Those of us in the IT infrastructure world know all about the tectonic trend of convergence on Ethernet. Just about everything, from SCSI to RDMA to PCI, is heading there these days. But Ethernet is conquering other worlds, too. A case in point: HDBaseT, the new standard for HDTV interconnect that sends audio, video, power, data, and USB over standard Ethernet cables!
Everything Must Converge…
The world of home theater is ripe for interconnect convergence. A current setup reminds one of the old data center birds nest, with dozens of different special-purpose interconnects. You might find aS/PDIF or TOSLINK cable for digital audio, or maybe an RCA-terminated coax, alongside standard RCA pairs and phone jacks for analog audio. On the video side you’ll see DVI, HDMI (with or without audio), a variety of component connectors, S-video, composite, and a different kind of coaxial cable. These days, there are likely to be a smattering of data connections (Ethernet, Wi-Fi, proprietary phone plugs, etc). And don’t forget the tangle of power cables, each with a different connector.
Convergence has long been a theme in audio/video circles. The Continent had SCART, which us Yanks totally missed, and we are currently witnessing the crossed HDMI and DisplayPort waves. These two can possibly include sub-gigabit Ethernet signals, and the latter also adds USB to the mix. Then there’s DiiVA, a newcomer with gigabit Ethernet, USB, and 5 watts of DC power to boot. Want to go wireless? Check out WiGig and Wireless HD, which send (compressed) HD video and audio over short distances.
…Or Not
None of these converged home theater interfaces has succeeded even to the level of SCART, though. Incompatibility, incomplete implementations, and customer confusion lead many to fall back to the basics. Many consumers continue to use red-white-yellow analog audio and composite video cables even though they ruin the experience of their expensive HD setups.
Greed is another factor limiting the use of converged interconnects like HDMI. TV retailing giants like Best Buy are thirsty for the huge profit margins they can pull in from overpriced HDMI cables, scaring consumers away from quality. Rather than spring for a $100 cable, many consumers just use the basic bits that come in the box or whatever is already lurking behind the entertainment cabinet.
Wireless interconnects could change this, since they accord automagical connectivity. But early wireless concepts using 802.11n and UWB (Tzero) can’t send uncompressed HD video, let alone 1080p, 3D, multi-channel audio, data, etc. A 720p signal takes about 1.3 Gbps, with 1080i and 1080p bumping this to 1.5 Gbps and 3 Gbps, respectively.
WireFreeHD, WHDI, WiHD, WiGig, and Wireless HD may manage short-range 1080p, but power is another matter entirely. Wireless power exists (see Powermat
HDBaseT: Everything In One Cable…
It’s not wireless, of course, but what if we stuffed everything into a single cheap cable? This appears to be the concept behind HDBaseT: Plenty of video and audio bandwidth, Ethernet, USB, and a full 100 watts of power. Plus, it uses commonly-available Cat 5e or Cat 6 twisterd-pair Ethernet cables and supports common daisy-chain and star network topologies at up to 100 meters per link. This is seriously cool stuff.
Imagine plugging your Blu-Ray player, TiVo, HDTV, and receiver into an Ethernet-like hub, then running links to a few more components in the bedroom, office, and play room. Many of the components wouldn’t even need a power cable, with the receiver or hubs pumping 100 watts into the same cable. This whole setup would be cheaper than a few over-priced HDMI cables and would be much more capable.
…Or Not
On the other hand, HDBaseT might not be all that attractive. It lacks the “awesome” factor found in wireless interconnects, and wired Ethernet hasn’t exactly taken the home market by storm. Consumers have shown they’re much more interested in wireless LAN even when a wire would reach and provider faster connectivity.
And will Joe Sixpack be all that interested in full-bandwidth HD video when he’s content to settle for composite or even modulated RF today? Experience has shown that “good enough” usually wins, and the sharpness and color accuracy of DVD-to-composite-to-HDTV seems to meet that definition.
One-wire connectivity between receiver and HDTV sounds cool, especially when that wire is cheap and thin, but are power cables that objectionable? Perhaps unified power-and-data is a solution in search of a problem.
Then there is the name. As IT folks, we love us some BaseT, but that name means nothing to the average consumer. There’s no special draw for HDBaseT, only its promised features and the probable bundling by sponsors LG, Samsung, and Sony.
And will it even work? The HDBaseT consortium is promising 10.2 Gbps of video and audio data plus 100 Mbps of Ethernet, and (presumably) 480 Mbps of USB over a star- or daisy-chain Ethernet topology. 11 Gbps (to start with 22 Gbps on tap) plus 100 watts of DC power is awfully optimistic for “existing network wiring and field-terminated connectors,” as promised on the HDBaseT web site.
Finally, who approved that terrible logo? Where’s the cool DisplayPort-esque icon?
Stephen’s Stance
HDBaseT might just work. If Sony, LG, and Samsung produce and promote HDTV and receiver pairs that interconnect only with a single cable, and if the TV needs no power cable, I think consumers will be interested. But these same companies are involved with DiiVA, too, and will likely continue to support HDMI for compatibility reasons. Then there’s DisplayPort, which is rapidly gaining ground. How many connectors can an HDTV maker afford to include?
I seriously doubt we’ll see the promise of multi-room HDBaseT realized. Consumers won’t rewire their homes for this, and the minuscule installed base of pre-wired homes won’t support it. Business, on the other hand, might really leverage HDBaseT for conference rooms and customer demo centers. And just imagine the awesome trade show setups enabled by single-cable, no-power connectivity to HD displays! Commercial applications and home theater in-a-box might end up being the two killer apps of HDBaseT!
They call out 100BaseT specifically, which only uses two pairs out of the four in the cat5e/6 cable. I wonder if they run 100BaseT on two pairs, and something entirely different on the other two pairs?
I agree that 100W seems weird. In the PoE standard there was a lot of concern about avoiding the various national electric codes, which all considered “low voltage” to be DC power at about 25W or less. The 25W limit isn’t inherent to the cat5e/6 cable, its a regulatory threshold where the cable plant installation would have to meet the electric code in Japan/Europe/US.