Four ways to manage digital audio files in your broadcast environment.
The increasing affordability of LCD and plasma panels is encouraging television viewers throughout the world to replace CRT-based receivers with larger displays. The quality of digitally delivered 525- and 625-line SD when viewed on a 40in monitor is extraordinarily good and will continue to improve as the industry moves forward with HD.
Larger screens inevitably encourage viewers to invest in surround-sound equipment, bringing the cinema experience direct to their homes. This in turn requires broadcasters to deliver multiple signal channels, transmitting Dolby 5.1 surround as a digital stream. The increasing globalization of program delivery has added to the requirement for multichannel audio by expanding demand for multilanguage soundtracks. Through case studies, this article will look at the various ways in which audio is handled in modern file-based broadcast environments.
Centralized audio storage
A digital server system forms the central audio storage resource at the Broadcast Center in White City, London, one of Europe's most modern playout facilities. Designed specifically for broadcast and post-production applications, it captures and files programs using an integral SQL database.
Audio files can be located quickly and used in playlists or played directly to air. Applications include radio announcements, program store-and-forward, automated voice-over playout and multichannel audio playout. A record screen enables announcers or voice-over artists to make their own recordings. Audio files required for playout can be recorded automatically from DAT or CD using the ingest screen.
Multiple tracks on one tape can be recorded in a batch, speeding up the process. Broadcast wave (BWAV) files can also be copied directly on to the audio server and then logged in the database. Files can be copied from server to server automatically or manually, allowing configuration for full one-to-one redundancy.
The server brings major advantages to production and playout operations, and supersedes traditional audiotape, cartridge-based and MiniDisc play-in devices. Recordings can be made to the server without having to load, label and archive individual tapes, cartridges and discs. Audio content held on the server can be accessed rapidly using its integral database, giving all the benefits of a massive library without the risk of items being lost in transit from shelves of individually archived storage media. The system also provides levels of redundancy not normally associated with traditional magnetic media.
Stored content is accessible from the continuity booths and automation-driven playout suites throughout the facility. The system's primary function is to store speech clips for end-credit and other program-junction announcements. As channels evolve their output content, the system can also be used to store other audio-based objects, such as music or effects. Each automated playout channel has access to at least one replay port from the server cluster. The continuity booths are equipped for record and replay.
The audio server is fully integrated into OmniBus automation systems chosen for playout operations and normally plays prescheduled clips to air without manual intervention. It can also be accessed under direct control by playout directors, retaining at all times their freedom to initiate an immediate manual audio playout for highly reactive channels, such as BBC One and BBC World.
A major advantage of the production workflow is its ability to populate the database directly from individual channel planning schedules. Very little metadata needs to be keyed as part of the authoring process. Recording productivity is therefore high, and expensive vocal talent costs are minimal.
UK commercial network ITV uses a server-based digital audio system as the core of its voice-over recording facilities at the Southern Transmission Centre on London's South Bank. This resource enables voice-over presenters to create audio interstitials for the ITV1 network and regions, as well as for ITV2 and ITV3, which are all played out under the control of the transmission automation system. The installation consists of three audio servers integrated to a database-driven media manager.
ITV's first digital audio server, installed in 2002, replaced a PC-based audio recorder that served ITV1 and ITV2. The three-server system has a total of 12 audio ports that are routed to the individual regional destinations served by the transmission facility. In addition, files destined for the northern regions are automatically sent by FTP to the ITV Northern Transmission Centre in Leeds for playout. Similarly, voice-overs recorded for HTV Wales in Cardiff, on a standalone digital audio server, are sent automatically in FTP format to London for playout.
The three servers are mirrored, so each of the three voice-over booths is able to record items for any region. The items are then automatically copied between servers.
Operators in each booth have access to a make-list and fingertip control of everything needed for fast recording and audition. Each voice-over is recorded in stereo as a 48kHz 16-bit BWAV file, including metadata.
The record workstation installed at ITV is easy to use, an essential requirement when a typical session may require more than 100 announcements spread across all services. The presentation suites all have the ability to preview recorded announcements via direct playback workstation. The software is modular and highly logical, allowing ITV staff to maintain and configure the system.
The audio server allows recordings to be made without having to trace, load, label, back-up or archive individual tapes, cartridges and discs. Stored content can be accessed rapidly using the integral database.
The server can also be configured to provide levels of redundancy not normally associated with traditional magnetic media. BWAV files can be copied directly on to the audio server and then logged in the database. Files can be copied from server to server automatically or manually, allowing configuration for full one-to-one redundancy.
Digital audio router control
BBC World Service uses an automated playout scheduling system at its Bush House headquarters in London. The system is IP-based and replaces earlier analog routing equipment. It enables an NTP digital audio router to be accessed and operated via 96 existing control panels in various parts of the building. It also gives each of BBC World Service's 43 language sections the ability to control transmission output using a Web browser.
The BBC's analog system used a proprietary control panel within each studio, showing the time to air and the network to be used. Integrating the new system into the existing panels was much more cost-efficient than supplying, installing and configuring 96 new panels. It also eliminated the need to familiarize staff with a new control interface. Software records no longer exist for the previous system, so a card was designed specifically to communicate with the panels and use XML-to-legacy protocol interfacing.
File-based multichannel audio
The increasing globalization of television broadcasting has greatly expanded demand for multilanguage audio channels. This is particularly relevant within the European Union. In 2004, Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia joined the EU. And more recently, the EU admitted Bulgaria and Romania. Satellite delivery was the great catalyst for the shift from national to international program delivery, enabling broadcasters to address multiple countries with a single signal stream.
A growing number of broadcasters are adopting file-based media asset management to increase their operational productivity. This introduces the far-from-incidental advantage that content files can be repackaged for multiple applications much more easily than was ever possible with tape.
Multilanguage audio channels are no longer limited by the number of independent tracks available on commercial videotape recorders. It also eliminates the inefficient and resource-consuming workaround of running two synchronized VTRs to gain extra audio tracks.
Three major international broadcast networks have recently implemented file-based multichannel audio handling as part of their investment in media asset management systems. Each company's audio workflow was constricted by the limitations of video-plus-four-audio formats and even of the eight-channel Sony IMX tape format.
The starting requirement was possessing the ability to ingest an unlimited number of audio channels from which the desired one or more could be selected for playout. The ability to add further languages after the original asset had been ingested to the archive was obviously essential. The objective was to allow easy management of the entire process, to show which channels had been acquired and which remained to be obtained, then to add the extra language dubs as these become available.
The solution combines the inherent flexibility of efficient media asset management with the open file wrapper format used by Omneon's Spectrum server. The only limitation is on the output side, with the number of audio channels that can be transmitted via SDI port.
A new wrapper file containing the extra audio channel is created for each program. Each audio channel is tagged with the relevant ISO language code inside the file wrapper. This enables the playout system to select any language from the file.
The system includes software that allows implementation of a rule to prioritize. For example, the first choice might be to play Danish if available, otherwise Swedish or, if that too is missing, then English.
The primary server is a large system that includes multiple encoders, decoders and storage. It provides ingest and playout capability and is used as an intermediate store. Online archiving is performed on an external high-capacity server. (See Figure 1.)
A typical operation would begin with a requirement arriving from the traffic system saying a specific program needs Finnish audio added to it. It is then held in the media asset management system among a queue of actions that need to be completed ahead of transmission. Each language dub arrives as a WAV file. The operator forwards the received file via upload screen on a control screen, which uses a standard Internet browser. The system then automatically adds the file to the existing asset and places it in a queue for approval. Once checked and authorized, it is ready for transmission. The system knows all the processes involved. Manual involvement is only required at specific points in the decision chain.
File-based media asset management plays a central role at Technicolor Network Services UK, part of Thomson. Technicolor's multichannel playout facility at Chiswick Park, London, supports eight 1+1 resilient channels and has capacity for upward of 30 additional channels. It supports both SD and HD playout and was designed from its inception with scalability and cost-efficiency in mind.
The system provides workflow management, which integrates asset tracking, rules management, reporting and status. It also provides the tools Technicolor needs to ingest digital content from different sources, generate MPEG-4 proxy browse copies and tag the audio with relevant ISO language descriptors. The new installation has reduced the network's costs and, more importantly, reduced the processing time required to deliver additional languages and channels to Technicolor clients' existing inventory of assets, compared with pulling tape stocks from a library and managing time-consuming re-edits and ingest. The system includes touch-screen-based master control with an intuitive graphic user interface.
The demand for multichannel audio capabilities is likely to increase further as IP-based television program delivery becomes increasingly common. File-based media asset management makes program repackaging and reformatting far easier than was ever possible with the traditional combinations of storing videocassettes on shelves and delivering them by hand.
Spencer Rodd is technical director for Pharos Communications.