Skip to main content

For all the time, effort and money it takes broadcast networks and stations to acuire media assets, often those assets are not treated so well. Tapes get lost the very day they are shot or aired. When the media moves to archive, the challenge of finding it and making it useful again increases.

While there is a rough correlation between the size of the news organization and its ability to catalog, store and retrieve items, probably all news organizations face the question: how can I best store and manage my media?

Some modestly sized stations struggle with masses of dusty tapes and ill-sorted, handwritten index cards alongside graphics lists that are outdated as soon as they are printed. Networks and major-market stations, those with more resources, might have the people to carefully log and store their media. But how can they make that index and the content it represents easily available to hundreds of correspondents and producers scattered across the nation and world?

We see broadcasters around the world achieving new efficiencies by integrating news editorial systems with production equipment, but they also want this integration and server technology to make media assets easier to track, retrieve and share, even after the items are moved off today's shelf.

Asking some key questions is a first step in implementing a media management strategy.

- How much material do I want to be able to air immediately?

- When I am through with this material, where do I want to put it and how quickly do I want to get it back?

- How much do I want to know about this material? The more media I have, the more descriptive information I need to store for each object so I can find it in the future.

- From how many places do I need to play this material? Do I need to do so simultaneously? Are these places in the same facility or are they geographically diverse?

- How much do I want to spend?

Answers to these questions will help frame the media management plan and other key server decisions, such as the choice of a unified vs. distributed video storage model.

In the distributed model, media storage is shared among more than one machine and capacity is increased by adding more servers. All are interconnected, so material can be moved among them. The advantage is diversity and redundancy in the case of hardware problems, along with the ability to dedicate a single machine to a particular production process. Potential disadvantages include possible delays in moving material from one server to another when needed and the difficulties of tracking and storing multiple copies of the same clip.

Multiple server systems can also make integration with desktop production systems more complicated because clients have to point to several machines in order to get to and manipulate their material. Generally, distributed server systems work best in concert with facility control or automation systems, which handle the processing involved in transferring objects between machines.

Adding more storage space (hard drives) to a single system expands unified systems. The number of inputs and outputs can generally be changed independently of the amount of storage available. All users have access to all objects simultaneously without the need for copy or dub requests between machine islands. This reduces the complexity required to integrate other machines and that required to link servers to the newsroom computer system.

Though RAID and redundant critical system components heavily protect all of the unified systems, this design still does not offer the same amount of protection as multiple independent servers. Another potential downside is cost, because some unified systems are significantly more expensive than their distributed cousins.

Accomplishing these goals will mean profound changes in the newsroom, including a major shift in how one thinks about media assets. Terms such as "video" and "still" become merely semantic distinctions. Video and stills become objects that can be easily transferred over wide area networks, stored on laptop computers and even on wireless handheld devices.

The current system of producing news on bits of audio and videotape has many flexibilities, not to mention an obviously dominating constituency in terms of rolled-out production product. In addition, it is likely, according to what we see in early integration plans around the world, that tape will continue to be the short-term solution for at least a news organization's archives. Though an older technology, tape is hard to beat for sheer density of storage per unit space.

But for daily news production, efficiencies in the object-oriented media storage system will be too tempting to resist.

What is a media object? In very general terms, an object is like an envelope into which you have dropped a videotape or maybe a photo. For convenience, you put a label on the front of the envelope describing what is inside so you do not have to open it up and look at it every time you want to use it.

The nice thing about simple envelopes and labels is that, ignoring size, they all can be handled the same way. You can mail an envelope containing video as easily as you can mail an envelope containing a photo. The post office does not care.

Similarly, a file cabinet that can hold photos can also store video. Within the context of a newsroom editorial system, as long as a system can handle an envelope (object) the system does not care what is inside. An object-oriented newsroom system can allow users to view, edit, move and store objects such as still stores, CGs, video, audio and automation events all in the same way: as files stored on a computer.

Once a user learns how to include a CG in a script, they have also learned how to place still stores and video, too. These possibilities can radically transform the way a typical news organization does business - leading to significant new efficiencies.

At the heart of this transformation is a key enabling object technology: the MOS (Media Object Server) protocol, a common way for media and newsroom systems to exchange information over local or wide area network links. MOS has been developed in a partnership between newsroom computer vendors and key suppliers of broadcast production equipment.

Although now primarily in use at single sites, extensions of MOS will soon open up exciting options for news organizations with stations or outposts across the nation and world. These extensions will also facilitate the movement of media assets from broadcast stations to newspaper and Internet content partners.

For example, using media servers, MOS and the object storage model, producers in regional news centers could send playlists to MOS servers hundreds of miles away. They could move media items from server to server just by dragging and dropping a MOS pointer into a script. The goal is to create a virtual worldwide or nationwide newsroom that makes the best use of resources, no matter where they are.

Moving media around transparently and automatically retrieving it from an archive - near-term or long-term - is more than just a technical challenge. From an editorial point of view, one must also deal with issues of sources, compensation and rights. Where did the pictures come from, and who took them? Will there be an extra charge for using them elsewhere or again, or are certain stations prevented from using those media resources?

Manufacturers are already working to address this issue by moving forward plans to convey extensive contextual information - metadata - along with the object itself.

In the AP newsroom model, the editorial script or assignment item from the planner will be the root of this metadata chain. The assignment editor will create a story/planner item, which will be linked to its script counterpart. As the story/script moves through the system, video, audio, stills and CG are added, each with its own collection of associated metadata - descriptive notes, durations, embargoes and the like. It is all kept together, and will keep growing as new, "derived" versions of the story are created for newspapers or the Internet.

A good way of thinking about this model is the menu one gets when right-clicking on a file in Windows, especially in Windows 2000. What you see is a listing of information about that file, some of which you can control, and some of which is "read-only." The key point is that it is all in one place, and it all moves with the file wherever it goes, an essential requirement for managing data and media on a large scale.

Although issues of metadata and workflow in a media object environment are somewhat distant, it is important to consider them as the media plan evolves. Regardless of what method or mode you choose to store your media, the method of storing, moving and retrieving it has to fit into the news production workflow. Without that human integration, the system will take longer to fulfill its promise of making the tough job of news just a little easier.