Saying that networking in AV is a hot topic is a bit like saying the Grand Canyon is a deep hole in the ground — no one is surprised. The forces driving the adoption of IP technology are a combination of cost savings, performance and, above all, flexibility. Those forces have changed the way we think about AV.
By "flexibility," we do not mean simply larger channel counts or longer cable runs; rather, we mean replacing the long-used model of connecting devices with physical, point-to-point cables with a system that permits endless combinations of signal flow, and with virtually no dependencies upon location or distance. It is a system in which old noise problems, such as ground loops, are nonexistent, and in which digital products from hundreds of makers all "just work," without clocking headaches or magic incantations. We're talking about an AV system that can be secured and managed to ensure reliable operation — a system that is a lot like a computer network.
The Economics Of IP
For most small- to medium-sized installations, no detailed knowledge of networking is required to create an audio network. Modern solutions are self-discovering and automatically configure clocking, making setup easy and essentially "plug and play" with simple software. Bandwidth concerns are almost nonexistent, as gigabit networks easily transport hundreds of channels of uncompressed audio without special accommodation.
In larger and more complex systems, it is beneficial to consult IT specialists. Real-time AV traffic is still a standard type of data, but networks can be optimized to better handle very large channel counts or large numbers of endpoints with careful configuration of network switches. The system-design rules for AV are very much the same as for any good computer network, meaning that one must be aware of any points at which data aggregates to avoid congestion in large setups.
It is also important to note that installing networking infrastructure is generally far easier than installing legacy AV. Ethernet cables are inexpensive, thin and lightweight, and they are easily run throughout buildings. There is no need for bulky "snakes" of analog or digital cables, and one may freely place access jacks everywhere they might be needed in the future.
With a network, systems integration is easily done. Once in place, the system can reconfigure endlessly, without getting into walls or ceilings to move cables, and it uses software from a common PC. All connections are made using one type of cable and one type of jack. By choosing to install readily available gigabit networking equipment, an installation is effectively futureproofed for years to come.
Lessons To Learn
Fortunately, the lessons that IT has learned over the years are available to help the AV community. Practical design choices in networks, such as isolation techniques, can help mitigate unwanted access, as can products that impose a user and privileges hierarchy upon access to system control.
Approaches currently being used closely mirror traditional IT solutions for security, and, over time, they will expand to include more device types and a higher degree of granularity.
AV Will Remain Unique
IP networks already comply with all the standards required for accurate, reliable transport of data that is real-time on a human scale. The Precision Time Protocol (IEEE 1588) standard has existed since 2002, and it permits networks to be time-aligned with great accuracy — 1μs or less. This, coupled with practical design choices in network topology and switch configuration, is a unique attribute of AV networking at scale.
Audio And Video
This implies that, for workflows that involve uncompressed video, 10Gb/s or higher networks are required. Although such technology can be purchased today, it is far more exotic and expensive than the ubiquitous 1Gb/s networking equipment we see everywhere. To be clear, this is not to say that video cannot run over 1Gb/s connections; indeed, with compression techniques such as JPEG 2000, high-quality post-production video can easily be sent to networked endpoint displays and monitors. Over the next few years, we can expect to see that approach become commonplace, alongside further adoption of very-high-speed networks (>10Gb/s) for production locations.
Converged Versus Segregated
Segregated systems are a good idea in some instances. For example, a very large system that carries a huge channel and device count might tax a gigabit network close to its limits, and, thus, it would be a good candidate for "standalone" operation. Segregation also aids in designing secure networks, as the audio network has no connection to other sources of data. Some installers also prefer segregated systems because such systems mean they don't have to worry about clashing with an IT department.
But, in many cases, segregation is not the most useful option. In settings such as conference spaces, corporate boardrooms, hotels and hospitality centers, data access and audio are frequently required on the same devices and in the same locations. In those instances, only a converged network will allow computers and AV devices to access necessary services and resources, making the system more coherent and easier for end users to operate.
There is no universal advantage to either approach. Instead, by using the common platform of IP networking, one can tailor solutions so as best to fit a client or situation.
10Gb Versus 1Gb
In audio, 10Gb/s switch ports are useful for "trunking" data from one switch to another in very large systems (typically >1,000 channels). In video, 10Gb/s and higher will be required to transport virtually any uncompressed video, and it will be useful for simultaneously streaming multiple sources of compressed video to many endpoints. At present, a 1Gb/s network can only handle a small number of high-quality compressed video streams.
Moving Up In The Stack
Much practical AV-network security is still done via obfuscation and limited physical access. If networks are kept segregated, then that is a logical approach. However, in larger systems that are part of broader networks — like in conference rooms and corporate spaces in office buildings, or even a large stadium — the need for a coherent, user-based method becomes clear because that is precisely how IT departments manage access.
With leading audio-network management systems today, users log in using credentials created for the network, or with common methods such as LDAP and Active Directory. Once logged in, users may only interact with the audio network according to the permissions set by the administrator. Different people might have different levels of access to different areas of the network, and devices are kept locked within defined groups, or "domains," so that new devices cannot be added without permission.
Multi-vendor-supporting solutions enable a level of security that would otherwise be very difficult to achieve in a mixed-manufacturer system.
Scaling The Network
To scale beyond that level, there must be time alignment between subnets. Specifically designed audio-network platforms can automatically configure clocks within subnets and link them to maintain the same levels of performance as before, but now at a much larger potential size. AV networks can scale to thousands of devices and continue to be managed from a single point of control.
For systems without comprehensive audio-network management, single networks are limited to a few hundred devices each. If those systems do not have to communicate with one another, multiple "standalone" AV networks can be deployed to achieve larger scale using virtual LANs (VLANs) to contain each one. Although that does limit some functionality, it is often a practical solution.
The benefits of networked AV remain as promising as ever: easier deployment, near-infinite flexibility and flawless performance. As the industry grows, those promises are being delivered on a larger scale, and they're changing the way audio and video are getting done.