Digital Audio Networking – What I Learned at InfoComm 2013
by Chris Foreman, Wordworks
In this article, Chris Foreman shared some information that he recently learned on Digital Audio Networking.
As an old analog guy, I sometimes long for the days when connecting things was easy. All we had to worry about was impedance and level, headroom and noise, polarity and pin-out, grounding and shielding and so on. Come to think of it, audio connections weren’t exactly easy, even in those analog days. And, now, in our digital era, I’m facing a whole new set of connection issues. I guess I’d get bored if things were truly simple.
Thus, one of my goals for this year’s InfoComm was to further my education in digital audio networking. To this end, I spent quite a bit of time with networking vendors and I attended an InfoComm University class, called “Modern Audio Networking”, taught by Rodrigo Ordonez of K2 Audio.
Ordonez’ class packed a detailed overview of the entire subject of digital media networking into an hour and a half – one of the best I’ve experienced at a trade show. Also, I got a lot of good information from vendors on the floor. So, I feel like maybe I’ve advanced from novice to beginner. And, from that limited point of view, here are some thoughts about what I saw at the show and where we seem to be headed.
In the future, most installed systems audio will travel on a facility IP network.
Ordonez’ class was based on this assumption and many InfoComm vendors were moving towards this model, most often via AVB compatibility. This trend may be delayed for applications like tour sound and residential where separate networks and proprietary technology are more acceptable. But, for better or worse, this is where our installed systems business is headed.
Proprietary digital audio networks may be endangered.
I saw any number of proprietary digital audio networking products. Many require a physically separate cabling system. The products ranged from simple digital snakes to complete networks that carry multi-channel audio, along with control and monitoring traffic. I even saw one ring network that promises to survive a cable breakage. Yet, if I’m right, and the world is headed towards the shared IP network model, many of these proprietary networks could hit the technology dust bin in a few years.
AVB is promising but I’m not holding my breath.
AVB has a good chance to be the standard around which the technology converges. But, AVB requires an AVB-compatible, managed network switch to reserve bandwidth for an audio stream. Unfortunately, only a few vendors are now offering these for sale and important majors, like Cisco, aren’t yet included in this list. And, since standards committees typically move slowly, it may be several years before we have a final AVB standard. During that time, other technologies could have a chance to prove themselves.
Why I’m still confused.
In the past, we designed, installed and managed our own networks, i.e., analog cabling systems. But, for better or worse, those days are ending. We’re climbing on board the same train with IP telephony, video streaming, data file transfer and email. We’ll be trusting our precious audio to an IP packet system that was never designed for real-time live media. And, we’ll be negotiating for priority and bandwidth on networks that also serve security and life safety systems. It’s a brave new world where IT skills will be as important as audio and diplomacy will be as important as design.
Thus, even when the facility network has sufficient bandwidth for our needs, even when we can get a VLAN dedicated to audio and even when the IT managers are trained to manage live audio, I’m still a little uneasy about giving up the control and technical performance we’ve always enjoyed.
Why, then, are we headed down this path? I’d be surprised to hear anyone say “we want to be part of the facility IP network because it provides the best technical solution”. So, is it the politics of the construction process? Is it an economic issue? My guess is a little of both and, if I’m right, I suggest that we may be asking the wrong questions and looking at the wrong metrics. I’ll explore these issues in a future blog.