Archiving Issues

I attended a great talk last night hosted by the Asian Film Archive. They brought up Ray Edmondsun from Australia’s National Film and Sound Archive to give a Q&A session about archiving, preservation and ethical issues in film restoration.

Some of the more interesting topics discussed were of course the film vs. digital debate. Of course film is still by far the best archival medium we have. Properly stored, it can last over 100 years, and the technology to read it is easy to reconstruct all you have to do is simply shine a light through it.

So my question was, if film is the best archival method, what is the second best for digital assets? The answer was there wasn’t one. Things get trickier when you move into the digital format. Tape or Hard Disk? Will that tape or disk format even be readable in 30 years even if the media survives? What file format to store it on? Essentially while film is a long term archival solution, digital requires a constant migration from format to format throughout its life. The downside of this is that it’s very likely that somewhere along the way, someone won’t migrate it fast enough and the data will be lost. The upside is that digital distribution means that it’s easy to propagate many copies around the world rather than striking film prints. Archiving digitally is a moving target.

Another interesting topic was on the difference between the restoration of an old film, and the reconstruction of an old film. Restoration involves keeping the film in as close to it’s original condition as possible. Removing scratches, dusts, splicing errors, and faded colors were fine, but you left everything pretty much as you would have found it. A reconstruction on the other hand is designed to make the film viewable for modern audiences. So for example, you could clean up the audio track, and remaster it, reframe some of the shots to match a specific aspect ratio, make color correction decisions, even fill in missing scenes with new footage. Whatever it takes to tell the story again, rather than simply preserve what was there. Archivists often battled between preservation and accessibility, with the results usually being somewhere in the middle.

You ever wonder why an audio fade is called a fade? Well I found out last night. It turns out that in the early days of sound, there was no way to mix audio. This was before electronic mixers, you had only one audio track for all your sound, and it couldn’t overlap (try editing with only one audio channel enabled). The sound was recorded optically on film and so if they wanted to fade out the music at the end of the scene, they dipped the actual film strip in peroxide! If you wanted a three second “fade” you took the film and dipped it in the peroxide from the bottom up, slowly lowering it in over a period three seconds, and removed it. The peroxide literally “faded” the optic track away!

From the I can’t believe I haven’t read this book already file: Ray Edmonsun recommended the book “The Parade’s Gone By” by Kevin Brownlow, as the seminal text on the silent film era. It was written in the 1960’s and has interviews with the great stars and directors of the day before they passed away. I’ve already got a copy on the way!

Another interesting topic was what to archive. Obviously the film itself, but Edmonsun stressed that the real find for archivists included the shooting or lined script, as well as the promotional and marketing materials. This helped immensely in any research and reconstruction work. He stressed that you shouldn’t wait until the movie was finished to think about archiving, but to have a plan in place before you roll the camera.

Submitting your film to your country’s film archive was the best way to preserve it for the long term, they’re not just after blockbuster feature films. Doc’s, shorts, travelogues, anything that sets a time place or culture is their purview.

All in all a very interesting talk on a very niche subject!

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *