Workflow Next Generation Solutions for Broadcast Workflows Matt Adams and Chris Lee, Omneon Inc.

One by one, products with broadcast-specific functionality are dying or becoming terminally ill. Videotape is giving way to disk drives. High-end graphics systems are giving way to Photoshop and Flash. What will be next?

The death (or illness) of many broadcast-specific workflows which will give way to more IT-like processes. Dry as it sounds, the shift of the broadcast and media industry toward IT has the potential to alter cost structures so radically as to rejuvenate the business of delivering scheduled television channels.

If an industrial engineer was introduced to a television station for the first time, one of his or her first questions would have to be, “Why do you ‘touch’ the product so often?” It’s a natural question, considering the typical workflow. One person must find the program or determine when it can be recorded; someone else records it; and a third person might preview the program and jot down segment timings. A copy is made so that someone can make a promo; another staff member reinserts the program into the on-air playback system; someone or something then starts playback. The inefficiencies of this model are even more striking if all of these processes are multiplied by the number of stations that typically receive, handle, and broadcast the same syndicated programming.

Real-world solutions have been developed and deployed to attack elements of this problem. Pathfire, for example, distributes dozens of television programs daily via file transfers, so no one has to roll a tape on the receiving end, and, in theory (at least), important metadata such as timing details is preserved. In practice, however, the majority of stations still dub the program to tape (in real time) before airing it. In some cases, they even dub it to a different server first.

Next-generation solutions addressing this issue will target a different customer: single organizations that program hundreds of channels. Such businesses already exist in much of the world where broadcasters traditionally have “farmed out” on-air playback to a service bureau. Consider what might happen when an organization with a huge content library (such as a Hollywood studio) decides it would like to run a dozen different channels in each of 20 different regions worldwide. To keep things interesting, also assume that the design goal is that once a piece of content is received, no one should ever touch it again. To the industrial engineer, this would represent a perfectly reasonable goal.

The primary pitfall to avoid in implementing such a model is working with proprietary and non-standard tools. Of course, there always will be some compromises when it comes to standards, but the broadcast industry and its vendors have settled on a handful of video and wrapper formats, and it’s a given that any large-scale playout system will be required to produce material in virtually any of these formats. The model simply requires that much flexibility.

38

Beyond those video-specific standards are additional standards created by the IT industry. Broadcasters and other media companies now can ride the same technology wave that the Internet rides. Together, Service-Oriented Architecture (SOA) and Extensible Markup Language (XML) provide a good framework for developing a variety of specific, individual functions and presenting them in a unified manner to the end user. In more concrete terms, it means that the Hollywood studio might develop a Web page and database through which its employees can create and modify content records. Instead of developing more complicated features itself, such as transcoding a video from one format to another, the studio could design the Web page to trigger that functionality from within a module purchased from a vendor.

Omneon has committed to this kind of approach in its Media Application Server (MAS). The MAS eliminates the need to write API calls to specific devices or to manage specific storage locations. This solution provides an SOA to locate (and alter) media and manage files, and similar approaches are being developed by other vendors.

How might a Hollywood studio put MAS to use in programming its 240 worldwide channels? For starters, the studio would create a centralized ingest and quality control facility. If the content is received as a video feed or videotape, it is encoded to a specific folder. Content that is received as a file might also be moved to this location. A media worker at the intake facility could open a Web page displaying the material that the bookings department has purchased and is expecting to arrive. This person could also open the folder containing the new material and link the actual content files with the information about their arrival. When this link is made, work rules could automatically subject the content files to an automatic QC process. All of this material might also be transcoded to the same “house format” and/or a file wrapper applied.

At this point, items would be placed in a viewing queue to be approved by media staff, who may manually log segment-timing information. Once this information is logged, each file would be stored in a super-secure, high-availability storage system such as the Omneon MediaGrid™ active storage system. In the meantime, schedulers would have been hard at work defining each channel in terms of language, time zone, aspect ratio, etc. From the universe of programs they know are available, they’d also schedule programs and commercials for each channel.

Of course, not all outputs are individual channels. Some might be groups of channels already groomed as cable TV transport streams. Programs could also be designated to be sent to mobile devices in certain territories for viewing in a certain date window, to the Internet for streaming, or to cable VOD or IPTV systems. In many of these systems, different commercials might be requested depending on the area served by the distributor.

AUTOMATION FOCUS