I had an e-mail from a reader in which he revealed he'd had a conversation with technical service people at Maxell. This is what he wrote:-
Thought you might be interest in a conversation I had with the tecnhical service people at Maxell Corp, USA today. They were aware of the recent article on storage of data on CD-ROMS, but a bit surprised at the observations therein.
They have been running life expectancy (LE) tests on some of their products for a while now, and this one person I spoke with said that with proper storage (out of the light, relatively constant heat and humidity), they didn't think a 50 year (read fifty) LE was unreasonable.
He also suggested using a commercial grade CD-ROM (which has a much heavier coating to resist abrasion) if a CD is subject to frequent handling. It is the scratching that causes more problems than degradation of the dye substrate.
Then, he says, make a second, untouched copy, for an archive and, if properly stored, this should last well beyond 5 years.
Not so keen on tape
He also seemed much less enthusiastic of tape as a better archive medium, contrary to the IBM consultant's comments. Magnetic tape is in very close proximity to other layers of tape, he said, and are far more susceptible to "dropout" from interaction of the magnetic particles from one layer to the next over time, particular in unfavorable storage environments (high heat and humidity).
He said the whole LE issue may be moot in the near future as the media industry migrates to the latest technology, holographic recording. Although currently available for commercial use, it could be quite some before the technique is cost-effective for home PC's. But holograpy will make it possible to store far more data than even a DVD, and in a much smaller size medium, and with far greater stability than current CD technology. Which may also help explain why the industry hasn't made any great push to develop longer-life CD ROM media.
Another little tidbit that might be of interest to you (and your readers) has to do with labelling. Maxell said that using alcohol-based markers on CD-ROM's (like Sanford's felt-tipped "Sharpies") can have serious adverse consequences, as the liquid can actually bleed through from the label side to the data side if "puddling" is allowed to take place (by writing too slowly, perhaps).
He recommends using only special water-based markers for labeling CD's.
When asked about using printed, semi-circle adhesive paper labels to avoid the problem, he cautioned against that, as well. Quite often, they found that the slightest misplacement of the label or smallest curl at the edge of the label can be enough to throw the rapidly spinning disc off balance.
This can ruin either the CD, the drive, or both! His suggestion was to get a special printer that will print directly on blank CD's. If alcohol-based markers must be used, he suggested a catalogue method, with only the code numbers, not full descriptions, written on the CD label (e.g.CD labeled #1 would be catalgued as "family pictures, July 2005". CD with #2 would be "Cariibbean Cruise". etc.).
Hope this information is useful. At the very least, I think we are a little more at ease about the storage and labelling of our photograhic data on CD-ROM's, and think we are now comfortable in that regard without resorting to external tape storage.
Commmercial grade CDs
What an interesting e-mail. It makes the relative life expectation of burnt CDs versus tape clearer and is a valuable corrective to the IBM Germany statement.
A Google search on commercial grade CDs revealed brands such as USDM, Acme, Taiyo Yuden and Verbatim. I'm investigating commercial grade recordable CD availability from Verbatim.