Magnetic tape deterioration

Jan Lyall writes:

>We are about to commence a cooperative research project between the
>National Library of Australia and the National Film and Sound Archive to
>investigate the deterioration of magnetic tape.
>
>In our Oral History collection we have a large number of magnetic
>reel-to-reel tapes that are exhibiting symptoms of the so-called
>'hydrolysis' problem.  When tapes suffering from this problem are played
>the sound is distorted, a high pitched squeal is emitted and the player
>eventually jams because the tape has exuded gummy degradation products.
>'Baking' the tapes in a warm, dry oven is known to make most tapes
>playable, however, this has not always worked on all tapes we have tried
>to restore, and we are concerned that the baking process may result in
>the distortion of the sound.  (All the research reports we have located
>deal with digital tape--all our tapes are analogue.)

My general attitude is that unless the chemistry and physical properties
of tapes change significantly, they won't last regardless of digital or
analog format.

I've written about tapes before, so I apologize to people who are
getting this as a re-run (I realize that winter is supposed to be a new
tv season with new programs....but.....)

You are apparently already familiar with the hydrolysis of the binder.

To start with, the magnetic "image bearing" material has been getting
generally less and less stable as information density has increased.
This is generally true of all information media and not just restricted
to tape. Consider the stability of cave drawings.  Low information
density and not very portable, but pretty darn stable.  We then had
writing on stone, followed by skins, rag paper, tree paper, still
pictures, b&w movies, sound movies, color movies, and tape.

Anyway, tape can generally be considered, like film, to have three
parts, the imaging material, the binder (that holds the imaging material
to the base) and the base.  In the old days, the magnetic particles were
gamma-ferric oxide--a version of rust.  Those of us in the NE US (and
Canada) know that once rust forms on cars, it doesn't change very much.
Along came cobalt-gamma-ferric oxide which know one was very sure about,
although it seemed to be pretty darn stable too.  Eventually we got to
metal particle tapes--tapes in which the magnetic media consists of fine
particles of metallic iron.  We all know how that should fair.  In
general, this has been one of the many fights that the manufacturers
have waged to produce a reasonably stable tape--how to prevent the
oxidation of the metallic iron.  On the plus side, the particles are in
a binder layer and are thus somewhat protected.  On the other hand, the
surface to volume ratio of the particles is pretty high and tends to
promote the disintegration process.  (There's a large surface area for
moisture and oxidants to attack versus the actual amount of iron
present.) Anti-oxidants and other mystery additives have been added to
try to reduce this problem.

The second part is the binder, and as you know, this doesn't stand up to
humidity very well, but is somewhat reversible.  The chemistry at least
is reversible in a dry environment, but once the binder has softened
enough that particles either move or come right off the base,
re-hardening the binder by reverse hydrolysis won't get the image or
sound back.

The third part is the polyester base (although in the 60's and I believe
part of the 70's, acetate was used as a base.  This base, of course,
suffers from vinegar syndrome possibly even worse than acetate film
does.) The polyester is pretty darn stable.

That takes care of the chemistry, now comes the physical interactions.
Surface friction seems to be one problem.  I heard a presentation by one
of the manufacturers about steps that they have taken to try to reduce
the lap-to-lap friction in order to improve the stability of the tape.
Problem two, as mentioned above, is the true reversibility of binder
hydrolysis. Whether the tape is digital or analog, once the magnetic
information has moved or been distorted, there is no chemical or
physical way of bringing things back.  Remember that a recorded tape is
an ordered system and spontaneous reactions have a tendency of
increasing entropy and decreasing order. Problem three is the physical
property of the polyester base.  Like many polymers, it doesn't like to
be under stress and undergoes cold flow to relieve that stress.

The thin polyester binder also makes tape reasonably prone to magnetic
print-through (something that most people have heard with old audio
tapes.) Each magnetic particle is a magnetic unto itself.  Each particle
can also be influenced by magnetic fields.  Thus the particle in each
lap of tape can act as a magnetic to change the information on the next
lap of tape. We hear this as ghost music before a tape actually starts.
It will also affect video tape, causing ghosts and sound prior to the
actual start of the tape, but I have to admit that I have never seen it
before.

The thin polyester is also not very stiff and provides little vertical
support.  This is the cause of edge damage.  Tracking or sensory
information may be lost due to edge damage.  I'm not exactly sure
whether the actual cause is physical stress of the binder (localized
flow problems) or stress demagnetization.  The old recommendation was
for people to retension their tapes every year or so, but I understand
from Ellen McCrady that there is at least one standards committee that
are not recommending such activity.   By re-tensioning, the tape doesn't
become loose and sag (causing the edge damage.)  Re-tensioning was also
supposed to move the relative positioning of the tape, thus reducing
magnetic print-through. One video tape company has recommending leaving
the tape at the far end (not rewinding after viewing) therefore putting
the print-through at the far end of the tape where it is less annoying.

The problem with re-tensioning is that the base, as mentioned above,
hates being under the tension and therefore undergoes cold flow.  This
means that the tape starts to sag and loosen, so the tape is
retensioned.  Which means that it's under stress so it flows, sags and
loosens.  So the tape is retensioned.   And so on and so on.  Of course
you can see where this is going.  Your tape may have started life as X
feet long, but eventually becomes X + Y feet long (meters for SI
people.)  Even if nothing else has happened to the tape, magnetic
particles that were a micrometer apart at the time of recording are now
more that a micrometer apart.  Whether the image is digital or analog,
this movement of relative particle location will again cause loss of
information.  The only advantage that digital tapes have is that they
can be copied (in theory) an infinite number of generations without
information loss.

As I have mentioned before, there is a running joke in the image
permanence field about couples who video tape their weddings.  In this
day and age, marriages at least, are deteriorating faster than the tape
images (estimated to last about 15 years or so I might add.)

-Doug

                                  ***
                  Conservation DistList Instance 7:45
                Distributed: Thursday, December 16, 1993
                        Message Id: cdl-7-45-002
                                  ***