Preservation Road: Visual and audio records say more than text ever could

NARA works to ensure that future generations can continue to learn from the past, and that increasingly requires the agency to digitize audiovisual records.

The hundreds of tourists who line up outside the National Archives to see the original U.S. Constitution and Declaration of Independence are a daily testament to how much Americans cherish their official history. What they see inside also shows the ingenuity of modern preservation science and the lengths to which the U.S. government goes to protect its founding documents in perpetuity.

The declaration, as signed by Thomas Jefferson, Ben Franklin, John Hancock and others, is sealed in a bronze, bulletproof glass case, protected with humidified helium to block oxygen and other irritants, plus a filter to screen out harmful light. It’s then lowered 22 feet into a vault every night. It’s also watched by a $3 million camera and a computerized system designed by NASA's Jet Propulsion Laboratory that can detect deterioration invisible to the human eye, such as minute changes in readability because of ink flaking or fading.

Those charters of freedom might receive tourists’ devotion — in addition to a lot of technological attention — but they don’t hog all of the federal government’s preservation efforts. The National Archives and Records Administration also is home to billions of pages of material that is considered important enough to be kept as permanent records of the nation’s history. NARA gets those records from all federal agencies under individually agreed-upon terms and timetables. It then puts them through a variety of processes to preserve them and ultimately make them accessible to the public now and in the future.

In this report:

NARA reinvents preservation for the digital age 

Library of Congress defines the state of the art at Culpeper facility

Increasingly, those vital records are not just texts of agency memos, national economic reports, census documents or e-mail messages. They are pictures, sounds and moving images stored in various formats on audio and visual files. They capture everything from NASA astronauts’ first walk on the moon to documentary images of the endangered American bald eagle, videotape of President Bill Clinton’s Senate impeachment trial, audio files of oral arguments before the Supreme Court, moving images captured by Landsat and other satellites orbiting the earth, and — someday soon — full-motion video of battlefield conditions in Afghanistan.

NARA’s audiovisual holdings include 360,000 reels of film, 225,000 sound recordings and more than 110,000 videotapes. And millions more are on the way as the digital revolution takes hold throughout the federal government. Those recorded sources, be they analog or digital, will tell the story of U.S. policy-making and information gathering to the ages in ways that plain text never could. And their preservation needs also present challenges.

“As wide of a technical variety as exists today — from professional broadcast media to low-resolution capture devices such as iPhones — you’re likely to find it somewhere in the federal government,” said Leslie Waffen, chief of NARA’s Motion Picture, Sound and Video Branch.

Although it's becoming easier for agencies to produce audiovisual content, it’s becoming more difficult to establish policies, procedures and technical guidelines for how they preserve the content — partially because the technology used to produce audiovisual content is constantly evolving.

Moreover, the preservation of nontextual records requires extra care, state-of-the art equipment and special storage conditions to ensure that future generations will be able to replay them and experience them the way their original viewers did. Even then, the ever-accelerating digital revolution has promulgated many standards and formats for different types of records, making it exceedingly difficult to settle on standards that will stand the test of time.

“If you give me a handful of digital files now, I can’t guarantee that they’re going to be around tomorrow, let alone a hundred years from now,” said Richard Green of the International Association of Sound and Audiovisual Archives.

And here’s the rub: There’s no enforcer in a position to impose such standards should they be resolved. Mission-driven government officials produce content to meet their agencies’ immediate, present-day needs, not those of archivists. Content from, say, a video recording device on an Air Force drone bomber must serve warfighting purposes first and foremost, even though those images might eventually become a permanent record at NARA.

Through its relationships with agencies — particularly the biggest producers of audiovisual materials, such as NASA and the Defense Department — NARA tries to influence the format in which agencies submit records. However, officials say, they have only so much influence, and their power is limited.

“You could say, 'Well, we want this particular format and this way to capture video,'” said Jason Love, supervisory audio/video preservation specialist at NARA’s Special Media Preservation Lab. “But when you look at the military and they’re putting these small, small recorders in the cockpits of jet fighters or in tanks, a lot of that technology is based around needs.”

Setting Standards for Digital Formatting

Although the digital world ushers in new recording and capturing possibilities, it also raises questions and perils for archivists. That’s where Carl Fleischhauer, program officer at the Library of Congress’ National Digital Information Infrastructure and Preservation Program, comes in.

Fleischhauer leads a working group of federal agencies that focuses on developing federal digitization guidelines for audiovisual materials. The group is part of the Federal Agencies Digitization Guidelines Initiative, and in addition to representatives from the Library of Congress and NARA, its members include Voice of America, the Smithsonian Institution, the Defense Visual Information Directorate and the Government Printing Office. A separate working group is dedicated to still images.

The audiovisual group’s goal is to disseminate information about standards and practices for the digital reformatting of audiovisual materials by federal agencies, according to the group’s charter. “We want to define a set of specifications for the creation of digitized content that are as much as possible in common between different federal agencies…in part to guide our own work when we do it internally but especially with vendor relationships and the acquisition of services,” Fleischhauer said.

The group has its work cut out for it. Unlike still images, for which work on digitization standards began years ago, efforts to develop standards for digital moving images are more recent, and formats vary.

If digitization standards for still images could be called mature, Fleischhauer said, he would describe the agreed-upon practices for digitizing sound recordings as semimature, while conventions for video recordings are less advanced. His group is heading toward the publication of an advisory guideline for sound recordings but is nowhere near that point for video recordings.

“Part of the reason it is challenging is that there are any number of practices in place for digitizing video,” he said, adding that the technical solutions that broadcasters have used aren’t considered by some to be appropriate for preservation-quality reformatting. “The point is, we haven’t hit a point of consensus about this, and that’s where that exploration continues.”

Fleischhauer said there are many questions about how to store and digitize video files because the file sizes are huge, and storage space and transmission time are considerations. There also is disagreement about file formats.

For example, for decades, the film stock used to make motion pictures was fairly constant, and there was a steady supply of it for people such as Audrey Amidon, a motion picture preservation specialist at NARA, to make copies for preservation.

“Mostly our goal is always to stabilize the original, to inspect it and assess its condition, store [it] in the proper housing and keep it in the right environment so that it will last as long as possible — and if necessary, we copy it to a new piece of film stock [that] should last another hundreds of years,” Amidon said. “The day that we will no longer be able to do that is coming soon.”

Changes in how movies are produced mean that companies that have traditionally produced film stock will not be able to continue to do so, experts say. Those changes aren’t limited to film. Other analog formats are also becoming obsolete, and agencies increasingly digitally produce sound, video and images that might eventually be considered permanent records.

A common thread running through NARA’s diverse audiovisual holdings is that agency originators generally didn’t create them with archivists in mind. The recordings' formats are often more a reflection of the needs of the agency than NARA’s preference.

Jeffrey Reed, supervisory preservation imaging specialist at NARA’s photographic imaging lab, said that although NARA officials have conversations with agencies about preferred formats, they don't wield enforcement power.

“Even though there might be a format that’s ideal for us to receive and keep things in, we really aren’t in a position to dictate that to the people who are creating the records.” Reed said. “But when asked, we’d love to be able to recommend to them, ‘If you start out this way, it’ll make our lives a lot easier.' ”

Meanwhile, Waffen said, the digital revolution has made it even more crucial for archivists to pay attention to the state of incoming records. “When we were an analog archive, you had less choices and you had a little more time before obsolescence kicked in,” he said. “When you get into digital formats coming from agencies to the archive, then you’re an active archive and you have to constantly be vigilant. There’s refreshing and migration and all kinds of digital asset management that comes into play, as opposed to a more passive archive that can let something sit on the shelf for a while.”

NARA employees say proprietary software makes it harder for archivists to maintain or play back records. The agency might think it was getting a great deal on software, but records that require hard-to-acquire licensed software can quickly become a headache.

Dan Rooney, an archivist at NARA’s Special Media Services division, said he would ask chief information officers to think more about long-term viability and stay away from proprietary formats, storage systems, servers and tape-writing capabilities. “Those are all things that are going to be problematic for NARA once they start transferring their records over here,” he said.

Drawing the Line on Metadata

The working groups' recommendations are voluntary for agencies, but officials hope that if the audiovisual or still-image group issues a guideline, agencies will insist that their employees follow it as much as possible, Fleischhauer said. Issues on the audiovisual working group’s to-do list include writing a useful, explanatory, broad guideline for sound recordings; consulting with experts to compare different digital video formats; and coming up with guidance for maintaining metadata about technical aspects of recordings.

Fleischhauer said a common theme in the group’s discussions concerns the metadata, or technical or content data, of the records. “It’s a line-drawing exercise, and not everybody agrees about where to draw the line on how much metadata to embed,” he said.

For example, how much data about the digital formats should be included in the video of the moon landing, which NARA has worked with NASA to restore?

The debate over metadata is especially important because such information can restore context that got lost in the digitization process. “Are we providing enough metadata that there is enough context that somebody could really understand those things about an individual item?” Rooney asked. “I agree that some of the context is kind of lost, and it suffers in just a purely online delivery. It’s a little colder.”

But even with great metadata and digitization technology, original analog materials can’t be completely replicated. NARA’s Waffen said he worries that analog material might not be saved in the long term.

“I think you need to retain the original material as much as you can to truly be called an archive,” Waffen said. “There’s something in the intrinsic value of the older material and the formats it’s on that I would hate to see lost.”