Modern master control rooms, like the one shown here in Turner Broadcasting’s new HD facility in Atlanta, have different paradigms of monitoring and control.
Modern needs, modern problems
Over time, master control has changed, but its basic function remains the same. Broadcasters around the world use master control systems to build program streams and insert interstitials such as advertising and promotions for future programs. This is not to say that the changes in master control have been minor or superficial. On the contrary, the changes have been dramatic, driven by factors such as FCC-mandated ancillary services, the requirements of the marketplace, the details of the content and interstitials and a broadcaster's desire to differentiate his output from that of the competition.
The FCC has mandated closed captioning, descriptive-video-service (DVS) audio, emergency alerts, ratings flags (V-chip) and, of course, DTV. Beginning early in 2005, broadcasters must add PSIP to their signals. Taken individually, these requirements are quite simple to accomplish. But they add cost onto the broadcaster's ledger that is unlikely to be offset easily by new revenue.
The broadcaster's quest for a unique look has led to a much more complex menu of switcher options. Among these enhancements are pushbacks, more and shorter interstitials, weather and other announcements over programming, complex voice-overs, rich graphics — sometimes over content (for hire), burned-in and moving logos, and network logo insertion at the station. To this, add HDTV content and several program streams multiplexed together, each of which might have all of the above elements.
To maintain the old functions and accomplish new ones, the tools of master control have had to evolve. In the early days of color TV, a master control switcher might have had 12 to 16 inputs, perhaps one or two key inputs, and a voice-over channel. It would likely have offered automatic transitions and a set of triggers for pre-rolling VTRs (or film chains). A preview output would have fed a separate monitor to allow the master control operator to look at network feeds on a larger monitor for quality control prior to air.
But such a limited master control could not adapt and expand to accommodate the changing demands of the FCC, station management and sales. Some of the new functions — such as V-chip, pushbacks, EAS, and even DVS and closed captioning — could be (and often are still) performed downstream. However, the explosion in the number of ancillary signals added to the output begs the issue of control, certainly of human control. Some functions, like closed captioning and V-chip, need only monitoring under normal circumstances. Other functions related to transitions in aural or visual program content (such as added lower-third graphics, pushbacks, voice-overs) add greatly to the number of things the operator needs to control and monitor. As a result, automation — first in the form of cart playback of interstitials, and later in servers — became a way to extend the amount of content an operator could handle reliably. The cart machine received a log from traffic and played entire interstitial pods as a unit, greatly easing the load on the operator and allowing more complex content.
We can credit (or blame) the rise of computers for enabling the next level of complexity in master control. If the growing demands on the master control operator had pushed him to his limit, might it not be possible for a smart machine to assume some of the burden? Manufacturers met this challenge by producing new master control hardware with rich feature sets that permit creative traffic, program and sales departments to squeeze more content and dollars out of a station's sole consumer product. Today's hardware offers rich graphics, layered signals and effective control of complex events.
Signal monitoring can now be done on plug-in modules in DA trays, as with Thomson Grass Valley’s Maestro.
To make all these functions work, the operator needs an easy to use interface. (The links between traffic and automation facilitate much of the manual selection in the background.) The key component of this interface is software that can configure complex displays understandable to the human operator. The GUI must include three elements that enable the operator to understand complex events and manage them when they go wrong.
First, the GUI must have a setup task, which a facility might use only when it installs the master control or when it adds or removes peripherals. Many manufacturers now include in their peripherals embedded features that stations can activate later through software authorization keys. In addition, the setup functions offer a set of graphics and operation templates that allow the station to create and manage a unique look.
Second, the GUI must provide understandable data about the current operational state of the hardware and embedded software. This allows the operator to see that the next event has been sent to the hardware for execution (hopefully), and displays it on the preview system. Taken together with automation, this GUI element must clearly indicate critical information and facilitate manual operations that are part of normal practice. It also must permit multichannel monitoring for situations like multiplexed SD/HD in DTV. A hardware control panel is often part of this element. Increasingly, touch screens or mouse-driven computer displays are supplementing or supplanting the hardware panel we usually associate with master control. At the very least, it is becoming desirable to have a touch screen as a backup to the hardware interface. The best ones look exactly like the hardware panel so the operator understands them instantly.
Lastly, the modern GUI must provide fault reporting and resolution. For a multichannel operation, this may be the most important piece of software. It is not enough to have the monitor go dark and leave the operator wondering if it was the video/audio processor that failed. The system must alert the user to the highest degree possible. Consider the case of an operator working in a duopoly. He would have two NTSC channels (at least for the next few years), and two DTV channels, each of which might have up to four or five streams present. At any one moment, the operator is looking at four to 12 streams and their associated hardware chains. Concise and useful data will make the difference between lost events or lost hours.
Control surfaces like Eyeheight’s playout module have evolved to simpler metaphors.
Master control systems have evolved from stand-alone devices to hardware that can integrate and centralize functions like graphics creation, CG, still stores, logo inserters and DVEs. Yet the industry seems to be moving to view crosspoints and processors independently. (The opposite viewpoint is covered later in this article.) We might consider the switching requirement as an unnecessary overlap with routing, which is almost certain to exist in the facility. Thus, the processor (audio and video) might have Ethernet connections for moving graphics, receiving live data streams for processing as lower-third character generation, and other control ports, including automation. It might also have ports to feed requests to servers and other external devices. But the number of inputs may be reduced to half a dozen, some of which are key and fill inputs. By doing this, several manufacturers have been able to reduce the size of the processor to a card in a DA tray (though not a small one). Other manufacturers have made boards that mount in a routing switcher frame, using the same control system to facilitate broader control issues and crosspoint selection in a seamless fashion.
This tight integration, combined with separation of function, provides for interesting possibilities in fault management. It is not hard to envision a system that takes note of a failed master control processor and substitutes a backup in a 1-for-N (or N-for-M) redundancy strategy. The operator might be asked if he wants to move to a bypass path or simply replace a failed board with one from a pool of spares. By reducing the redundant switching functions and using an existing routing switcher or slot, the manufacturer saves on costly design, documentation and duplicative manufacturing of power supplies, frames and connectors. Cost comes down and reliability goes up, which is just the trade-off you need in a complex master control environment.
Even when the processor is not card-based in an existing or multi-use frame, the economies achieved by not duplicating crosspoints are clear. This approach enhances and simplifies the master control switcher. Modern system designs seldom have sources uniquely fed to master control switchers. Most have at least a few routing switcher outputs dedicated to master control inputs. The strategy of building the processors without internal crosspoints seems to be an acknowledgement of this predominant design strategy and a way for manufacturers to guarantee the sale of a router.
Integrating graphics into the master control processor is also a powerful and fundamental change in architecture. Take promotions, for instance. The ability to generate the graphics for backgrounds or fill in text fields in a template derived from live or other sources can simplify the creation of a playlist for a complex channel. For instance, traffic might only call a template when inserting a weather crawl; an outside service provides data to fill in the text. The same method can apply to EAS. A keyboard is often available for simple text generation in the processor. You might even load a set of templates that would allow a reduced feature set for a backup to a production control room. But, clearly, these devices are not intended to replace complex character generators.
Some master controls, like Leitch’s DTP, operate entirely in the compressed domain.
There's a second way to look at the dynamic of moving crosspoints out of the master control switcher. What if, instead of moving the crosspoints out, you moved master control functionality into the station's central routing switcher? The connector count goes down further, the integration with facilities management software might be enhanced, and the manufacturing cost should drop even further. It would also simplify system design, though at the expense of the ability to monitor paths inside the routing switcher that are no longer accessible.
In part, this is a generalized change in the way we look at routing switchers. Because all of the signals are present in the facility's hub router, you might ask: Why move them outside? At least two manufacturers have taken slightly different but parallel strategies. One is developing specialized cards to put in a frame that also holds routing. The other is adding daughter cards to a router that offers features like keyers and voice-over capability.
Whither goest master control?
This last approach plays well in the multichannel DTV environment. Many stations, particularly those in public broadcasting, see a need for simplified master control streams with fewer options at lower cost. Putting this functionality into the router allows it to route key and fill signals from any source seamlessly to any processing card for output as a simplified master control stream. This might also work well where a broadcaster uses his DTV bandwidth to provide functionality as a cable headend might, while also providing local branding at low cost. There is always a downside and, in this case, you might have to add a character generator and manage the content to allow a single generator to function on a time-shared basis with many outputs.
Lastly, you might consider having no master control switcher, but rather a processor that operates entirely in the compressed domain. Systems like this have been available for several years. They allow you to link together and brand the MPEG server output (and, perhaps, a network signal delivered as MPEG). In addition, this type of processor allows you to add many normal master control functions without going back to baseband for processing and switching. The first wide-scale deployment of such a system in the United States is the FOX Network HDTV distribution. That system uses a number of devices that can combine local and network signals into a single output (which, after all, is the fundamental function of master control).
Several things are clear. The pressure of cost and flexibility requires new tools and operational techniques. Everyone who plans or manages a facility needs to understand the dynamics at play and the range of options available. Life will never be the same again. And there is much to learn.
John Luff is senior vice president of business development at AZCAR.