The purpose of the technology behind a LED video wall or a digital signage system is to get content (typically video and images) to the monitors or projectors that make up the system. That technology may be simple or complicated depending on how many inputs and outputs are needed and how the content needs to be presented.
The industry has been moving away from traditional Audio Video (AV) technologies and more towards Information Technology (IT). Recent editorials in Commercial Integrator (http://www.commercialintegrator.com/av/adam-forziati-av-industry-impressions/ followed by http://www.commercialintegrator.com/av/defense-adam-initial-impressions/) show this transition is well under way. The technology has become digital, and though the industry still needs AV knowhow (acoustics, sightlines, user experience), it also needs IT capabilities for networking, security, scripting, and understanding of resource usage.
Since the goal of content distribution for video walls and digital signage is to get content to the displays, I’m going to use a network analogy to explore some of the technologies. Some of us may remember the great debate between circuit-switched networking and packet-switched network. The phone companies, who were used to connect calls (thus switching circuits), were proponents of circuit switching. Circuit switching is a process in which a dedicated connection is established from the source to the destination before information can be sent. This dedicated connection offers reliable bandwidth, but has very poor resource utilization (much of the bandwidth may be unused at any given time).
The other approach, packet switching, shares bandwidth by routing chunks of information (packets) from their source to their destination along a data path that can be shared by many other packets. This approach allows for much better resource utilization and has been nearly universally adopted. As a side note, the telephone companies tried to compromise by proposing Asynchronous Transfer Mode (ATM) that was packet-based, but provided virtual circuit switching at the time a connection is established. It is used for ISDN and deep within some of the backbone networks.
How does this networking analogy apply to visual information distribution? The traditional matrix switch that switches an input to one or more output monitors is like a circuit-switched network – the path is dedicated from input to output. If that functionality is all that is needed, then such a non-shared approach is appropriate. We can argue that the traditional server-based video wall systems are also like the circuit-switched networks. They may not appear so at first glance because the inputs can be shown on any monitor or maybe several monitors, so the connection from source to destination (monitor) is not obvious. However, we need to think of the hardware server as the destination in this case. It is the component that renders the video to the displays, thus it is the constrained resource. While the server is a powerful machine, it has to accept all of the inputs and drive all of the displays so it can be overwhelmed if it has to do too many things at once.
Distributed system-based content delivery systems tend to lean towards the packet-switched network analogy. The network connecting all the sources and the many computers that drive the displays is a shared resource that is designed to support the delivering of many feeds to many destinations simultaneously. In addition, the amount of resources (computers, memory, network switch bandwidth) scales as the number of displays grows. It is entirely possible to overwhelm individual resources, but that doesn’t break the entire system. For example, one display computer could be overwhelmed by receiving and trying to decode more video feeds than its CPU can handle, but in a well-designed distributed system, only that display is affected while the rest of the system operates normally. As with packet-switched, even if one node gets bogged down, packet traffic still flows everywhere else in the network.
For example, the Hiperwall software treats all the connected displays as a huge canvas. Any of the content can be shown anywhere, from a small portion of one display to partially covering several displays to the entire video wall or signage system. All the content is digital and shared as needed on the network (capture cards can digitize video sources, for example), and thus can be distributed to any or all the computers driving the displays. Of course, putting too many 4K videos or high-bandwidth streams on a single display computer may cause that computer to stutter while playing the content, but the rest of the system is unaffected by that overload condition. Because of the distributed, shared nature of the system, it can display a huge amount of content, including multiple streamed or stored content items per display, while scaling to hundreds of displays showing live streams.
There are many products in the industry that take advantage of IT (IP networking, modern processors, etc.), but are really providing point-to-point solutions (digitize video here, display it there). These are like Asynchronous Transfer Mode, in that they provide packet-based virtual circuit switching. There isn’t anything wrong with these products if you need the equivalent of a really long HDMI cable, but they don’t scale and aren’t as flexible as systems that provide a shared canvas to draw many content items.
In short, the content display industry, including video walls and digital signage, has been transitioning over the past decade from traditional dedicated AV hardware (circuit switching) to a commodity-based IP networking IT infrastructure that can provide resource sharing, flexibility, and fewer constraints. This transition is lowering costs, enhancing capabilities, and supports unprecedented scalability, which is important in a world where displays are all around us. As displays become ubiquitous, the hardware, software, and networking technology to drive their content must scale appropriately.
