Reformatting video

Scott Campbell <scottcampbell [at] adidam__org> writes

>Jim Lindner <vidipax [at] panix__com> writes
>
>>For archival purposes we normally recommend composite digital
>>formats that use no compression--these formats are known as D2 and
>>D3....
>...
>Based on your posting, I have a question about the difference
>between composite digital and component digital for archival
>transfers. Why do you "normally recommend" D2 and D3 composite over
>component D1, D5 ,or digital Betacam?

The questions you have asked require are fairly technical and would
require an extensive response that would include a background on how
television works as well as color and encoding theory--but here is a
very short version.  Component video formats (both analog and
digital) separately record the color component signals.  These
colors in the additive color system are Red, Green, and Blue--the
combination of which will make white and the absence of which will
create black.  In practice the colors that are recorded separately
in component signals are often the "color difference" signals--those
being the difference between the luminance (or Y) portion of the
signal and the chrominance portion--and as a result component
recording in video will record the Y, Red - Y, and Blue - Y signals
or a variant of them.  This is what component recording is whether
it be analog whereby you are recording a variable signal or digital
whereby you are recording a sampled representation of that same
variable signal. Some digital component systems use more bits then
others to represent the signal, and some use higher frequencies than
others to sample the information.

At the current time, television component signals are *not*
broadcast and never have been commercially due to limitations of the
bandwidth that such a signal would require as well as for
compatibility reasons with black and white television.  Color
broadcasting in the United States is encoded using a system called
NTSC which encodes the color difference information into the black
and white signal--this combined signal is called composite
video--and this is the type of signal that is recorded on your 1/2
inch tapes. Other countries have similar color encoding schemes
(PAL/SECAM and variants).   In order to change a composite recording
to a component recording it needs to be decoded (as you may
correctly assume--getting it composite in the first place required
some fancy encoding footwork), and there is no perfect type of
encoding or decoding nor are there just a few types. There are
many types that have many different results, so by taking a
composite format which is analog which is what you have and
transferring it to a digital betacam format which is component you
are changing the basic way the signal was recorded and you can also
easily change the way that it appears as well.  As such composite is
"closer" from an archival perspective to the original recording.

In addition, in general composite formats are less expensive then
component formats--as an example D2 and D3 machines are generally
less expensive then D1, D5 or Digital Betacam--this is always true
of both  machines and usually true for the tape that they use--which
is a significant cost in a migration effort.  For example D5 tape is
used twice as quickly in a D5 machine then D3 tape in a D3 machine
and therefore the D5 tape is more expensive.  There are some
differences of course between formats and vendors so Digital Betacam
tape might be less expensive then D5 tape (or maybe even D3 tape
depending on the marketplace at any given time).

Of the formats discussed, Digital Betacam is a compressed format--it
has fairly low compression compared to others--but it is compressed
at a ratio of 2:1 or 50% using lossy compression which means that
50% of the information is thrown out in the digitizing process (this
is quite different from lossless compression that does not discard
the information). Because of this most people in this field try to
steer clear of compressed formats in general--but people have
preferences for other reasons depending on their internal use needs.

Getting into a competitive evaluation between vendors and products
is not possible here--they all have their plusses and minuses.  At
VidiPax we do *not* charge additional for recording 2 copies at the
same time for precisely this reason--normally our customers choose
to record one tape in a component analog format like BetacamSP and
the other copy in a digital format of their choice (often digital
composite like D3 or D2).  They each have their disadvantages and
advantages--and we support ALL of them.

Jim Lindner
VidiPax
The Magnetic Media Restoration Company
450 West 31 Street - 4th Floor
New York, N.Y.  10001
212-563-1999
Fax: 212-563-1994

                                  ***
                  Conservation DistList Instance 12:4
                  Distributed: Thursday, June 18, 1998
                        Message Id: cdl-12-4-014
                                  ***