The digital revolution in audio happened slowly. Digital devices replaced analog gear (single devices that performed one function); however, the digital devices remained single devices that performed one process, just in the digital domain. Years passed before industry professionals began to make single devices that could perform multiple digital processes. Networks of digital devices soon emerged, and that’s where we are now.
A Networked Media System could easily be mistaken for any group of digital devices that are connected to each other by Cat-5. That’s not what a Networked Media System really is; that’s a digital network of devices. A Networked Media System is more than just connecting digital devices via network cables and network switches. It’s an actual system of media devices. With Networked Media Systems, the benefits are based on using the network itself as one of the tools, not just as a means to connect. The network lends the ability to be more efficient, and the ability to share resources. Those resources are audio channels themselves, and also the DSP capacity, which can be shared among devices or with a highly evolved system like Tesira. You’re actually sharing the processing power.
Until recently, each device had to be programmed individually. You’d program Device A to perform a specific function, address it with the computer, and give it its instructions. Device B would receive a different set of specific instructions, and so on. With Tesira, you design a system file. As a system designer, you can choose to never involve yourself with which processing blocks go in which of our devices. You can simply compile the device and plug it in. The device will choose the most efficient set of hardware in our configuration engine that will support that design, and will decide where the blocks go (some elements, such as input blocks and output blocks, have to be in certain devices). Processing blocks can be shared among many DSP devices if you have a design that is using many DSP devices. There is no need to program individual devices; you’re using the network to spread the resources for efficiency and balance. That’s one of the examples in which the network itself becomes a tool that improves the system.
Making the network one of the tools presents tremendous advantages. There are different architectures that can be used, and that’s where we’ll begin discussing centralized Networked Media Systems, which are one of the four types of Networked Media Systems.
The idea of a centralized Networked Media System involves a single point of processing that is really powerful. This processor is typically driving a large system, and the endpoints are relatively simple. The endpoints may actually be somewhat sophisticated, but what they’re really doing in terms of processing is next to nothing. They’re not manipulating ones and zeros of audio. These endpoints (an in or an out or both) allow analog audio to enter the system through the analog domain, and the endpoint devices communicate back to the central hub. This communication occurs over switch networks using AVB switches with Ethernet cable (Cat 5 cable, Cat 5e, Cat 6). Many people are uncomfortable with the idea of a single point of processing, and that’s why our Tesira SERVER, which is the best tool for that job, can be deployed in a redundant pair.
Tesira SERVER offers configurable DSP slots. With some of the devices on the market, the DSP included in the main processing chassis is the only DSP provided; if you need more DSP power, you have to get another chassis. Both our Tesira SERVER I/O and our Tesira SERVER have additional slots to accommodate more DSP cards if needed. Tesira SERVER can support up to a total of eight DSP-2 cards, and each one of those cards does as much as two of our AudiaFLEX processors in terms of capacity, so that’s a significant amount of processing power crammed into one chassis. With just a single Tesira SERVER, you can operate a large, sophisticated system with significant processing and many endpoints.
You can also perform these functions in a redundant pair, without the need to create two servers with two identical files. To do this, you create one file and tag it as “redundant” in the software. The servers identify each other and one is designated as the secondary. The primary is the one that actually runs the audio, and the secondary SERVER observes the primary and follows any live changes to presets, changes to levels, and all of the adaptive processing is happening with our adaptive algorithms and adaptive filters. The secondary is following everything that happens live in the box, and is performing identical functions, just without audio running through it. If the primary is lost for any reason – a network problem, a cable that goes bad, a faulty power cable – the secondary takes over in a matter seconds. It will find all the expanders that are attached, and begin driving the system in the state in which the primary had been operating. There’s no human intervention involved in making sure that system is still running.
Depending on what disrupted the primary – a momentary power outage or circuit breaker issue, for example – it may restore itself. In that case, the original primary becomes the new secondary and will continue to operate in that capacity when it boots back up. It will register that the secondary is now in active state, so it doesn’t try to take over the system. Instead, it enters background secondary mode even though it’s still technically designated as the primary. It will remain there and wait. If, for the rest of life that system, the secondary never has a problem, the secondary will continue to run the system.
In the extraordinarily unlikely event of equipment failure with the device, or if a network cable went bad and human intervention is required to get the primary back online, you can choose to manually fail it back to the primary if you wish to have the original primary remain the primary. This ability is also useful for testing. The primary/secondary moniker doesn’t really matter after the initial boot up. Picture it like a driver’s training car that has brakes and a steering wheel on the passenger side so the instructor can take over if you’re about to jump a curb.
One of the key points is how little additional work it takes to have a redundant system. Only a couple of mouse clicks are needed to indicate that a file is redundant. Then you just put in four more screws to add another server to the rack. It’s so little work that the rack screws are part of the list of extra steps. With some systems, implementing redundancy can be so time consuming that some people choose not to do so. We wanted to create a low barrier to entry to make sure systems will be robust.
A centralized system has a number of advantages, including a single point of control and access to the system. You have one place from which the entire system is driven, and the bulk of the financial investment is in that one place. Endpoints are much less expensive, and the necessary cabling is all probably in the building already. If the building is expanding or if it’s a phased installation, you can simply continue deploying endpoints on an as-needed basis. The additional investment is minimal because you have that core set of servers that are driving all the processing. You don’t have to reinvest; you just keep going. The percentage to add endpoints decreases, so expanding or changing the system to move endpoints between rooms is a low burden because the bulk of the investment remains in one fixed place. It will always be capable of driving the system no matter how much the system may change overall, because that one central hub can be flexible. You can change the file or the design easily to adapt to changing needs, and reload a new configuration that will operate differently based on the endpoints.
With a low burden for modularity, centralized Networked Media Systems offer flexibility, scalability, the ability to launch in phases, and ability to support expansion. Centralization provides a high total ROI. The power of a centralized Networked Media System with a pair of redundant Tesira SERVERs, and especially the ability to grow over time and add DSP cards and expanders, provides a powerful brain for an amazing system. Ultimately, centralized Networked Media Systems are versatile, cost-effective, and well suited for a broad variety of applications.