Note from the Editors: We have written many articles about the emergence of digital media adapters--networked devices which convert media content typically stored on PCs for use on TVs. We have used DMAs in our own home for many years to listen to music and to view photos and videos on our home theater system. We are pleased to present this guest article by Keri Waters of Micronas USA, which raises the question of whether DMA functionality should better be incorporated into networked digital TV sets.
Keri Waters is Director, Strategic Marketing of Micronas USA. Ms. Waters is responsible for software strategy across the entire product line, including features for development in-house, as well as all relationships with software partners. She brings over 10 years of experience in engineering, sales and marketing. She previously held positions at Molino Networks, Area Systems and Cisco Systems. Ms. Waters holds a B.S. in Mechanical Engineering from Massachusetts Institute of Technology.
Micronas, headquartered in Zurich, is a semiconductor designer and manufacturer with worldwide operations, and a leader in innovative global TV system solutions with expertise in the digitization of audio and video content; the Micronas Group employs about 2000 people and had 2005 sales of USD 676/EUR 545 million. Micronas USA, a wholly owned subsidiary, develops and delivers a diverse portfolio of integrated real-time streaming audio/video encoder and decoder products targeted at the world’s most significant digital communications markets.
NDTV--A New Type of Product Within the Home Network
Recently, discussion of the convergence of the Internet and television within the home has focused mainly on IPTV, a new type of managed service for delivering premium television content to subscribers via broadband. While field trials are underway and the market is emerging nicely, IPTV alone will not address all of the consumer expectations for simplifying storage and access to their digital media. The expanded menu of on-demand premium content offerings available via IPTV will certainly be attractive to consumers, but in the truly converged home, personal digital content—-photos, videos, MP3 playlists—-will be accessed from the home theater just as easily.
Today, this type of personal content tends to be stored on the home PC. The PC is an ideal platform for organizing and storing media, but not necessarily for viewing it with others. Digital Media Adapters (DMAs)-—networked set-top boxes designed to be a bridge from the PC to the TV, with a user interface for browsing and playing media-—have been on the market for the last 3-4 years, but have sold poorly. So far, consumers have been unwilling to accept an additional box in the home theater stack, and the DMA units themselves have been plagued by poor reviews of their quality and usability.
However, personal digital content continues to proliferate (how many people have gigabytes of precious photos stored haphazardly on their hard drives?), and interoperability and digital rights management standards have matured to the point that the quality concerns are diminishing. Many people feel that DMA functionality ultimately will become integrated into some other device already in the home theater – in another set-top box, perhaps, or in the television itself.
The latter type of product shows a great deal of promise. According to S2Data, the market for so-called Networked Digital Televisions (NDTVs)—-that is, digital TVs with integrated DMAs—-will grow to 27 million units by 2010. LCD-TV manufacturers are scrambling for differentiating features to avoid commoditization, and consumers are beginning to recognize a need for DMA functionality within the home.
NDTV Design: Analog and Digital TV
A typical NDTV brings together three functional elements in a single chassis: analog TV, digital TV, and a DMA. An 8-bit microcontroller runs software that manages the chassis control (parsing of remote control inputs and so forth) and all of the functionality required by traditional analog TV. The complexity of just this piece can be surprising, since there are a large number of broadcasters that may deviate from broadcast standards just slightly, and the onus of decoding whatever broadcast stream is sent rests firmly with the TV manufacturer. Thus, several megabytes of software may be required just to support basic analog TV functions.
With the government-mandated analog broadcast spectrum shut-off rolling out in most developed nations over the next few years, new televisions must also contain a digital tuner and digital video decoder. The complexity of the digital TV stack is far higher than that for analog, in that digital broadcast standards include specifications for advanced features like program guide data, secondary audio streams, advanced closed captioning and teletext--all transmitted in the broadcast stream. The software to support all this must run on a real embedded operating system like Linux or WinCE, and requires massive amounts of field testing to support the intricacies of each local region. The digital video stream may be in one of several commonly-used formats such as MPEG-2, MPEG-4, or H.264/AVC, may be in standard definition or high-definition resolution, and may be in either traditional 4:3 aspect ratio or widescreen 16:9. Thus, a sophisticated hardware multi-format video decoder is employed in the design to reduce the cost of managing all the variations.
Adding DMA Functionality
Since the TV design already includes an operating system and a multi-format video decoder—-which form the basis for DMA platforms—-manufacturers see an opportunity to add outstanding feature differentiation to their digital TVs with only the incremental cost of the DMA software and networking components. The attached diagram shows how DMA functionality could be incorporated in the NDTV.
In taking this logical step forward, however, manufacturers are once again greatly increasing the complexity of the design. DMAs must be able to play all of the media that users collect on their home PCs, in all of the commonly-found formats and resolutions. Video formats include the traditional broadcast formats mentioned above, but also include other codecs common on the Internet such as Quicktime, Flash, DivX, and Xvid. Other media formats that must be supported include MP3 audio and JPEG images. Decoding for all of these additional formats may be done in software if the hardware decoder contains a powerful enough CPU for applications.
The DMA must also include a rich and intuitive user interface for browsing and selecting media from a remote PC, as well as support for many popular media streaming protocols such as RTSP, WMS, and DVB-IPI. To discover the media repositories available on the local network, the DMA stack must include support for common protocols such as UPnP and Bonjour. All of these requirements and more are generally dictated by manufacturer specifications such as Intel’s Viiv and Microsoft’s Rally Technologies, and by emerging standards bodies such as the Digital Living Network Alliance (DLNA); certification by these bodies becomes a paramount concern.
The NDTV requires a networked connection to media sources. Most early models of NDTVs will include wired 10/100 Ethernet and wireless 802.11b/g support, given that these are the de facto home network standards. However, the average home network today exists mainly to allow family members to share a single DSL or cable connection for high-speed Internet access, and perhaps also to share a common printer. DMA functionality demands a new level of performance from home networks.
Today, the most likely use for a NDTV will be audio and photo playback, for which 802.11g is probably good enough. However, the average lifespan of a LCD-TV is expected to be 5-10 years, so whatever wireless solution is chosen, it must be future-proof against upcoming need for video streaming. A 10/100 wired network may be able to deliver the performance needed for video, but most homes don’t have Ethernet jacks near the television. 802.11e, the extension to the 802.11 standard that introduces the concept of QoS to 802.11 networks, will allow NDTVs to enjoy the full wireless bandwidth available between it and the PC. With modern high-compression video codecs, this will come close to being good enough for one HD stream. Still, this may not meet the stringent video quality requirements of the consumer, who will invest a significant amount of disposable income in this new product.
Since NDTVs will be used mainly for file-based playback, they will have some advantages in the home network over IPTV systems, which must deliver some real-time content. File-based streaming may be buffered, improving video quality over narrow bandwidth networks. However, consumers will have little patience for long initial waiting periods after selecting a piece of media to play, and will likely want to play media in more than one room of the house at a time.
For second generation NDTVs, there are some emerging alternatives for the home network. Powerline networking may be attractive, since the television is sure to require power from the home. However, powerline hasn’t yet reached the bandwidth of wireless networks, so it’s not yet a serious contender in places where wireless is feasible. Coaxial-based networks, such as are promoted by MoCA, are another alternative, particularly favored by cable companies that wish to play in this market. New forms of wireless such as UWB may also be considered, but the current limited range will limit usefulness in larger homes. 802.11n, which allegedly will increase bandwidth over 802.11g networks ten-fold, is a compelling option, but the standard is not due to be complete before the first generation of NDTV designs will be finalized.
In all cases, however, TV manufacturers are cautious about increasing their bill of materials (BOM) cost for new standards that may not catch on. Close cooperation between networking companies and TV hardware companies will be required to develop an industry-wide recommendation for a standard. In the meantime, NDTV manufacturers of early devices should consider how to build in expandability for future networking technologies. Integration with the home network is a critical and complex aspect of NDTV design.
Consumers are counting on the consumer electronics industry to deliver on the promise made by the flatpanel TV ads: the television as the centerpiece in a sleek minimalist living room, floating unencumbered on a cool white wall; a serene-looking woman sitting on the stylish sofa, effortlessly navigating a crisp, gorgeous UI to find any piece of media she wishes. Pointedly missing from the visual message are stacks of media, set-top boxes, wires, user manuals, and bad image quality. The bar is high, but the building blocks for the successful design exist, and soon the television experience will be transformed.