Wednesday, 27 July 2011

Preserving born-digital video - what are good practices?
Interesting to see Killian Escobedo's post on digital video preservation over at the Smithsonian Archives' visual archives blog. Our trainee, Emma, is working on questions of these sort at the moment as we start to develop strategies for preserving the vast amount of born-digital video being deposited in our archive collections. While there's quite a lot of material out there on digitising analogue video, we've found a real shortage of guidance on the management of born-digital video collections. With that in mind I'd be really interested in hearing how other folks are dealing with this kind of material. Can you give us any pointers? At the moment we're particularly interested in learning more about existing practices, good tools, realistic workflows, and preservation-grade standards (for metadata and content - which ones and why?).

So, what kind of digital video do we have? It's a good question, and one I can't answer fully for the moment. What I can say is that our collections include digital video deposited on CDs, DVDs, Bluray discs, miniDV and mediumDV cassettes, and hard disks. Much of this material has yet to be captured from its original media so we don't have that inventory of codecs, wrapper formats, frame rates, metadata, etc. that Killian talks about. This kind of detailed survey work is a next step for us, but one that will have to wait until we have developed a workflow for initial capture (bit-level preservation comes first). I wonder if we'll see the same diversity of technical characteristics present in the Smithsonian's materials. It seems likely.


Simon Spero said...

Since you're starting with born-digital, you at least get to skip the tricky part and go straight to the really hard stuff :-)

A few thoughts off the top of my head:

1) VideoCD discs are MPEG-1, so they're pretty easy to handle.

2) DVDs are MPEG-2 encoded; some commercial releases may be CSS encrypted; CSS encryption is solved, and open source libraries are widely available to remove it.

3) BlueRay may use HDCP protection, which has also now been cracked; software is not widely available at this time.

4) Under UK law, unless there is a specific license agreement to the contrary, copyright owners are supposed to provide a means to allow archives to exercise their right to make preservation copies (or at least, you petition the secretary of state to lean on them). If the copyright owner is also the donor, then copy protection is unlikely to be a problem.

5) DV tapes put the "not archival" in "that's not archival"... get the bits onto spinning media asap. The copy should be as good as the original, assuming that original data is still good. The content is encoded in a standard format without much metadata, so it's especially important to keep track of which file came from which tape. Once you've got the data, it is supposed to be better if you don't rewind the tape until needed. When a tape snaps , Reel-to-reel splice-fu can be used to keep going, with some loss for the spliced part. If this happens it might be an idea to do the rest of the acquisition on your least favorite deck.

6) At some someone will bring up MPEG-7.

7) The AVATAR-m project may be of interest:

Susan Thomas said...

Cheers Simon :-) Couple of links there we hadn't reached yet.