White Paper: Digital Archive Transforms News Processes At CNN

Solution
Media Information Retrieval Application (MIRA) - a digital asset management system built on IBM DB2 Content Manager, IBM DB2 Universal Database, IBM WebSphere, IBM Content Manager VideoCharger, IBM pSeries™ and IBM xSeries™, Intellistation workstations, FAStT700 Disk Subsystem; Consulting and system integration services from IBM Global Services.

Launched in the United States in 1980, and introduced in Europe five years later, Cable News Network (CNN) is one of the most pervasive and prestigious global 24-hour TV news channels in the world. With around 4,000 staff worldwide, 39 news bureaus and more than 700 affiliated TV stations. CNN delivers fast, accurate reports on live breaking news events all over the world. Noteworthy CNN footage includes the fall of the Berlin Wall, Nelson Mandela's release from prison, coverage of the war in Iraq and the worldwide Millennium celebrations.

CNN and CNN International are now available to more than 1.5 billion people in 212 countries via 15 cable and satellite television networks and eight Web sites. CNN's interactive services were launched in 1995 and currently receive an average of 30 million page views every day. CNN also has a financial news channel in CNNfn, CNN Headline News, CNN en Español, CNNRadio, the CNNtoGo mobile service, CNN Airport Network and AccentHealth.

Following the merger of Turner Broadcasting System with Time Warner in 1996, and the subsequent merger with America OnLine in 2001, CNN is now part of one of the largest multimedia production, marketing and distribution companies in the world.

Content concerns
Most companies operating in the television and video production industry face a similar challenge – how to manage and ultimately exploit their large repositories of content. Achieving sustained industry success is driving many companies to develop the technology infrastructure to create, manage and distribute content through today's multiple channels, whether over the air, via the Internet, or through interactive cable services.

CNN has always been committed to producing high-quality content, managing it efficiently and extending its use as effectively as possible. Therefore, the company was quick to recognize that its future systems needed to be able to manage, access and deliver content with greater flexibility.

The decision to go digital was imperative for CNN. Without the move, it faced an ongoing, time-consuming tape dubbing process. With the huge volume of footage being added to the archive annually, this alternative was becoming increasingly undesirable. In addition, much of the older footage was disintegrating due to age and repeated use. CNN needed a system to preserve this historic and valuable asset as well as be able to use, modify and deliver this content in a more accessible and interactive way.

Digital developments
With its Archive Project, CNN was one of the first broadcasting companies to move into digital asset management, and its archive is set to be one of the world's largest.

The multi-year Archive Project is aimed at providing CNN with the infrastructure to manage digitally 200,000 hours of archive material gathered over the past 24 years, as well as 40,000 hours of new footage arriving annually. The Archive provides an online digital library allowing easy tracking and sharing of archived content across all of CNN's businesses and which will facilitate sales and exchanges with other companies.

CNN wanted to preserve the network's archive, be able to distribute content in new ways, and improve the efficiency of the existing news operation. The system to address these goals was based on open standards for ease of integration and avoided vendor-specific formats or applications to allow ease of migration and expansion.

The archive project , began with an in-depth analysis of the user and design requirements. Considerable time was taken with the users to understand how the system would be used on a day-to-day level. As part of the detailed design, screens were developed and revised with the users until the project team was sure it was developing the best, most user-friendly digital asset management system possible. A flexible workflow system was developed that covered capture/ingest, index and catalog, search/review and select, and fulfillment as well as comprehensive system administration functions.

Core components At the heart of the system is the scalable, resilient IBM IBM Series servers and a custom application developed by IBM.

Video from analog tape is brought into the system via ingest workstations. The library specifies the content needed, the priority of the request, the browse-quality proxy video clips and the treatments required, such as closed captioning, scene detection and speech-to-text. The Ingest Operator locates the source – usually on tape – views it to find the required segment, and processes the ingest request. The video is captured both at high-resolution and low-resolution browse-quality proxy. The content is MPEG encoded, and monitors attached to each encoder help to ensure the Ingest Operator can detect any problems immediately. The system also takes input from CNN's Newsroom system. IBM integrated Virage's VideoLogger into the system to provide users with advanced image and text analysis, as well as intelligent video segmentation and metadata creation to enrich images and enable easier searching.

Once ingested, the content is cataloged and indexed - essential for easy retrieval. The cataloger plays back the video proxy file while writing the video annotation. Keywords are added into the catalog record. These keywords are checked against the customized vocabulary dictionary and challenged if ambiguous or spelled incorrectly.

The metadata content is stored on the large RAID disk array attached to IBM pSeries servers running WebSphere and DB2. The digitized video is stored on a robotic tape library in compressed digital form for economic storage and fast retrieval with the most frequently used video stored on a cache on the disk array. IBM DB2 Content Manager is used to store and manage all the metadata as well as the storyboard "blobs", low-resolution video and text. The Content Movement Locator (CML) keeps track of the location of every asset and performs the move operation when content needs to be stored or retrieved. CNN plans to extend the system to use satellites and its own closed network to distribute content around the CNN enterprise at high data rates.

Streamlined searching
Locating video clips was previously a complex, time-consuming process. A production assistant sourced the analog tapes and mailed the video from CNN's headquarters.

By putting CNN's analog videotape archive online, the company has provided a digital TV news library or ‘virtual filing cabinet' for its journalists and news producers making it faster and easier to assemble, edit, retrieve and distribute high and low resolution video clips. CNN staff can now access digital video, audio, graphics and text files electronically via their desktops.

Users of the new system can search the catalog from their desktop via the search application to locate content using free text and parameter-based queries. Queries can be refined to help identify the best content, whether digital video, audio, graphics or text files.

With the new procedures in place, CNN is increasing the efficiency of its infrastructure, developing new revenue opportunities and meeting new product and distribution opportunities. CNN's historically significant archive is being protected from further deterioration and damage, thus helping to ensure its quality, accessibility and conservation.

"The emerging digital archive is a core part of CNN's Integrated Production Environment. This all digital file-based environment will allow for transformed workflow and provides the potential to re-purpose content for many different functions with amazing flexibility."
-- Gordon Castle, CNN Senior Vice President of Technology.

Source: IBM North America - Retail