The Moving Picture
I Love Digital Video
Not surprisingly, given the nature of
this column, I really like digital video. What might come as a
surprise, however, is why. Itís not that Iím a big fan of the
stuttering frame rates of todayís Web video or even of the flamboyant
special effects enabled by digital video editing tools. I like
the idea that digital video is data and, as such, not inextricably
linked to a physical videotape or film reel.
I like that digital video can be copied
without generation losses, eliminating the reliance on a single
ìmasterî that must be guarded like a rare but biodegradable coin
or lost to history forever. Itís robust, too, and doesnít accrue
drop-outs or color bleed from friction with the tape deck heads,
nor does image sharpness fade into a two-dimensional blur like
with VHS tape.
Whatís more, digital video data doesnít
turn to dust like the aging film reels from Hollywoodís golden
age. CD and DVD discs last longer than film and tape and, while
files can be corrupted, digital video can be backed up like any
other valuable data. Since copies are identical to the original,
nothing is lost. To me, this ethereal nature is more like pure
art and ideas.
Benedetto Croce, the late-19th-early-20th-century
Italian philosopher, believed that the artistic process was completely
a result of what he called ìintuitionî and, as such, accomplished
in the mind rather than through external or physical communication.
For Croce, artistic vision itself was paramount, not the mechanical
outward creation needed for communication. Admittedly, Croce wrote
more about poetry than the more tactile forms of art, and his
arguments make more sense in that context, but I suspect Croce
would have liked digital media, too.
My wife is also a fan of digital media,
and works with it daily in her career as an archivist. (Not to
be confused with anarchists, alchemists, or Acadians from the
Maritime Provinces of Canada, an archivist works with historical
documents, records, photographs, and videos.) Unlike a librarian,
who essentially works in the present organizing and referencing
information, an archivist works with history for what it is, gives
it context in the present, and decides what relics to keep for
future generations and which should be weeded out.
My wife also regularly works with current
technologies, including digital media. Her recent projects have
involved building topical Web sites for portions of a collection
using photographs, streaming media, and Java.
My wife, however, hastens to point out
the shortcomings of digital media for archives. She sees the potential
dangers of non-standard media formats that may not be viewable
years in the future, like early computer data stored on 5 1/4-inch
floppies and early tape drives. Without standardization, digital
video faces the same problems.
Digital media files can also be changed
or altered far more easily than physical assets, often without
reference. For example, simply opening a file and changing one
bit alters the file date, and its history becomes confused. The
lack of a single physical master, in this case, also introduces
questions as to whether one version of a file is actually true
to the original.
There are few standards for keeping track
of data and organizing digital media so it can be stored logically
and retrieved efficiently when needed. Video producers often have
their own methods for organizing materialósuch as timecode, shot
type (close-up, medium shot, landscape, etc.), topicóbut those
systems are usually personal. For an archivist working with all
types of digital media, a standard way to account for what information
is tucked away in all those 1s and 0s is critical.
These problems are what the MPEG committee
is trying to solve with MPEG-7. Unlike its past work creating
highly efficient audiovisual compression, this future MPEG standard
takes a completely different tack by creating an infrastructure
within which all present and future digital media can be used.
Formally known as the ìMultimedia Content
Description Interface,î MPEG-7 attempts to lay out a standard
manner in which digital media can be referenced, archived, and
retrieved. This standard does not preclude new coding methods,
but essentially will work with existing media files of all coding
types, including even analog media. In other words, MPEG-7 takes
the metadata object of MPEG-4 (see ìWhatís It Good 4: Encoding
and Applying MPEG-4î in this issue, pp. 38óEd.) and gives it some
More than just copyright information,
an MPEG-7 file could theoretically contain file creation and compression
history, as well as project data from that non-linear system so
the file could be reopened if necessary. Security information
could restrict file access where license or royalty fees are due
and could even include past and future play schedules for widely
Perhaps even more importantly for archivists,
an MPEG-7 data file would contain standard descriptors about the
audiovisual material. It would allow smart searches on content
topics, specific video images and objects, audio sounds (including
perhaps even individual voices), and scene types so archivists
could easily find and use media in presentations about the past.
Media management on this scale is a major
undertaking, ifÝ a fairly unglamorous one, for video people accustomed
to the pace of moving pictures. But for an archivist itís exciting
stuff. And, if successful, it could be the missing piece for corporate
video as well. While the idea of using corporate video to enliven
presentations, training, and promotions is an appealing one, those
intentions quickly fade if the fast pace of business leaves us
no way store and find the material.