Newswatch e-newsletter

Current Issue - February 2015

SMPTE Newswatch Masthead

Hot Button Discussion

The UHD Chain, One Link at a Time
By Michael Goldman  
 
As new technological developments make it increasingly clear that the broadcast industry's commitment to pushing past HD and into the Ultra High Definition (UHD) realm is continuing to pick up steam, it's also clear that this is a broad, open-ended, and somewhat fuzzy process. When the initiative gets boiled down to specific aspects of the broadcast chain, the industry remains far from having all the answers in terms of how to implement a full UHD broadcasting ecosystem any time in the near future. Andy Quested, the BBC's head of technology for HD and UHDTV initiatives and a longtime participant in several bodies that are working on these ongoing changes--including SMPTE, where he serves as a Fellow, the UK's Digital TV Group, and the ITU's (International Telecommunication's Union) Image Dynamic Range group-emphasizes that broadcasters will eventually need to examine the UHD paradigm from the perspective of the entire chain, link by link, and how best to connect all the links together. Therefore, although certain links in the chain, like high-resolution digital acquisition systems, or 4K-capable UHD televisions for consumers, are already starting to mature, the soup-to-nuts infrastructure to create, manage, store, transmit, and view true UHD (4K or above) imagery as it moves from the content creator to the broadcaster to the home viewer to enjoy a clearly improved viewing experience still has miles to travel before it becomes ubiquitous.
 
"We started getting high-resolution digital cameras a few years ago with the like of Arri and Red and Sony cameras," Quested says. "Now, many of these can do 4K or better on the imaging sensor; we have a pile of cameras that have far better resolution at the front end than we had before, in fact. And then, on the back end, we are getting displays available that have reasonably high resolution relative to HD. So when it comes to UHD, we are already able to make longform programs with cameras capable of delivering UHD resolution imagery. We do still have a problem with those cameras in terms of depth-of-field for sports and live events. Large sensors and shallow depth of field are not the best for events where all the play needs to be clear and sharp. But we are starting to see that get sorted out, with newer 2/3-in. sensor 'system cameras' just arriving. So I hope we will see more productions using multicamera trucks very soon.

And then, unlike when HD first came in, there are alternative ways of delivering the signal. In fact, the BBC has already done its first UHD 'transmissions' without making a single change to our transmissional playout chain. Check out the BBC's blog series from 2014 on how the network used IP methods to transmit UHD signals from Brazil during the World Cup in the summer of 2014. That's because we didn't use it. We sent the signal directly to BBC iPlayer [the BBC's television and radio programming Internet streaming service]. So broadcasters are now seeing they can use hybrid systems for delivery--traditional transmission for non-UHD signals and IP transmission to give you a better experience with UHD signals."
 
As previously examined in Newswatch, consumers on what Quested calls "the back end" now have suddenly readily available UHD-capable televisions on which they receive and watch content. Thus, the "front end" of acquisition, combined with IP methods of delivery using a growing number of streaming technologies, and the "back end" where consumers can view 4K content on first-generation UHD televisions are all now readily available, he points out. Yet, there is perilously little UHD content to be found, and no dedicated UHD broadcast services right now. Why is that?

Quested explains that this due to the "middle bits" in the chain-the ongoing hurdle involving transitioning broadcast plant infrastructures, with some broadcasters still not having fully completed the transition from an SD to an HD infrastructure in certain places. As discussed in Newswatch last year, the conversion process for the broadcast industry from physical hardware and media to IT-based infrastructures has been slow, expensive, and difficult due to the high costs and cultural shifts that need to be made across virtually the entire broadcast industry. This makes it feasible for broadcasters to move 4K-and-higher data around the new, state-of-the-art, data-centric facilities they will first have to create, and then master and store it as needed to "churn out" content efficiently for production and distribution.
 
"So we are moving from both ends toward the middle, if the middle is mastering and storage, playout, and archive," Quested suggests. "We have the acquisition, and it is gradually getting easier to move further up through the post chain toward the final master. And if you look at the other end, the consumer is able to receive the content one way or another. We have the HEVC codec and the bandwidths are coming down for how we need to transmit it using IP protocols, and we are getting the encoding systems together. But that bit in the middle is the expensive problem-the traditional infrastructure for what we could call a channel, and a traditional infrastructure for mastering long-term storage and moving content around. Our own infrastructure is slowing us down. Who has written off their HD investment yet? And when do they think they will want to do their next investment? And what of several possible technology choices will they want to go with? So it's really a business decision, because the technology part of it is getting there quickly."
 
Quested points out that dozens of models exist for IP-based networks and studios. The BBC, in fact, has an ongoing R&D project that details its approach to the problem, and as noted in the January 2015 Newswatch, ESPN's new DC2 broadcast facility is entirely IP-based, while others across the industry are on the way-the types of facilities that could potentially be producing and broadcasting UHD programming when the time is right. However, for the time being, in the grand scheme of things, all these developments-from high-resolution cameras to IP-based studios and IP-based delivery to UHD consumer monitors-are essentially providing, for lack of a better term, an enhanced HD viewing experience, rather than a true UHD viewing experience, Quested states.
 
"In a sense, it is a major risk to make HD look too good, because why would anybody want to throw huge amounts of money into production and bandwidth to bump up the viewing experience just a small amount more?" he wonders. "I have worried about '4K' ever since it was put into the UHD standard. I prefer the term UHD, since we are talking about getting something significantly better than current HD out of the technology at the end of the day."
 
This is why Quested believes the industry won't be able to settle for simply improving the front end or the back end or various other links along the chain to improve the HD experience indefinitely-the rewards of doing so in terms of a visibly better picture that consumers will notice, appreciate, and spend money on are rapidly reaching their peak. He points out that even today's finest HDTV's and UHDTV's are succeeding in translating 4K imagery, largely through the use of over-sampling techniques, "basically getting the best HD you can by using UHD cameras and displays," he explains. But, he believes that methodology has gone as far as it can go, considering that studies still show that consumers are not sitting at recommended distances from their fancy new monitors to fully absorb the better picture quality that can be provided today, and they are not likely to do so any time soon. Therefore, only pristine UHD content getting to their homes for viewing on clearly larger and improved UHD monitors one way or another is bound to make a true, discernable difference with consumers. The only question he asks, therefore, is how long will it take for consumers to care enough for it to make good business sense to go that extra mile on every link in the chain-rather than just some of them-to make this happen.

"So there is a dichotomy regarding how much better we can make the HD viewing experience in a way that people will notice or care," he says. "Also, testing on viewing distances by the EBU and a recently published white paper by the BBC recently published a white paper on viewing distances showed that people have not moved either closer or further away from their televisions than many years ago, but their televisions have gotten much larger. In the UK, [people] are still sitting about 2.7 meters from their television, like they were the last time [the BBC did such a study in 2004]. This means we are still sitting too far away from the recommended viewing distance for most televisions, [moreover] we are getting close to the threshold on screen size in the UK, and I suspect most other places, where we could benefit from an entirely new system. We aren't quite there yet, but we are getting close. In other words, HD will eventually start to fail in the sense that the viewer will soon be able to easily see the difference between a good HD picture and a UHD picture, regardless of your transport stream or camera or display."

Quested says, although the landscape is greatly in flux and the UHD chain is being built incrementally rather than link-by-link in linear order, there is no doubt that the industry is moving toward the creation and distribution of significantly higher-resolution imagery to replace HD as the standard broadcasting format. He just does not see the transition as replicating how things went down during the transition from analog to HD viewing, nor does he think we are headed to a one-size-fits-all UHD broadcast standard any time soon.
 
"In fact," Quested adds, "the big difference between how HD arrived and how UHD is coming in does not actually have much to do with the images themselves. It has to do with the options for getting them to people."
 
Instead, Quested says he expects the industry to put great resources and attention toward continuing to expand the Internet-based, streaming distribution methods it has been pioneering in recent years with various services like Netflix and others, with increasingly sophisticated set-top box technologies for essentially automating the process of matching signals heading to consumer televisions. This is similar to the notion of HD "channel swapping" with set-top boxes in certain territories around the world, whereby the consumer's set-top box maintains the same channel number for particular channels or networks but selects the correct signal-SD, HD, or eventually, UHD-that best matches the consumer's television. These innovations, he suggests, will allow broadcasters to increase the volume of UHD content they produce and distribute, regardless of how long it takes the industry as a whole to transition the aforementioned "middle bits" of their infrastructures to result in a complete and total UHD ecosystem.
 
For the foreseeable future, however, all of this action is aimed at the first level of UHD listed in ITU-R Recommendation BT.2020-so-called UHD-1 (2160p/4K resolution). The second level is called UHD-2 (4320p/8K), as discussed previously in Newswatch, which is often called Super Hi-Vision after the name given by Japanese broadcaster NHK to its specific flavor of UHD-2. NHK recently reaffirmed its commitment to broadcast the 2020 Tokyo Olympics in Super Hi-Vision/8k resolution via satellite, thus making that date something of an unofficial demarcation point for when some people think 8K broadcasting will somehow become viable. Quested, however, isn't sure that 8K broadcasting will ever translate into a dedicated broadcasting service.
 
"Even the Korean broadcasters working on UHD-2 and NHK are not saying there will be dedicated [8K] services for many years," Quested says. "But there could be a mixture, which is interesting, where you might never have an entire 8K service or channel, but you might have 8K programming, like the 2020 Olympics, which is something the Japanese are very focused on. I doubt we will know their plans, or anyone else's, for standard Super Hi-Vision services, until after that event in 2020. But those Olympics do tell you how short five or six years can be. They have to build an entire infrastructure just to do the largest broadcast ever in the world. And in the meantime, experiments will continue. We will need to see some content.
 
"But I tend to think consumers will be taking their time with UHD. There is some consumer fatigue out there. and there are business decisions for the broadcasters to make. Do they wait, because it will be cheaper to jump in later? Or do they jump in early to take whatever premium they can? These are the ongoing debates, but for UHD-1, the pieces are coming together. SMPTE has an HDR standard, primarily targeted at non-broadcast content, and the ITU is working on HDR standards and HD SDI options. The [Digital Video Broadcasting Project] and MPEG are working on [these issues]. Also, HEVC, and many of groups, are working on standards and reporting on the status of all the new technology and innovations. The truth is, it will be IP-based, since we have vast new increases in capabilities for getting data to consumers. Maybe cellular networks? Think about 5G--will your television soon have a SIM card in it? Is there any reason why it shouldn't? Will the consumer even know or care where their TV is pulling the content from, as long as it looks good? For them, it will just be about the quality of the experience, and those possibilities are great. For the broadcaster, we have to remember that we are aiming for those eyeballs sitting on the sofa or watching a high-resolution portable. But consumers don't really care how they get the pictures, as long as they don't have to do anything difficult to get it."
 
News Briefs
HEVC's Low Bandwidth Benefits
A recent TV Technology column by Ian MacSpadden discusses the expected consequences of the widespread adoption of the High Efficiency Video Coding compression scheme (HEVC or H.265), which MacSpadden suggests is likely to "make the biggest impact of any new TV technology in coming months." The reason for this, he believes, is that HEVC and the corresponding improvements that it will permit broadcasters to bring, in terms of greater resolution, color depth, and faster frame rates, among other things, will not only make UHD broadcasts more viable in the coming months and years, which is the primary reason for the codec's introduction following its formal adoption as a new video compression standard last year. It will also have directly benefit ongoing broadcasts of high-definition and even standard-definition content, as well. The article quotes one compression expert, who explains that "low-bandwidth viewers" may end up being among the primary early beneficiaries of HEVC, because they will suddenly be able to receive 720p images over the same delivery path that once could only provide them with 480p images. This means that "leveraging the new codec to deliver content over existing distribution systems" will be among its primary, initial accomplishments. The article also examines the many eventual benefits of 4K delivery, which will now be possible on the high end, thanks to HEVC's arrival.
 
Quarter Century of Photoshop
The Internet has recently been flooded with tributes to Adobe's Photoshop photo editing and graphics software because the historic, game-changing technology turned 25 years old in February. A quick Google search of the term "Photoshop turns 25" will result in dozens of tributes, articles, blogs, and videos about the anniversary's importance. Among the most interesting is an NPR article and corresponding All Things Considered radio interview with Photoshop co-creator, Tom Knoll, about how he and his brother, John Knoll, now a well-known Academy Award-winning visual effects supervisor for ILM, developed image processing software while experimenting with early computer graphics' technologies 25 years ago-software that eventually turned into Photoshop. They sold the software to Adobe in 1990, and as the article points out, "25 years later, Adobe Photoshop is still the industry standard for photo editing software." Elsewhere, this Adweek article discusses Adobe's all-out effort to celebrate by creating a tripped-out Photoshopped ad set to Aerosmith's Dream On, in honor of the software's birthday. Also, this Seattle Times article examines how new social media services like Instagram are offering some of the basic functions of Photoshop for consumers who do not need to do professional level work to touch up their own photos--applications they would likely never have thought to offer, were it not for Photoshop's influence. Indeed, as the article points out, we can understand Photoshop's importance merely by the fact that it has "attained the rare status of a product that has become a verb, like Google and Xerox."
 
Another Virtual Reality (VR) Approach
January's issue of the Newswatch pointed out the arrival of the Microsoft HoloLens hologram VR technology system  as another step forward in the increasingly vibrant virtual-reality field. That system allows users to blend holographic 3D imagery with things they see in the real world through a sophisticated headset. Such developments are now coming at a furious pace, with more news out of the VR world about the arrival of the Zeiss VR One headset, which offers up yet another paradigm shift on the basic VR concept. Zeiss' approach with VR One is to offer a headset that fits with app-enabled smartphones to provide both virtual reality experiences such as gaming and augmented reality experiences, in addition to an entirely new way to watch movies, particularly those of the immersive variety, depending on the mode and application users select. A recent Pro Video Coalition article reported that the VR One system recently earned an iF Design Award in the Audio/Video product category for its "virtual movie theater" capability because, among other things, it allows users to enjoy movies (currently at 720p resolution) in immersive environments similar to a surround-style 3D theater, in which you can add yourself and others as virtual characters, if you choose.