1 of 8
A Dante MY-16 interface card is one of many networking options for Yamaha digital consoles.
2 of 8
The Roland S-2416 supports 32 inputs (8 digital and 24 analog) plus 24 output (8 digital and 16 analog) and has an additional REAC port to cascade and additional Digital Snake for expanded I/O. REAC is Roland’s audio-over-ethernet protocol, 40 x 40 channel, 24-bit either 96 kHz, 48 kHz, or 44.1 kHz.
3 of 8
For more interfacing flexibility, Aviom is now offering a Dante option for its D800 network distributor.
4 of 8
The Core 500i from QSC offers multiple connectivity options for its proprietary Q-SYS networked systems, based on Gigabit Ethernet.
5 of 8
MyMix by Movek employs a modified AVB protocol that allows use of standard (non-AVB compliant) network switches
6 of 8
The Waves MGB interface connects digital consoles using MADI to the Waves SoundGrid network.
7 of 8
Based on AVB protocols, the Pivitec personal monitoring system integrates control via iOS devices.
8 of 8
Allen & Heath’s ME-U personal monitor hub features a standard input card to interface with GLD dSnake, iLive ACE or Aviom A-Net. MADI, Dante and EtherSound cards are available, as well.
That pretty much sums up the development of digital audio networks, at least thus far. The recent proliferation of non-compatible networking protocols has left us with a profusion of networked audio devices from various manufacturers—all equipped with identical RJ-45 network connectors—that cannot talk to one another. The high-resolution, multi-channel audio “spoken” by one device is “heard” by another as incomprehensible digital babble.
AES-67: Tackling the problem
This problem has not gone unrecognized by the industry and professional audio societies. The most sweeping effort so far at finding common ground is embodied in the AES-67 standard, published late last year by the Audio Engineering Society.
AES-67 is not a new networking protocol, but rather a “road map” for enabling various existing (and future) networking protocols to communicate with each other. The end goal is seamless “interoperability”—the current buzzword—so that a device using one networking protocol can plug into another and transfer the signal transparently with little or no added signal delay (latency). The fact that we have a proposed standard is a major step forward, but manufacturers still have to engineer the details into their new products while also maintaining backward compatibility with existing products.
No easy task, but networks already working to implement the standard include Audinate’s Dante and QSC’s Q-LAN. The standard also outlines scenarios for interoperability with the AVB protocol, developed by a different professional standards group with a broader mandate, the Institute of Electrical and Electronics Engineers, or IEEE. (More on all of these in a bit.)
A creeping revolution
Although the lack of a universal protocol has introduced confusion, the revolution continues as audio networking technology steadily improves and costs come down. Rather than waiting for across-the-board interoperability, churches are integrating audio networks on a piecemeal basis. Think of it as a “creeping revolution.” Networks often start with on-stage personal monitoring, next advancing to the stage box-to-console connection, then into recording systems and outboard effects systems, and finally to amplifiers, loudspeaker drive processing and now even directly into some powered loudspeakers.
The revolution is gathering momentum. If you don’t want to be left behind, you need to a.) get a grasp on audio networking basics, and b.) familiarize yourself with the current landscape so you can make solid near-term decisions.
What you need to know
If your background (like mine) was in audio and not IT, you might have some catching up to do. Understanding networking is now as critical as understanding analog systems and PCM digital audio. To start with, you need to know:
1. The basics of packet-switching networks and how they differ from point-to-point connections.
2. How Ethernet works, and the implications of the speed levels (Fast, Gigabit and beyond).
3. The TCP/IP stack, and how its layers relate to the function of networking protocols.
If you don’t have time to take an online course or read a textbook, a good introduction can be found in a pamphlet from Harman Professional: “The HiQ Net Guide to Audio Networking,” available online as a PDF. (Google “HiQ Net Guide.”) It’s a bit dated—two years is a long time in networking—and far from comprehensive, but nevertheless it presents a useful overview.
Surveying the landscape
In this brief overview, we’ll only look at specific systems intended for live production audio or, secondarily, for distributed audio around church facilities. Systems intended primarily for broadcast applications are not included here.
Group one: open-standard and third-party “backbone” networking systems
Designed to connect a broad range of devices, these networks—at least potentially—could form an integrated system connecting most or all of the audio devices in your church, even when sourced from different manufacturers. It’s a mixed bag: some were widely adopted but have lagged recently, while others are surging ahead propelled by newer networking technologies.
AES-50
Defined as an open standard by the Audio Engineering Society, AES-50 is a Layer 1 protocol that uses Ethernet wiring but is not a true-switched audio network. In this respect, it is more comparable to MADI. The SuperMAC and HyperMAC networks, based on AES-50, are used by Midas consoles and Klark Teknik signal processing for low-latency connections with up to 192 bi-directional channels at 96 kHz. AES-50 is also employed as a network by Behringer consoles.
Audio Video Bridging (AVB)
Developed and published by the IEEE, AVB is another open standard that broadly defines a suite of protocols for time-synchronized, low-latency streaming through networks. Because critical timing information is carried out at the Layer 2 level, special AVB-compliant switches are required for AVB networks. Also, since the standards are broadly defined to allow flexible implementation, a dedicated industry group—the AVnu alliance—has been formed to test devices for guaranteed interoperability. As it leverages the power of Gigabit Ethernet, AVB can offer up to 400 audio channels with no fixed ceiling on either bit depth or sampling frequency. A number of major pro audio companies have committed to developing AVB-compliant products, and several—including Biamp Systems, Meyer Sound, Lab.gruppen, and Harman Professional—have introduced AVB networkable products. The Pivitec personal monitoring system is based on AVB protocols.
CobraNet (Cirrus Logic)
The granddaddy of them all, CobraNet is a Layer 2 switched network based on Fast Ethernet. It has been very successful as a system for distribution of multiple audio channels in large installations such as stadiums and convention centers, and among rooms in a few larger churches. However, it has found limited use in production applications because channel capacity is limited if both high audio resolution and low latency are required.
Dante (Audinate)
Introduced a full decade after CobraNet, Dante leverages speed of Gigabit Ethernet to offer a switchable, low-latency network with up to 512 bi-directional channels and a maximum sampling rate of 192 kHz. Dante has also won kudos for its fast set-up, with auto-discovery of all devices on the network. Audinate cites a list of more than 150 licensees of Dante technology, a list that now includes the majority (but far from all) of the major players in pro audio systems. If you have a favorite piece of gear that can’t plug into Dante directly, chances are the Focusrite folks can accommodate your needs with their Dante-based RedNet products. RedNet devices communicate with each other via Dante, but translate audio into MADI, AES3, S/PDIF, directly into ProTools and, of course, into analog.
Ethersound (Digigram)
The Ethersound protocols define a hybrid between Layer 1 and Layer 2, with audio data switchable only in one direction. (Two-way communication is possible only when units are directly connected in a daisy chain.) Latency can be as low as 125 microseconds, and channel capacity is up to 512 per link, with a maximum sampling rate of 96 kHz at up to 24 bits. EtherSound has given way to newer protocols in recent years, but the network is still supported by several major digital console makers.
MADI (AES 10)
Short for Multichannel Audio Digital Interface, MADI is not a switched network, and it uses coaxial cable (or fiber optic line) for transmission, not Cat-5/-6 cable. Widely adopted, it is essentially the multichannel version of AES3. MADI is mentioned here because any supplier of true-switched audio networks is well advised to supply a MADI bridge to accommodate the many MADI-enabled systems currently in use.
SANE (Optocore)
A relatively new entry from the German maker of fiber optic systems, SANE (Synchronous Audio Network + Ethernet) offers synchronous and redundant transport of up to 64 channels of audio with very low latency (41.6 microseconds). Naturally SANE interfaces seamlessly with Optocore’s fiber optic-based systems, but to what extent it will catch on as a separate network remains to be seen.
RockNet (Reidel Communications)
Also out of Germany, RockNet is a dedicated Layer 1 system that essentially functions as a highly flexible digital snake. At the top end, a system can carry up to 160 channels, with up to 96 kHz sampling rate. All components and hardware are touring grade, though installations are easily accommodated.
Group two: proprietary and focused application
Although there are gray areas between, the networks following differ primarily in that they are proprietary to the maker of the audio devices to be networked. In other words, it’s likely that the network will accompany the product(s) in question as part of a “package deal,” and less likely the network protocols will be evaluated separately.
ACE (Allen & Heath)
Short for “Audio and Control over Ethernet,” the ACE link offers connection via Cat-5 cable for up to 64 bi-directional channels of audio plus network control. An ACE card can be fitted into the company’s iLive and GLD consoles for expansion and integration with other systems.
A-NET (Aviom)
The first of its kind and still a dominant player, Aviom’s A-Net is a proprietary Layer 1 protocol that is best-known for linking the company’s popular personal monitoring systems. It is also used in Aviom’s digital snake systems.
BLU LINK (Harman Professional)
Although using Cat-5e cabling, Blu link is not Ethernet-based, but rather functions as a high-speed digital bus. Devices are wired together in a ring topology (usually in the same rack room) for redundancy. The protocol supports up to 256 channels at 48 kHz/24-bit resolution.
MOVEK (myMix)
Movek’s personal monitoring and recording systems employ a network based on AVB protocols, but with modification to enable use via common Fast Ethernet hardware. This lowers costs and simplifies configuration, but allows “interoperability” only with other myMix devices.
Q-LAN (QSC)
Q-LAN is an up-to-date Layer 3 protocol similar to Dante regarding both performance and configuration flexibility. However, in this case, QSC reserves use to its own Q-Sys integrated system platform.
REAC (Roland Systems Group)
The Roland Ethernet Audio Communication (REAC) protocol links the company’s digital mixing, recording and personal monitoring systems with a 40 x 40 channel, low-latency link at 24-bit/96 kHz resolution. Standard switches can be used for signal splitting, and the same Cat-5e cable also carries MIDI and remote control data.
SOUNDGRID (Waves)
SoundGrid is a proprietary Layer 2 network principally designed for linking SoundGrid servers (for processing plug-ins) with digital mixing consoles and, for processing control, host computers.
ULTRANET (Behringer)
Ultranet is a proprietary network designed to interconnect Behringer mixing consoles and the company’s P16 personal monitoring systems. Recently, Ultranet capability has been extended to the company’s IQ Series powered loudspeakers.
No need to wait
If you study your basics and plan ahead, there’s no reason not to delve into networked audio at the earliest opportunity. Yes, a more standardized future will be welcome, but the added operating flexibility and lower installation costs will make it worth joining the networking revolution when your next system upgrade comes around.
