RE: [AV Media Matters] CD labelling-why not?

[e.ha.zwaneveld@nfb.ca]

Greetings,

In response to: Robert Brian Levy [mailto:xernaut@yahoo.com]
who asked:

I would appreciate a few more opinions about this idea
that it's best not to write on a CD at all, when that
CD is intended for archiving information for the long
haul.  Are even 'CD-safe' pens such as that made by
TDK, which those of us working with the Caddo language
and oral traditions project have been using, as well
as the other CD-safe pens mentioned in recent postings
to this list, are these kinds of pens also likely to
compromise the life of the CD?  If so, why?

Thanks for the on-going discussions.

             Brian Levy

First I would like to comment that I feel the advise of David Seubert was
very sound, given the intent of keeping these CD media as long as possible.
I also would keep an 'archival' disc in storage unused, and another copy of
the same for consultation.

As to why?

Well, the label surface of CD-R media is the most vulnerable side, having
been coated with a 'protective' acrylate plastic or epoxy coating that is
supposed to protect the metal reflective layer under it from scratches,
oxidation and contamination. That is why additional hard coating is such a
sound option. As already implied, the so-called typical protective coating
(some CD-R media don't have a protective coating at all) is minimal
protection. DVD media are protected by an adhesive layer (or dual space
layer) and substrates.

The CD industry has had experience with reactive label inks, usually
alcohol-based, which ate their way in through the protective layer and
oxidized the metallic layer. These chemical reactions are referred to as
"growing" in time, so what seems insignificant when the disc is new may grow
worse. That is why any pen used to mark a CD-R should have ink based on
water and be recommended by the disc vendor. In addition, because the
typical protective layer is so vulnerable, the felt pen tip must be soft, so
it does not engrave the writing. When that happens you can see it from the
disc reading side, it does not take much to penetrate through the lacquer.
That is how a video manufacturer friend of mine, could not play his CD-R
discs the day after he wrote them, I have seen them. To avoid accidental
content loss, I believe that it is a sound policy to use the unique serial
number on the disc to identify it in your database, rather than writing or
worse, attaching labels or stickers to the disc and cause instability.
Again, it is amazing how easy it is to cause disc unbalance.

Risk avoidance when the disc is new can multiply the useful lifetime of its
content, if the new disc has been proven by independent testing to be
suitable for the intended purpose. Given the highly competitive marketplace
and great pressure on the manufacturer's profit margins, if any, such
qualification testing up front is a good policy, as many failure mechanisms
are already built into (or are already out of tolerance) inferior new discs.
You may find yourself re-transferring your' archival' discs forever i.e
every few years (assuming you catch their maximum tolerable error rate
before they refuse to play back any signal at all).

Finally, there was someone who asked about writing more data on discs than
indicated on the package,such as going past the 63 or 74 minutes capacity.
Keep in mind, that the lower the capacity of a disc, the easier it is for
the Error Correction circuits to correct larger defects. Given that a good,
new CD already has an average of 100,000 errors when new, (out of some
19,300 million channel bits per disc) for the error detection and correction
circuits to chew on, why press your luck to save a few pennies? The error
rates of optical discs don't get better in time, they get worse, so their
life style from the moment of conception, as with people, does matter a lot.

Best regards,

Ed H. Zwaneveld
Technological Research and Development
National Film Board of Canada
May 4, 2001
Tel: +1-514-283-9143