[AV Media Matters] Compression

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

My original posting - in case it was not clear - was meant to refer
specifically to using compression to preserve material which was
originated as standard definition video colour material (SDTV). This
is the starting point where no better quality original exists, so
all the discussion about interlace, colour bandwidths etc. is
irrelevant. I remain convinced that the 50Mbps &/or DB are
effectively non destructive means of preserving such material at
100% of the original quality. There is no advantage in specifying
existing and new non-compressed format conservation for this type of
material, and many disadvantages, as discussed in my previous
posting. It is an act of pedantry to think non compressed to the
exclusion of other much more important factors .

For material which was originated in better quality than SD TV, and
in particular film, (and Jim's Bill of Rights) then the arguments
above do not apply. I am surprised that anyone even contemplates
preservation of film by SDTV transfer - the inherent definition
limitations, plus interlace, colour bandwidth, aliasing, even before
video recording, plus artefacts from digitilisation and the video
recording defects itself argue against it. Surely only HDTV or the
Hollywood definitions are realistic candidates with the required
bandwidths/definition. And where is compression/format in this
paragraph of discussion? again it is overshadowed by the rest. Can
one take seriously the discussion between D1/D5 and DB for these
types of materials when the vandalism - Jim's chain saw - came long
before the video recorder.

The point about access vs preservation is of course important.
Preserve at 100% quality (which can include compression) but access
at lesser qualities is good practice. Equally, there will be not one
but a family of preservation formats (cf ATSC) as Jim says.

To repeat my argument - for that subset of original SDTV material,
50Mbps and DB compressed formats represents non destructive 100%
preservation. It is irrelevant whether there will ever be a final
data format - there will never be. The fact is that NOW there are
established good data formats and stategies for preserving SDTV
which can be future proofed for tens of years at least. Storage is
always going to cost money - a lot less per GB in the future maybe,
but I am talking now. The argument is to start today and get some
preservation and access work done, and set some de facto standards,
with the good tools that exist now. Use the opportunity now to
cooperate with the broadcasters and manufacturers to mutual
advantage. There is good business money out there in these archives.

Tony Gardner
EC A/V, Brussels.
These are my own personnal views and do not represent my employer's policy.