How is the digital world affecting the role of audiovisual archives? Last week the Metropolitan New York Library Council (METRO) and New York University's Moving Image and Preservation Program (MIAP) presented a workshop on preserving locally-produced digital audiovisual content, which tried to provide some answers. Hosted by MIAP's Howard Besser, it featured four other speakers dealing with relatively small, community-based projects:

• Sady Sullivan, of the Brooklyn Historical Society's "Crossing Borders, Bridging Generations";
• Natalie Milbrodt, of the Queens Memory Project;
• Brendan Allen, of Democracy Now!; and
• Yvonne Ng, of WITNESS.

The latter two, being broadcast archives, touched on issues we face daily at the New York Public Radio Archives.

The workshop directly and indirectly sparked some thoughts on the subject. Below are three of them.

1. Digital preservation is young

The first "wax" cylinders identified performers only through a slate in the recording itself; unless the user manually labeled them, the only way to know what was on the recording was to play it. As people's collections grew, it was clear that this was impractical, so pretty soon cylinders had some information engraved on their edges and on a label affixed to the (otherwise generic) box. Similarly, "78" shellac discs used to only have labels affixed to their center; but even though the label had much more detailed information than what you could engrave on a cylinder's edge, you still had to pull the disc out of the shelf to get an idea of the content. Eventually, "albums" (as well as vinyl LPs and Compact Discs) developed sleeves with spines that could be read on a shelf, without having to pull out the disc. Note that as collections grew in quantity, so did the ease of access (although that does not mean that collectors did not have to do a little organizing).

Howard Besser showed that with digital audiovisual materials, created in staggering numbers, our current discovery methods are still primitive. However, they keep getting better: inexpensive speech, music and image recognition are surely coming soon. Will neatness count? As big data discovery methods become more sophisticated, neatness may count only if you want to be independent of profit-driven discovery methods.

2. The onus on preservation will always be on those interested in accession

A challenge common to broadcast archives is the ingestion of files inconsistently formatted and named by disparate producers (what, in OAIS terms, would be disparate SIPs). This frustrates broadcast archivists, but we need to realize that producers have developed methods that usually work: most of the time  they find what they need. Archives will necessarily have to implement different ways of dealing with information which may actually hinder an individual producer's workflow or even their ability to access their own materials.

More broadly, the complexity of archiving depends on the scope of potential access:

• how long we want to keep that file and
• how wide is the circle of our potential consumers.

As Yvonne Ng pointed out, the longer we need to keep a file, the more complex its archiving. If a producer wants to make sure her heirs can benefit from her copyrighted work, she will probably take extra steps to ensure the safeguarding and viability of her files. For example, hopefully she will not expect her hard drive to function long after she is gone.

Similarly, a producer may have a quirky naming convention for his footage, but if it works for him, he has no reason to change it. However, if he wants to publish his finished story on a national platform, he will have to submit a description along, and possibly transcode his product into a different format. Finally, if he wants to post it on youtube, he will probably have to add some tags that youtube recognizes. Each widening circle of potential consumers requires a different approach.

What all this means is that, although archives can try to suggest strategies that follow the principle that "archiving starts at creation" (a laudable principle that theoretically gives good results), it will often be an imposition on other workflows and thus hard to justify. Which brings us to our third point:

3. Archives need to make friends

Digital preservation requires a constant commitment of resources. This translates into having to justify preservation much more frequently than in the old days, when a tape on a shelf could almost be forgotten without much investment (or dire consequences).

Due to the staggering volume of material being produced, archives need to be smart about managing the complex processes in digital preservation. This means creating automated workflows when possible, and also creating a consortium of stakeholders that can ensure a longer life for the digital assets. Stakeholders can be the umbrella organization, individual producers, access platforms, copyright holders, the public at large, etc. Working together will aid in developing consistent workflows that will result in long-term benefits for everyone. Now more than ever, no archive is an island.