[78-L] "archival" CDs?

[Antony Pepper]

Good practical advice.

Personally, I would recommend laser etching all *really important*
archived digits into granite slabs as 2D (matrix) barcodes... This
will enable the information to be recovered quite easily by equipment
yet to be designed, for probably many thousands - maybe millions - of
years to come.

Do make at least two sets, stored in different locations, each
containing at least 100% parity/redundancy info (that way the data can
be recovered from just half the total number of slabs in each set -
any half).

If that's not possible for you, you could try inkjeting the same
barcodes onto rag paper. Those sheets or rolls will be okay for least
100 years in good storage, probably a lot longer, and can be scanned
back in by whatever optical device is then available without any great
difficulty. You'll need a lot of paper though.

Do include a human-readable definition of the data format!

If you want ongoing access to the data, not just to store it,
practically speaking you need a copy on a live array of come sort.

Best,
A


>
>Hi Rodger,
>
>As others have already pointed out, perhaps the greater risk is
>obsolescence of the CD format rather than deterioration of the
>media.
>
>Once you are in the digital world, you need to commit to data
>migration in order to keep your digital audio in a current
>format.  Figure on a major migration every 10 years.  Fortunately
>the migration process can usually be automated if the files are
>on a server.  At some point you'll have to ingest all those
>gold CD-Rs - long before their supposed 300-year lifespan.
>
>Mitsui/MAM-A markets their discs very well.  The fact of the
>matter is that the dye layer (pthalocyanine) is still organic.
>And organic materials have a finite life.  Unlike shellacs and
>LPs, you can't put CDs on a shelf for a few centuries and expect
>them to be usable - even the gold CDs.
>
>So store your discs dark, cool, and dry if you want maximum
>life.  You will still need to check the data on the CDs every
>few years to make sure there's no corruption.
>
>Assuming that you are saving WAVE data files, I strongly
>recommend generating MD5 checksum files (a type of digital
>signature) for each digital file.  This will allow you to test
>the files for data corruption.
>
>See
>
>http://www.theaudioarchive.com/TAA_Resources_FTP.htm#Checksum_Tutorial
>
>to learn more about checksums.
>
>If your university has an IT department, you may want to speak
>with them about their resources for long-term archiving of data
>if you have a large number of digital audio files.
>
>Hope this helps!
>
>Eric
>
>
>-----Original Message-----
>From: 78-l-bounces at klickitat.78online.com
>[mailto:78-l-bounces at klickitat.78online.com]On Behalf Of Rodger Holtin
>Sent: Thursday, January 29, 2009 1:38 PM
>To: 78-List
>Subject: [78-L] "archival" CDs?
>
>
>I've recently transferred some reels to digital for the university where I
>work and the archivist says she wants them on GOLD CDs as they are "archival
>quality."
> 
>Since they don't carry those at CheapIsUs Office Supply or Radio Shark, this
>is new to me.
> 
>True?
>
>Rodger
>
>For Best Results use Victor Needles.
>
>.
>
>
>
>_______________________________________________
>78-L mailing list
>78-L at klickitat.78online.com
>http://klickitat.78online.com/mailman/listinfo/78-l
>
>_______________________________________________
>78-L mailing list
>78-L at klickitat.78online.com
>http://klickitat.78online.com/mailman/listinfo/78-l