Conservation DistList Archives [Date] [Subject] [Author] [SEARCH]

Subject: Stability of videotape and optical disks

Stability of videotape and optical disks

From: Doug Nishimura <dwnpph>
Date: Sunday, May 19, 1991
I was at an SMPTE (Society of Motion Picture and Television Engineers)
mini-conference.  I was horrified to hear a couple of speakers from the
industry tell the group that video tape is more stable than film!! I've
heard this hype from the manufacturers for a few years and just
dismissed it as a marketing ploy (claims that your video taped memories
will last forever).  To hear such nonsense at a technical meeting,
however, tells me that it is not just a bunch of marketing people
claiming this, but that it goes right up into the management sector.
The three were all managers of some sort (product manager, national
technical manager, and a technical service manager).  Their claims were
challenged, of course, and they showed that they didn't know very much.
The three were from two different companies.  Just in case library
people are being given fantastic claims from tape and optical disk sales
people don't believe it.

First the stuff from the companies at SMPTE.

1) They claimed that color dyes fade (true) and that some Quadruplex
tapes were unearthed from a hole in the ground and (after they built a
machine for them) they played.  The Quadruplex tapes were made back
around 1957, but I don't know when they were buried.  This was one
isolated case and lab tests indicate that this was not a likely event.
The tape from that time was likely cellulose acetate and the binder
(that holds the magnetic particles to the tape) was polyester urethane
-- both highly sensitive to high humidity.  The both suffer from
hydrolysis.  The tape support becomes brittle and shrinks severely while
the polyester urethane turns into a thick goo.  I suspect that the hole
was pretty dry and that color materials would probably have done about
as well, although color dyes in the '50s and early 60s were not so great

2) I accept that improvements have been made to the binder and to the
iron particles (on new metal tapes) to help make them more stable.
Apparently a thin, controlled oxide covering over the metal particles
makes them more resistant to oxygen and water.  In addition, some
improvements to the binder has made it act as a more protective barrier.
However, unless they have changed the basic polymer, it will still
suffer from hydrolysis. Water scavengers in the polymer wouldn't help
(like trying to dehumidify the atmosphere).

3) They have put in a pre-shrinkage step in the manufacturing process.
This helps stress-relaxation shrinkage.  They have also improved the
smoothness of the back side of the tape, decreasing pattern imprinting
on the binder layer of the next lap of tape.  The patterning decreases
the head to tape contact and ruins the image and sound.  However, some
of the improvements have counter effects too.  To help decrease the
stress shrinkage even more, they have made the tape even thinner which
makes it more susceptible to magnetic print through.

4) They claim a predicted life of 25 years based on a single condition
(60 C/90% RH) incubation!! Obviously they haven't heard of Arrhenius and
reaction kinetics.  Arrhenius worked out a temperature and reaction time
relation for special reactions called "first order".  In some cases,
other types of reactions can be manipulated such that pseudo-first order
reactions can be derived.  However, Arrhenius predictions depend on
establishing a RELATION between temperature and reaction time.  The
weakness of single condition incubations is that at a particular
condition, material A may last 30 days longer than material B.  However,
material B is much more temperature sensitive than A.  Thus if we
extrapolate back to real life conditions, the reaction rate for B is
decreasing per degree change faster than the reaction rate for A.  At
real life conditions, it is quite possible that material B is many times
more stable than A.  There simply is no law of kinetics that allows life
expectancies at real life to be calculated from a single condition

5) Real life cases were sited that said that certain tapes in a library
collection handled about 500 plays (for one type of tape) and 2000 to
3000 plays (for another type of tape) before they suffered from case
failure. This is like predicting that my car will last 1000 years based
on the rate of rusting of the bolts on my seats!  Depending on the abuse
and condition shock, I would expect the case to far outlast the tape.

Optical disks are another problem.  There is very little experience with
the stability of this medium.  THey suffer from heat, humidity and
physical stress.  THere is very little data on the adhesion between
layers.  One disk, for example, has 5 micro-thin layers of alloys bonded
together as a sandwich.  These layers are then bonded to a sheet of
polycarbonate.  How are they stuck together?  How well?  Aluminum, as a
reflecting layer has been found to oxidize by may manufacturers and
therefore some have gone to trying platinum and gold -- good choices,
but rather expensive.  When they finally get the physical and chemical
bugs ironed out, there is still the problem of playback equipment. About
twenty different disks have been made in the last ten years. No
manufacturer will guarantee that they will support a machine
indefinitely.  No one wants to support old technology at a loss.  True
disks (being digital) can be copied indefinitely, but this means that a
copying program must be initiated to handle the growing numbers of
disks.  This will no doubt snowball out of control in the decades to
come. THis will also require replacing all playback machines every 10
years or so. Copying must also be done soon enough that the format of
disks can be updated before the support for the old machine is lost.
THis may sound like a rather dumb comment, but if OD becomes as popular
as film, (and it may), there will come a time when copying must be
started as soon as the new machine has been bought or support will be
gone before copying of all disks has been completed. In the case of
film, all that is needed is a lens system and a light. A chip that is
currently made (for an OD player) may cost pennies or less, but once the
chip is not manufactured, it could cost 10's or 100's of thousands of
dollars to make ONE chip.

To put things into further perspective, ANSI (American National
Standards Institute) has started groups looking into optical disk and
magnetic tape standards for permanence.  The photo people have always
looked at "archival" as at least several hundred years.  The optical
disk people said that archival was 15 years and the tape people claimed
that archival was 2 years. As a result, the permanence standards
committee (IT9) decided to discontinue the use of the term "archival".

Both media have their uses.  Certainly optical disk systems are great
for cataloging and tape is good for short term use.  It's easy to edit
and doesn't need time for processing.  However, don't believe that
either will provide "archival" copies that will last forever.


                   Conservation DistList Instance 5:2
                  Distributed: Saturday, May 25, 1991
                        Message Id: cdl-5-2-001
Received on Sunday, 19 May, 1991

[Search all CoOL documents]