4k Directions

August 5, 2014, Emerging Display Technology Conference, San Jose, CA—A trio of pundits and technologists talked about the drivers and markets for 4k TV sets. Paul Gagnon from NPD Displaysearch joined Graham Loveridge from Pixelworks and Mick Grover from Sharp in a set of presentations.

Gagnon noted that the market is growing for all types of devices from TV to PC, tablet, cameras, and phones. All of these areas are changing resolution over time as the technologies and manufacturing capabilities improve. At the large end of the spectrum, TV sets have seen waves of adoption. For a TV, 4k is relatively easy, only requiring about 100ppi. On smaller screens, on the other hand, getting beyond HD is fairly hard, wit minimum resolution of over 400ppi, depending upon the size of the display.

TV is likely to see a breakthrough year with a possible 13M 4k units sold. To keep the number s in perspective, the total TV sales will be over 250M units. The area with the greatest adoption is China, with three quarters of all sets claiming 4k resolution. The caveats for China are that the 4k sets are social status indicators and performance doesn’t matter or the users don’t care. The sets have poor upscaling capabilities as brands and retailers sacrifice functionality for lower prices. The premium for a 4k set is only 40 percent greater than for a HD set.

In other areas, the 4k sets are positioned as ultra-premium devices. Although some content exists, the lack of choices for content and any realistic delivery mechanisms matter to the buyers. In addition, buyers are distrustful of the lists of features as a reflection of the debacle for 3-D. In general, buyers are more concerned with size than with 4k capabilities.

The strongest target is for TVs that are 50 inches or larger, and the 4k sets have a lower percentage premium at 50 inches versus a HD set. The supply chain is in transition and may cut the supply at certain levels and sizes to drive demand for other sizes and feature sets. For example, expect to see discontinuation of HD sets at sizes over 55 inches. The manufacturers are pushing the 4k sets and have targets of over 20M $K TVs in ’15.
Overall, 4k TVs are seeing a fairly low premium of between 7 and 54 percent over HD in all sizes.

In non-TV displays, notebook and tablet sales are growing fairly slowly, at 1 to 7 percent growth per year. In the mobile space, the 4k displays command a high premium, but there are no native 4k apps. In addition, the phones need to have high-end processing to fully use the 4k capabilities. As a result, the phones will see the high prices for 4k displays as a niche with at most a 1 percent market share.

To make the transition even more difficult, there is a hardware-software mismatch. There are no standard formats and we just went through the HD changes. Content distribution is a major challenge as broadband is too slow and Blu-Ray hasn’t settled on a standard format. The only reasonable alternative is to scale HD up to 4k. This process, however, only provides marginal benefits, as 4k needs high-dynamic range, high frame rates, and wider color gamut to realize the full potential of the technology.

Grover stated that the road to 4k will take a long time before many users adopt the technology. The content ecosystem and availability of viable distribution media will hold back the market growth. The ecosystem must engender a full value chain, from capture through edit, infrastructure management, to delivery. Some of the challenges are related to the content.

Digital cinema is a non-real-time function that provides the creators the luxury of time to get the images and effects carefully crafted. News is a pseudo-non-real-time function that allows some pre-processing of content that has to be mixed with real-time images. news can be captured on a 2/3″ UHD camera and gets transferred to the studio in a bandwidth-limited medium. In both cases, the fine details are processed off-line.

The hardest material is live production, which requires real-time interactions and image processing. The live action needs better small cameras and still lacks a cabling standard. The 4 SD-I cables for HD need to be replaced with something, but nothing is defined as workable yet.

Content delivery is another issue. There is no equivalent to the Blu-Ray standard for disks and video quality for 4k. Some content is available through OTT vehicles, but there are long buffer times or downloads. Currently there is no mechanism to deliver content over broadcast or cable.

The adoption timeline will have to consider 4k as a novelty for the intermediate term. This year Netfix will start UHD delivery content through broadband channels. Some experiments are ongoing for satellite video on demand, and cable is expected to start trials next year. Slowing the adoption is the expected release of the next ATSC specification and a Blu-Ray disk late next year.

Early adopters will have access to content in ’16, with a big push to get the world’s infrastructure ready for the next World Cup in ’18. Wide scale adoption will start in earnest in ’19 to be ready for the ’20 Olympics. 4k distribution will be on satellite and cable and the other media will move to a client-server structure using DLNA, UPNP and DTCP/IP standards.

New technologies for cable will move to quadrature division keying and ATSC 3.0, but the technologies still lag the displays. Global 4k adoption is progressing faster outside the US. Japan is getting daily UHD broadcasts from 11AM to 7 PM in selected markets and NHK is planning to broadcast the Tokyo Olympics in 8k. The EU is proposing a formal launch in ’16 and China is looking to a ’18 broadcast start.

The mobile space is interesting. The MHL 3.0 and Miracast 2.0 standards allow for viewing 4k TV from a small device on a large display. Direct broadcast to mobile devices is likely to come sooner than to fixed devices. Beyond the base resolution, new technologies like HDR, HFR and wider color gamut are intriguing to the studios, and will be the driver for 4k display adoption.

Loveridge described how to unlock the potential of 4k display processing. Most observers agree that 4k increases realism, and provide crisper images, but many have doubts about the value of those benefits. The doubters say that 4k has no content, only works on big screens, and takes immense amounts of power and memory bandwidth.

Content already exists and is available in some OTT services. Much of the content is user generated, but the various studios are starting to create in native 4k in anticipation of delivery capabilities in the near future. The static acuity of the eye defines the minimal resolution of an eye at a distance. Dynamic acuity indicates the maximum frequency for resolving separation of two lines.

The eye can resolve sub-pixel images as changes in brightness or sharpness. When considering 4k content, the eye can resolve 4k images on a 5 inch screen at 10 inches and the same image on a 60 inch screen at 10 feet.

As a result, all displays can benefit from enhanced video image processing to reduce artifacts. When done well, the image is improved with higher resolution. When done incorrectly or not at all, the image will display artifacts independent of the viewing distance. For example, judder is the temporal displacement of an image relative to our predicted displacement. This motion blur decreased perceived resolution and image quality.

Although traditional image processing is memory, bandwidth, and compute intensive, changes in architecture can return the need for higher clock speeds and increased power consumption. Video display processing eases power and memory bandwidth issues of higher resolution displays through concepts like PSR (panel self refresh) and MBO (media buffer optimization). These processes can be easily implemented to enable CPU and GPU to enter a lower power state more often. Even elements of UI can be offloaded to the video display processing circuitry. Adding video display processing to a system is necessary in mobile systems to remove artifacts, reduce judder, and reduce system power and memory bandwidth issues.