in. televisions. I think [the manufacturer's conversion] to making UHD-only sets in larger screen sizes will happen pretty fast, probably by the end of 2016 or 2017. So this is a meaningful transition."
Putman clarifies, however, that this transition is not merely about "getting more pixels onto the screen." At the end of the day, it's about moving display technology toward what he calls "next-generation television." By that he means combining higher resolution with the entire suite of picture and feature improvements he preached about in Newswatch a year ago, particularly when it comes to higher dynamic range, higher frame rates, and so on. So far, he suggests, they are off to a good start in pushing Ultra HD televisions in this direction, although the process is ongoing.
"We are now in a new transition to a TV system that is capable of a lot more: More pixels, more colors, more shades of colors, more dynamic range, a wider range of frame rates as all these things fall into place," Putman says. "By offering all this, consumers might say, yes, that is worth an upgrade, especially when prices are dropping so much right now. And from the numbers I am seeing, the sales of Ultra HD televisions are running ahead of what people were anticipating two years ago."
However, as Putman told Newswatch a year ago, "the highway remains under construction," but so do "the performance cars" that will travel the highway. By that, Putman means that the industry has once again entered a hybrid period somewhat similar to the transition period in the late 1990s when it started moving away from analog to high-definition broadcasting, except, of course, that there are alternative digital paths for sending 4K signals to UHD displays today. This means that, besides making today's UHD televisions affordable, manufacturers need to make them more attractive by making them as flexible as possible in terms of features.
"Just like back in 1997, today we can buy an Ultra HDTV, and then scratch our heads wondering where the content will come from," he says. "For the time being, there are lots of tricks [broadcasters] can adopt along the way, and one of them is to improve the 2K HD viewing experience on Ultra HDTVs. After all, Ultra HDTV's have scalers built into them that can take a 1080p signal and scale it up to Ultra HD. That's easy to do because it is a linear scaling process--you double the pixels horizontally and vertically, which was not the case when we went from standard def to high def. But unlike 1997, our communications' platforms have essentially all migrated to digital, so it is a lot easier to implement it today. There are services setting up to stream 4K, of course, and that is where the bulk of the action seems to be, with Netflix and others streaming 4K. Some people claim the quality could be better, but they are streaming it, and if you have a sustained data rate of 15 Mbits/sec coming into your house through an Internet connection, then you can watch certain shows in 4K. Plus, we have a new codec with HEVC [H.265] coming out, and so, you are now starting to see widespread support for that. TV's are shipping with HEVC software decoding built in. We also have the Blu-ray Ultra HD standard [called Ultra HD Blu-Ray] that was recently announced. I expect you will see those products on the market in the fourth quarter of this year."
Such applications are examples, in Putman's words, of "building the plane while flying it," meaning that while the full UHD broadcast ecosystem is being built, as discussed last month in Newswatch, the ability to view 4K content for the time being depends on the features that are added to the UHD monitors already rolling off assembly lines, the formats available 4K programming will accomodate, and the media being used to view 4K content on 4K monitors.
Putman adds that one of the "sticky wickets" in this shift, is that it will, over time, likely lead to other related shifts in broadcast standards or infrastructure details that relate to the consequences of moving 4K signals into televisions for viewing. Among these consequences are the possibility of an eventual "sunset for eight-bit color," he says, leading to the possible need to upgrade today's display interface technology for future generations of Ultra HD televisions.
"I always tell people [with Ultra HD displays], they shouldn't obsess over resolution," Putman says. "You should focus on the total number of pixels, the refresh rate, and therefore, the clock rate. What you have to remember is that, unless you are streaming and decoding in the TV, for any physical media coming from outside the TV, the weak link is the HDMI connection. Right now, the dominant version of HDMI, the current one-HDMI 1.3 or 1.4--has a maximum bit rate of 10.2 Gbits/sec. So, if you want to pass 4K Ultra HD content through that, you can, but you can only do it at 24 frames/sec and 30 frames/sec. You can't do it at 60 frames/sec. That is why they launched HDMI 2.0 [in late 2013].
"That raises the clock rate high enough to be able to do Ultra HD at 60 Hz a second, albeit it is limited to 8-bit color. So you can already stream 4K content out of a set-top box through a legacy HDMI connection, get it onto a TV, and watch it. But if you want the higher frame rate stuff, the deeper color, then you will have a roadblock. If you are content to do 8-bit color at lower frame rates, then the existing infrastructure with TVs and set-top boxes will already support that. You just need to get the file into those boxes. The caveat there is that as people are getting to know Ultra HD acquisition, editing, and production better, and are also cognizant of things like high frame rates and high dynamic range and wider color gamuts, we are coming to the conclusion that we are probably going to see the sunset for 8-bit color. It won't be adequate anymore. So we are going to go to 10-bit color as a minimum, and possibly to 12-bit color in some applications. One example is high-res gaming off a workstation going into a monitor. That is one sticky wicket that needs to be resolved.
"HDMI 1.4 isn't fast enough to transport Ultra HD signals at frame rates higher than 30 Hz, and HDMI 2.0 isn't fast enough to handle a 60 Hz RGB format Ultra HD signal with anything beyond 8-bit color, as I said," Putman notes. "But the thinking of SMPTE and other standards groups and many people that work in this area is that 8-bit color isn't enough for Ultra HD. Since we need 10 bits to do color, to do expanded color gamut and higher dynamic range, and possibly might even, eventually [need] 12-bit color, then we need a new format for the display interface that is faster than HDMI."
Putman says there are two exciting developments in this area that he expects the industry to explore--DisplayPort and superMHL™.
DisplayPort, of course, is the Video Electronics Standards Association's (VESA)video interface standard developed originally as a computer display interface. But recently, DisplayPort interfaces have begun showing up on some types of 4K televisions, largely to accommodate users interested in gaming applications, according to Putman.
"Remember, momentum is often created to move to higher resolutions and greater bit depths that come out of people using workstations to do gaming, virtual reality, and 3D," he says. "Those video cards are very powerful, so some TV's are now adding DisplayPort specifically for that reason. If someone wants to operate strictly in the RGB domain, and they want to do a 5K, 5120 x 2880 format or something like that, you can forget about getting that through a single HDMI interface--you are going to have to go through DisplayPort, because HDMI just isn't fast enough. That is why manufacturers are already popping DisplayPort onto some TV models, thinking that they should have at least one connector for people who need to go with higher bit rates."
There is also the superMHL™ format, which recently evolved out of the Mobile High-definition Link (MHL) format developed by Silicon Image and members of the MHL Consortium, a mobile AV wired interface that permits users to link mobile devices like phones or tablets to televisions and audio receivers. superMHL™ is the newest flavor of MHL, which is a significantly advanced version that Putman likens to "HDMI on steroids."
"MHL hasn't really caught on for televisions mainly because, if you think about it, a lot of people who have a video on a phone would typically stream it to the TV with WI-FI, if it has that capability," he says. "But [some manufacturers] have supported MHL inputs, and it is quite clever in the way it multiplexes the signal--it can do power, USB, and HDMI over the same connector. And now, the MHL Consortium has announced superMHL™, which as I've seen it, is a much larger connector, a 32-pin symmetrical connector that is more like the size of an HDMI connector that can carry data up to 36 Gbits/sec, which is slightly faster than DisplayPort 1.3. It also uses Display Stream (DSC) compression, so for the first time, we can actually compress display signals. We previously compressed everything else in the video stream, so [DSC] was the last frontier. superMHL™ may be another solution to the puzzle of building up the 4K Ultra HD ecosystem in the long run. It's too early to know who will end up adopting the format, and for the near future, I expect more display manufacturers will use HDMI 2.0, but with the significantly faster bit rates offered by superMHL™, we do have to ask the question if we will need HDMI 2.0 in the future [for televisions] once we go widely to 4K."
Another interesting step forward in the past year, according to Putman, revolves around advances in backlighting with quantum dot technology (QD). A year ago, Putman pointed out that manufacturers were just starting to build the nanocrystal-based semiconductors known as quantum dots into Ultra HD LCD displays to improve backlighting capabilities. Today, he says, "I have lost count of all the companies now showing models with quantum dots. Almost all of them now have it--all the major television brands. This is a technology that very few people had last year, and now, we are seeing it start to go completely mainstream."
Putman says that is a good development, because quantum dots permit displays to reproduce a much larger color gamut.
"Unlike color filters], with a quantum dot, you can get very bright, intense colors without sacrificing saturation," he explains. "You need to have that to be able to do high dynamic range--you need to have the ability to do nominally 15 stops of light from black to full white. And then, you have to be able to do all the color shades and all the shades of gray that fall into those 15 stops, which is, with 10-bit color, billions of colors. Quantum dots are the only thing right now that can do that, outside of lasers, which are pretty much being limited to projectors."
Putman refers to developments like quantum dots and new data interface technologies as "happy accidents" for the display industry, because they illustrate that UHD displays capable of the wider color gamuts and higher dynamic range, which quality 4K content require, are now available and ready to go.
"As usual, the hardware always leads the acceptance," he states. "As with the transition to HD, a lot of other things have to happen for the UHD transition to be total and complete. But from a hardware standpoint, we are already well into it."
David Roth Weiss
David Weiss Productions
David is a Creative COW contributing editor and a forum host of the Apple Final Cut Pro forum.