AVB for Video?

At InfoComm 2011 last June, I purposely went looking for video equipment manufacturers that were planning to implement or make use of the new Ethernet AVB standard in their gear. My search was driven by three things: the future, the now, and marketing hype.

Almost everywhere I turned, I found that the future had not yet become the now. The overwhelming majority of manufacturers were ignoring AVB (though they weren’t ignorant!) and had made the decision not to go down that road. AVB is seen as a better alternative to CobraNet, but not something you’d want or need to transmit video. Too many limitations, and their gear already did what was necessary: H.264 video compression with embedded audio, and no need for additional audio or data channels. Another interesting item to note was their perception of IT managers as being reluctant to accept the 25-percent bandwidth limitation for data in the AVB standard.

In other words, the video technology had leapfrogged AVB, leaving it behind as an audio-only eddy. AVB does offer some stunning advantages over CobraNet (as seen by recent product announcements and roll-outs), but it does seem a shame that video-plus-audio applications will apparently be limited.

As for the now, I’ve been left to implement an IPTV network to carry video with embedded audio channels and to use a separate CobraNet approach to carry some dedicated audio-only channels in an application. It’s not the most elegant solution — AVB would be — but it will work. Which still leaves me drooling over the AVB marketing hype and frustrated with manufacturers and end users who, for the most part, appear unwilling or unable to work towards a further integration of AV and IT.

I did find one manufacturer that will be implementing AVB fully in its gear. But not until sometime in the 2012. Sigh. All that’s left is the marketing hype.

About Thom Mullins

Thom Mullins leads the audiovisual design services group for Affiliated Engineers Inc. in Seattle. He is currently chair of InfoComm International’s ICAT Council and has taught classes for both InfoComm and NSCA. For more, visit the Blogger Bios page.

9 Responses to “AVB for Video?”

  1. As promised, here some thoughts on AVB at the ISE in Amsterdam last week.

    There was a booth from the AVnu alliance, who is “an industry forum dedicated to the advancement of professional-quality audio video by promoting the adoption of the IEEE 802.1 Audio Video Bridging (AVB)”. I talked to Lee Minich, Marketing work group chair of AVnu Alliance, and he agreed that there is only a very small amount of companies who even think on developing devices for AVB, especially for Video transmission.

    More worse in this context is in my opinion, that Cisco, one of the founding promotors of AVnu, does not sell, until now, any product which is AVB certified,

    In this context one of the major problems of AVB occures. AVB can only rise, if the basic infrastructure is implemented, this means without the broad availability of network components as i.e. switches, it would be very very dificult to convince your clients to invest in AVB equipment.

    If you talk about Standardization especially about AVB, you have to talk also about HDBaseT, an alliance with some important players in the CE & the ProAV markets (i.e. LG, Samsung, Sony). The alliance wants to set a ‘proprietary standard’ for the 5Play convergence of uncompressed full HD digital video, audio, 100BaseT Ethernet, power over Ethernet and control signals. It was not astonishing that the HDBaseT Booth on the ISE was a little bit ‘consumer’ driven, even if they always try to mention that her ‘industry standard’ is very important also for the ProAV business.

    As a result of these cognitions, I would predict that it will take a few more years until really useable and certified AVB products with the ‘V’ part come out. And maybe the CE market will be a stronger driver than the ProAV market.

    Last, but not least, I want to point out that I will publish a more detailed report on these topic next week on G+ and on other platforms.

  2. I will do a similar thing as you on the ISE in Amsterdam next week. There is an education session scheduled named “AVB: The future of standards-based AV networking” but the speaker (Jerry Placken) comes from Meyer Sound and is in the audio industry for over 30 years. And no word about video in the session description! So I’m afraid, it will not be a better result on Video at all. I will post news after the show…

  3. Dont sure I get your points reg. AVB vs. H.264.
    Isn’t it perfectly ok to send a H.264 compressed stream over AVB?
    Maybe codecs like H.264 could be adepted to use less bandwidth when there is almost zero risk of packet lost, etc. because AVB takes care of all the pitfalls compared to sending H.264 over a dumb network.

    • Yes, it’s perfectly acceptable to do so. In fact, that’s what I hear chip makers recommending. There are two problems. First, most people transmitting audio over UTP either embed the audio or have a need for a limited number of separate audio streams. Second, as a result of this, most manufacturers (HaiVision, etc.) see no need for the AVB standard as they are already doing what they need, successfully, in other ways. Therefore, AVB appears to be headed for a backwater life in the pro audio arena.

  4. I share your frustration. It appears the video world prefers to focus on compressed streaming; H.264 seems particularly popular. But, many of my applications involve integration with production facilities, and these folks want uncompressed video to do with as they need. At first, I thought the lack of effort in developing AVB video endpoints was related to network limitations. But with the advent of 10G, and now 40G/100G backbones, it’s clear this isn’t the case. I expect it’s more a matter of economics; there just aren’t enough applications requiring uncompressed video. So, for the time being, I’ll have to continue using point-to-point solutions. FYI, XMOS is demonstrating MPEG-2 over AVB at CES and NAMM.

    • This has always been part of my frustration with AVB – it seemed short-sighted to create a system that has inherent limitations on the optimum video stream size. Even with the advent of wider bandwidth networks, running multiple streams of full-bandwidth video around a facility will clog a pipe pretty quickly. The extensive use of H.264 also speaks to the wide acceptance of Lowest Common Denominator in video (and audio!) quality. I think that will shoot us in the foot one day.

      Thanks for the heads up on XMOS. I’ll check them out on line.

  5. Was there ever any discussion of AVB being used as an alternative to standard source-to-destination, PC-to-Projector cabling?

    • Not that I ever heard, Rob. I think the concept was more for backbone applications. Providing an endpoint at a projector might be relatively expensive, given the need for software codecs at each location to compress and reconstitute video.