Connecting Systems To Displays: On The Air

This blog post references my cover story 'Connecting Systems To Displays: What We Got Here Is A Failure To Communicate' in EDN's January 4, 2007 edition.

As I alluded to at the end of the wiring-filled first two paragraphs of my writeup, it's no longer a given that cabling is required to route A/V information from one place to another. A number of wireless schemes already exists, and others are planned, that can ship standard- and high-definition images and their accompanying soundtracks both across a room (several-inch to –foot PAN, or personal area network, spans) and room-to-room (LAN, or local area network, links).

Belkin's PureAV RemoteTV, which first went on sale in late 2004, was in many respects a pioneer in this field, albeit an expensive one. As Matt Miller's early-2005 hands-on review pointed out, Belkin claimed that PureAV, which employed Magis Networks' proprietary Air5 5 GHz protocol, was able to reliably handle 40 Mbps of data. However, RemoteTV's reliance on mature-but-bit inefficient MPEG-2 encoding, coupled with the cost-driven decision to support only up-to 480-line interlaced resolution inputs, left it incapable of handling high-def (or even progressive-scan standard-def) streams.

RemoteTV didn't accept digital video inputs or offer digital video outputs, meaning that the signal chain from (potentially) digital video source to (potentially, with a non-CRT monitor) digital display destination was marked by multiple quality-degrading digital-to-analog and analog-to-digital video conversions. RemoteTV also only handled two-channel audio, which meant that while it could comprehend matrix surround encoding such as Dolby Pro Logic, discrete greater-than-two-channel audio sources like Dolby Digital or DTS would need to be permanently down-mixed prior to transmission.

More recently, in early September, Tzero Technologies announced (more from Ars Technica) that it'd be using Analog Devices' wavelet compression chips to implement a wireless version of HDMI over a UWB (ultrawideband) backbone. Wavelet compression is the foundation of the JPEG 2000 format and is also employed in the Digital Cinema Initiative standard. Although UWB backers vigorously tout the technology's claimed 480 Mbps peak bandwidth, real-life protocol overhead coupled with distance and intermediary interference factors (translating to packet delay and loss), mean that fairly aggressive lossy wavelet compression of the HDMI video stream will likely be required to hit the partners' 720- and 1080-line high-def ambitions.

As a 1998 (!!!) article by yours truly (also check out the website addendum) points out, wavelet compression's artifacts are often, at a given compression ratio, judged by viewers as less objectionable than the blocky artifacts of traditional compression schemes like MPEG-2, MPEG-4 and VC-1. The proof's in the demo, however, and the interference-rich show floor at CES will provide ample opportunity for me to see what reality's behind the Wireless HDMI hype….that is, if the promised demos come to fruition. UWB vendors don't exactly have stellar track records, after all, of meeting their schedule, range, or performance promises. Stay tuned.

If the idea of lossy-compressing otherwise pristine digital video doesn't thrill you, then the WirelessHD consortium wants to bend your ear. Their spec-in-early-2007 scheme employs a 60 Ghz (!!!) unlicensed-spectrum protocol developed by SiBEAM to deliver 5 Gbps data rates at over-30 foot distances, with no more than 15 ms source-to-destination latency. To SiBEAM's credit, it's assembled a formidable list of consumer electronics systems partners; LG, Matsushita (the parent company of Panasonic), Samsung, Sony, and Toshiba.

But wait, there's more….SiBEAM's late-October announcement was followed by, earlier this month, a third wireless A/V announcement, this from a company called Neosonik. The firm plans to use the 5 GHz unlicensed band to transmit up-to-1080i video streams but, because its protocol is proprietary, it claims it won't interfere with IEEE 802.11a traffic. I'll believe it when I see it. And, according to company officials, I'll be able to see it at CES in a bit over a week's time. Neosonik is pursuing a two-prong strategy; selling company-branded audio-only transmitter and receiver-inclusive speaker sets, while pursuing licensing deals in non-competing audio and A/V applications.

And speaking of 5 GHz, let's not forget Amimon, who plans to leverage conventional 802.11n, albeit in a QoS-optimized manner….

It'll be very interesting to see how this current startup skirmish of dueling press releases and PowerPoint presentations sorts out when the financing and development schedules start getting tight. In Las Vegas at CES, or wherever you do your gambling, how are you placing your technology bets?

Digg This | Slashdot This