While researching a feature on archive management, I received somewhat opposing views on the media that should be used for long-term storage. One (old) school likes to keep the original negative or videotape. Another group is a strong believer in data tape. The third group believes disks provide the complete solution, with the “cloud” and virtual data centers being the future. To be pragmatic, all three groups have merit in their arguments, but is there a right answer? I suspect not.
Excluding film, most storage technology uses the principles of ferromagnetism. There are newer technologies, with solid-state drives finding application especially for acquisition and playout servers, but at today's price, they are expensive. Holography has often been hailed as the future, but it remains just that. So ferromagnetism remains the primary means of storage as disk or tape, video/audio, or data files.
An archive can serve several functions. For a production company, the archive is somewhere to store the source tapes, production notes, EDLs and transmission master. A broadcaster is more likely to store finished programs, indefinitely or for the duration of their rights window. News organizations must run large archives, and a local news operation will also want to archive important local stories.
Excluding work in progress, content storage falls into three classes: backup, disaster recovery and archive. Backup is local storage of current work, a precaution against equipment failure. The boundaries between the classes can blur depending on the storage policies a broadcaster decides upon. Move the archive to another site, and it can potentially double as a disaster recovery (DR) facility. However, the program library is not the same. A DR system will likely store the programs and commercials to be used in the next weeks or months to tide over a business while it recovers from whatever disaster shut down the primary site. This is very different from the content stored in the archive.
Inevitably, it's all a balance of cost or storage versus the value of the asset. But the costs are falling, and the future value of an asset is difficult to judge. Deciding whether to archive a content item is often at best a guess. Today, an entire season of a TV series can be stored on one LTO tape that costs about €30. In isolation, the cost is insignificant when compared with the cost of production. Run a library of 100,000 hours, and costs add up.
After deciding to archive a piece of content, the next choice is format — uncompressed or compressed. Herein lies another problem: Uncompressed files obviously cost more to store. But what is considered acceptable compression quality today may not be acceptable in 10 or 20 years time, when that archive file could be uprezzed to UHDTV. This is a new problem. A videotape was archived as recorded. Saving files in an archive presents many options.
I find it interesting to follow the different approaches that broadcasters are taking to address this problem. One needs a crystal ball to build the optimum solution, so a broadcaster's decision represents the best balance of risk versus cost that it feels is right at the time. It is not economically feasible to store everything, although one day new storage technologies may mean that the cost is so low that nothing needs to be thrown away.
Let me know your views, especially on the use of the cloud. Do you want control over your assets? Is security an issue with outsourced storage? And what is the best format: uncompressed, DCT, wavelet or something else?
