RE: [AV Media Matters] Video Compression - a PLUS for archivists (but not too mu

[tony gardner <anthony.gardner@cec.eu.int>]

>At this time, there is no ideal digital video format because
>they either use compression or are very expensive.

It seems to have become a conventional wisdom amongst a section of the
'archive community' that video compression is undesirable. Like many fixed
positions in any walk of life they need to be reviewed and questioned
incessantly: not to do so will leave behind those that cling to it and those
who don't make the effort to understand it in context, isolated from the
main stream and real life.

A further problem - and this is one that I have to deal with - is that a
little knowledge is dangerous. Non technical but influential managers can
quickly understand that compression does (in theory) introduce artefacts and
conclude therefore that compression is bad under all circumstances. Where is
the perspective in such a conclusion? Other factors are much more difficult
to understand - and therefore don't enter into the decision making so easily
- questions of types of compression, types of tape/binder, differences in
DVD and CD recording technologies - the difference in durability between a
pressed CD you buy in the CD shop and a CDR I make on my PC.

I think there is a chasm opening between the professional broadcasters and
the purest archivists. Archives equal money and many of the video archives
are owned by private companies who are going to keep and exploit them for
profit. The managing engineers have got to find cost effective business
justifiable technical solutions. Are you saying that I have to store
everything on D1 or D5 - old formats no longer actively marketed (but for a
time supportable) (at least in Europe) - on machines which are going to be
difficult and expensive to maintain. Why can't I use a mainstream format -
say digibeta or one of the older or newer 50Mbps formats. If you now say to
me I can't because they are compressed, out of principle, even though I can
barely if ever see a difference between the original and the compressed
picture, then I have to say to you that the machines are current,
maintainable and reliable and that's a big part of the practical equation as
well.

Compression reduces the business costs. Good business. It makes it possible
to have a coherent justifiable station archive preservation policy. The key
is for the archivists to take the initiative and control of the compression
to be used. Certainly not DVD-V, probably not systems with long GOPs and
perhaps not the lower acquisition bit rate formats (20-25Mbps). But the
various 50Mbps intraframe only systems are very very good for practically
all TV pictures.

What about the prospects for an archive format for video? This of course
goes hand in hand with the EBU/SMPTE work on exchange formats - also
compressed of course. They always will be, sorry. The compressed version of
the same machine with the same tape and transport will be cheaper to run
(per programme hour) than the same thing recording uncompressed. Who is
going to buy the most machines - the broadcasters or the archivists? Where's
the market research? If the archivists continue to stick out for an
uncompressed format then that is going to be a format ignored by the
majority of broadcast users more concerned with today than tomorrow. That's
going to make the machine even more expensive and less likely to survive the
decades required of it.

Apply the same logic to robotic storage - compression equals more in the
same machine.

For most organisations the sums just don't add up - the revenue value of the
archives does not cover the investment. You have to make it cheaper.
So many organisations are still just ignoring the problems or dabbling
without taking the plunge, hoping for the grail 'soon'.

If the archivists make an alliance with today's broadcasters, their input
will be taken seriously. Make a few compromises,(eg) accept and find the
right type of compression, and you will have a format which is accepted and
used by both the archivists and the broadcasters - and you will be much more
sure that decades of survival are possible.

The most significant advances now coming is where the data format of the
video is support independent and available spearately on a connector on the
output of the video recorder/player: up until now the compression and data
system went hand in hand with the tape machine and tape type, at least in
digital systems. If I can chose which type of support to use to put my video
stream on, without changing that video stream as such, then I am better
prepared for the future (particularly if the decoding is softaware based).

Technology changes - if you miss a couple of IBCs or NABs you don't
recognise the show anymore. If the archive community doesn't move hand in
hand with the broadcasters and find some commonality of approach, then what
each side means by archiving is going to be so different and so entrenched
such that neither will have any empathy with the other. That would be a
shame because it is the professional broadcasters who are producing the
moving picture archives of the future and we are going to do it our way
because we have got to make the decisions now and justify them in financial
terms. The heads have got to come out of the sand and entrenched positions
must become flexible and accountable.

There is a window of opportunity now since integrated
acquisition/storage/transmission/archive systems that work are being sold
and interchange standards work is well advanced. Reassess these principled
conventional wisdoms in the real world and grab the opportunity that won't
be there for long.

Tony Gardner
European Commission A/V studios, Brussels
These are my own opinions and do not necessarily represent those of my
employer.