State of Address:
Data Networking in Pro Audio
by Stephen Murphy
Data networking, a nearly three-decades old office-oriented technology, has officially arrived in the wide world of professional audio. The use of a variety of existing data-sharing and archival schemes in the audio industry has been building steadily over the last decade. This trend has been propelled in recent years by significant technological advances, manufacturer alliances, the adoption of standards and good old pioneering ingenuity.
Several new companies have risen to become near-ubiquitous in the audio networking industry; and many of the old favorites have added network compatibility to exiting product lines or have introduced entirely new divisions to take advantage of the new paradigm. While this is old news to some, many other audio systems designers, engineers and facility owners find themselves wondering what the fuss is all about. This article offers an overview into two of the biggest trends in networked audio: Ethernet-based systems and storage area networks (SAN).
The most popular standard of networking in use today, Ethernet, is actually the oldest. Ethernet even predates personal computers, having been developed around 1973 by Xerox engineer Robert Metcalfe, who later founded 3Com.
In an oft-repeated quote, Metcalfe exemplifies the era: “I came to work one day at MIT and the computer had been stolen, so I called DEC to break the news to them that this $30,000 computer that they'd lent me was gone. They thought this was the greatest thing that ever happened, because it turns out that I had in my possession the first computer small enough to be stolen!”
In keeping with the manifest-destiny atmosphere at the time, Metcalfe was asked to build the first data networking system to facilitate the inter-office use of Xerox’s latest prototype: the laser printer. His challenge was to enable over 100 computers within one office to, in an orderly fashion, use the same device,
The key to his design was in assigning unique addresses (taking a nod from the recently developed Network Control Protocol, precursor to what became the Internet) to each “node” in the system, thereby allowing multiple disparate bursts of data packets to be routed in an orderly fashion. Metcalfe and his research team’s Ethernet specification defined both the physical connections and media and the packets by which information is transmitted over a local area network (LAN). Their patent describes Ethernet as a "multipoint data communication system with collision detection".
This original Ethernet protocol, also known as 10BASE-T, transmits information between computers at a rate of 10 Mbps (10 megabits per second). Improvements in data transmission speed have spawned new Ethernet protocols including Fast Ethernet or 100BASE-T, supporting transfer rates of 100 Mbps, and Gigabit Ethernet or 1000BASE-T, supporting data transfer rates of 1000 Mbps (1 gigabit).
Ethernet installations use inexpensive interconnecting cable similar to install-grade telephone cable called Category 5 (CAT-5). CAT-5 is an unshielded twisted pair (UTP) data grade cable which supports lengths up to approximately 100 meters before electromagnetic interference (EMI), signal radiation and attenuation problems threaten data integrity.
Ethernet is the most widely implemented networking standard in use today, connecting an estimated 50 million “nodes” in LANs worldwide. As can be expected with technology of this vintage and ubiquity, Ethernet components are generally inexpensive and readily available, making it attractive to developers for use in a number of alternate applications, including pro audio.
The adoption of Ethernet into the contracting, install and other segments of professional audio has been fast and furious — some say revolutionary. Protocols developed by Peak Audio (CobraNet) and Digigram (see Digigram EtherSound First Look, page XXX) are actively being licensed to many original equipment manufacturers (OEM) for incorporation of their technology in a surprisingly wide variety of pro audio products. Other manufacturers making proprietary Ethernet-based systems include Yamaha (mLAN protocol), Peavey (MediaMatrix CobraNet-based system), Otari (ND-20/mLAN compatible) and Gibson Labs (MaGIC protocol/system). A sure sign Ethernet-based systems had arrived in pro audio was when Neutrik introduced its innovative EtherCon series of professional CAT-5 connectors.
Use of a data network based on Ethernet offers several powerful advantages over traditional multi-zone analog wiring schemes. Ethernet’s inherent data handling capabilities, vaguely like a bi-directional multi-lane highway where every car has somewhere to go and knows how to get there (unlike driving in DC), makes it a natural for carrying and routing signals in a complex multi-zone audio installation.
Traditional analog wiring of a facility requires a large amount of cable and conduit, and installation is often disruptive or destructive. Ethernet-based systems use inexpensive CAT-5 cable (or fiber optic cable for longer runs). Also on the favor of networked audio is the elimination of discrete wiring “home runs” to (and from for bi-directional communication) each zone to the source master. In typical Ethernet systems, there can be multiple master or source nodes contributing audio/control data to the rest of the network. Downtime, repair efforts and cost are also reduced commensurate to the reduction of cable and connections in use in networked audio as compared to analog systems.
CobraNet, introduced by Peak Audio (now a Cirrus Logic company) several years ago, is by far the most conspicuous use of Ethernet in pro audio to date. Peak has established CobraNet as a standard for the transport of multi-channel audio and control data over a Fast Ethernet LAN. By aggressively licensing its technology, Peak Audio has created a audio and data communications protocol that allows the connection of products and systems from many different manufacturers which can operate together or independently on the same CobraNet LAN.
Peak Audio’s website lists as development partners twenty-four audio manufacturers, all of whom are developing or have released products designed to operate on a CobraNet LAN. Companies producing CobraNet products include: QSC, Peavey, Crest Audio, Rane, Whirlwind, Crown, Shure, BSS, Yamaha, Digigram, Symetrix, Gentner, Golden Sound, Linker, Renkus-Heinz, Richmond Sound Design, LCS, Ivie Technologies, Clear-Com, Creative Audio, Bose, Bi-amp, EAW and Mackie.
Peak Audio’s website (www.peakaudio.com) provides links to all of the above manufacturers for more information on their range of CobraNet products.
SAN Gets Everywhere
One of the other major movements towards networked data in pro audio is the proliferation in production studios of the storage area network or SAN. The rate of adoption among studios of bandwidth-intensive technologies — high resolution audio, multichannel surround audio, high-definition digital video and other “rich media” formats — is outpacing the ability for traditional storage solutions to keep up with the increased load. Enter the SAN.
A SAN distributes its resources over a scalable, redundant network of storage devices and servers, allowing many users to simultaneously access the same media without delay or reduction in quality of service (QoS). Devices attached to the SAN can be managed centrally, and thanks to their distributed form of architecture, data can be backed up with very little impact on network throughput. The SAN concept is based on a high speed interconnection technology called Fibre Channel, which allows transfer rates of up to one gigabit per second (far faster than the current SCSI maximum of 160 Mbps) and allows optical cable runs of up to 10 kilometers.
Several new start up and existing storage-oriented companies have focused their sights on the burgeoning market of rich media production, tailoring Fibre Channel SAN systems specifically for the audio/visual markets. Companies such as Rorke Data, StorageTek, PESA, Studio Network Solutions, Glyph and others offer a variety of scalable hardware systems and custom systems designs tailored to suit the facilities need.
Some companies are producing self-contained SAN solutions that install in minutes and possess the familiar studio control room just-another-rackmounted-piece-of-gear look, removing much of the confusion and work involved in “going SAN.” Glyph, for instance, popularly known for their long association with Digidesign Pro Tools audio and Avid video workstations, have added an array of self-contained SAN solutions and networked products including the Coba NetMedia Manager.
Perhaps Studio Network Solutions (SNS) has made the biggest splash in the pro audio market, going from a regional storage solutions provider to a full-scale Fibre Channel SAN hardware and systems design provider recognized throughout the industry in an incredibly short period of time. The company, founded in 1998, gathered a large amount of attention at the last several AES audio conventions with impressive demonstrations of their A/V SAN and A/V SAN PRO rack-mounted Fibre Channel products.
A/V SAN is a single rack unit/single user SAN that can be fitted with up to four drives and is designed to provide an easy-to-implement Fibre Channel storage solution at entry-oriented prices. A/V SAN’s bigger brother, the three rack unit A/V SAN PRO is completely scalable in storage size (up to 15 drives/enclosure) and can accommodate to up to 20 simultaneous users.
The trend towards the use of networked technologies in pro audio is ever increasing, seemingly to the benefit of all working in the industry. Increased performance, better reliability, more user choices and easier installs are just a few of the gains to be realized by the proliferation of network-based audio systems.
While the subject of networking and storage in pro audio could fill a massive book — a book that needs revising every other month as the frontier evolves — an article such as this can only cover a small portion. Look for in-depth coverage in Pro Audio Review of the above technologies in upcoming product reviews and features.
Stephen Murphy, contributing studio editor for PAR, has recorded over 300 vinyl and CD releases, including a Grammy Award-winning and a Platinum-selling album. Steve can be reached at firstname.lastname@example.org.