For the next four months this column will focus on four interrelated and interdependent areas of television technology: automation software, data archive systems, newsroom automation systems and media asset management systems. Each one of these topics is represented by manufacturers with specific expertise, but all of these areas of technology are dependent on other classes of products to automate portions of the operations at a television broadcast plant. This month the topic covers the first of the four, which is sometimes misunderstood as a monolithic single-product industry. To the contrary, the spectrum of offerings is quite broad and leaves plenty of room for manufacturers to define a niche and fill it well.
In the most general sense, automation is used to simplify the recording and playback of television programming for on-air operations. Software products are thus needed for a range of operations differing in scale and complexity. At one end of the spectrum are products suited to large operations, like DirecTV or EchoStar in North America, or BSkyB in Europe. These DBS operations handle the ingest and playback of hundreds of channels in one facility. The needs of the air log are unique to the type of continuity string they manage, which often has fewer unique events per stream than a large-market television operation. Also, the portion of daily operations that is conducted live is more restricted. To be suitable for such an operation, an automation system would need to be highly scalable and provide redundant databases and hardware to insure the system is quite tolerant of potential faults.
The monitoring of such a complex system by a small number of operators requires a different type of operator interface, one that can easily highlight potential faults and missing media and bring the details to the foreground of the operator's attention. One might, for instance, have a simple monitor screen with many channels displayed with a minimum of information about each. When a fault occurs the channel requiring attention could be brought to full screen with all the detail about the status and potential current fault, or abnormal condition sensed for later in the air log. Some manufacturers have designed “heads up displays” that put essential information over the output video in a bid to simplify the visual field of the operators.
A full-service television broadcaster used to worry about only one outgoing program. In the last decade many local stations have taken on regional feeds for cable channels (usually news, but not always), complicating the single-stream operation. While such situations are abnormal, it is no longer unusual to have duopolies or LMAs operating in one building with multiple outputs. A two-stream operation is considerably easier to monitor and control, with the simplest case being two logs on two separate screens, or perhaps two logs on one screen. It is quite normal for a single operator to man the control point during non-live programming, though live sports may well require more attention than one operator can give to two outputs.
At the far end of the scale are systems simply intended to manage spot inventory and playback. The system does not handle program content, and often is sold bundled with a video server. In small markets where labor cost is low this may well suffice to increase the reliability of spot playback over aging video cart machines, which are not well supported today.
Then there is recording of media, which is assumed, but may not be focused on during initial evaluations of products or product families. Ingesting spots to a server is not trivial by any means. There are serious questions about how to set up the database so that the traffic system air log can be easily tied to the database of spots. The system must provide a method of handling kill dates and purging media that is no longer needed. Some of that information can be gathered from the traffic system, but caution is always advisable. If a spot will be run for only four weeks in the current buy, that doesn't mean it will never again be used. Human intelligence may provide better guidance in some cases.
There is a broader question on the input side of a system, though. Assuming program content is to be automated, one must get complete detailed and accurate information into the automation system database about each program and each segment. The metadata needed certainly includes items like the program series, episode name and number, and start and end of message for each segment. The traffic log may have approximate timings delivered by the network or syndicator, but the precise times could be many seconds off. If the content must be acquired by satellite it is appropriate to have the automation system control the dishes and tune the receivers. Unfortunately, not all systems have satellite recording modules as standard equipment or even as options. Some vendors offer satellite packages that are closely coupled external modules purchased from a third party. While not necessarily a problem, you should be cautious when asking about support for such packages.
What about the software and hardware itself? In this age where labor is cut short it seems prudent to look hard at the options for redundant hardware. This might mean backup computers running in parallel as the ultimate in protection. Some vendors utilize stand-alone device control engines for connecting to VTRs, servers and switchers. Connecting multiple control engines to a single source requires effective switching of remote control lines without introducing a new failure mode. In addition to control hardware and automation engine redundancy, it is wise to look carefully at a strategy for redundant databases of automation information. A name brand computer may make you more comfortable, but most vendors utilize industrial computers, or house brand preferences without allowing the user to choose the brand. This permits the vendor to offer simplified support for fewer devices with a staff less skilled in general support of complex computers, but leaves the customer at the mercy of the vendor's support department. It is wise to ask loads of questions about support, for it is one of the defining differences among many vendor choices.
In general, most automation vendors support all popular products for playback, record, switching and ancillary functions like EAS and other requirements. It is wise to carefully go over the list of devices you anticipate with a vendor to be sure they have interfaced to your specific choices in the past. If they have not, get a specific quote on the custom work, in writing.
Economics rule the day in many operations, and indeed the motives behind instituting automation often begin and end with money. Believe it or not, even with a single-stream station the costs can vary by a factor of 25. Systems for simple operations can be purchased for under $10,000, or over a quarter of a million. The system may be a one-time purchase with annual maintenance as an option, or the client may only lease the software and continue to pay forever. Choose the package carefully. Cheap does not always win, but neither does buying the most expensive system always get you the protection you need.
The options are many, but when coupled to other pieces of the puzzle the options in some cases become narrower. In the next installment I will discuss archive systems, a topic that becomes more important as the industry swings away from linear tape and into the tapeless world.
John Luff is vice president of business development for AZCAR.
Send questions and comments to:firstname.lastname@example.org