Our June Broadcast Engineering world edition carries a feature on field production and how workflows have changed now that cameras are tapeless. Once the production is finished, this raises a question: What happens to all of the files?
Many sit on hard drives — not a place to leave them if they are to be archived. With the insatiable demand for content, an increasing number of content providers are mining the archive. Today’s production may have a future value. To meet that possible demand, the program master can be archived. But how much else should be archived. What about the original camera files, the production notes and the EDLs? One could keep everything, but there is a cost attached.
With taped programs, the tape may well have been stuck on a shelf somewhere, and left for future worry. Archivists generally class a short-term archive as less than twenty years, and long term for time longer than that. Videotape should last at least twenty years before it starts to deteriorate, so a short-term archive can be implemented by storing videotapes. But, what if there are no tapes? The Academy of Motion Picture Arts and Sciences has dubbed this “The Digital Dilemma” (although they were more concerned with the motion picture sector). In its reports from 2007, the lack of archiving strategy was identified throughout the media and entertainment sector.
The film guys have been here before. When they moved from color separation negatives (long-lasting RGB silver images) to color negative, the early dyes were found to fade faster than expected. Some films literally faded away. Fortunately, the chemistry improved as a consequence.
Much has changed in the intervening years, with developments in file formats, metadata standards and the acceptance of LTO data tape for archiving programs. That said, large collections of hard drives are accumulating.
The big problem with digital archives is migration. Technology becomes outdated, so files must be migrated every four or five years to the latest storage medium. This requires management. A tape on a shelf is a real estate problem (although it does needs climate control), but digital migration requires an ongoing maintenance cost. One only has to look at CMX EDLs stored on eight-inch floppies; how do you play those back?
Some might say “stick it in the cloud,” but that is not an answer. Actually, it’s just kicking the can down the street; how many of today’s cloud vendors will be around in 50 years from now? Two clouds would provide better security, but cost would double. Most cloud vendors come from a generic IT background, and their services are suited to office storage — documents and databases.
There are a few media-savvy clouds emerging for this market that are designed to handle very large files and high bandwidths. As storage gets cheaper over time, IT systems will more and more offer costs that are in balance with the low-revenue business of serving long-tail content. As we move forward, mistakes will be made, and content will be lost forever. But, we will learn how to balance archive running costs against sales revenue of legacy content assets.
For production companies the big question is how much to keep. Does the program lend itself to future repurposing, in which case as many of the program assets must be archived, or is it ephemeral?
