CD-R for long-term storage

There is a current article with interesting links about
this on /.:

http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus

From my experience: you get what you pay for. I've only had cheap white
disks (memorex) failed after a few years. Some blank media would not even
record after been stored for a year. Some DVD-R Memorex media also would not
record after a year.
Blue disks from good brands (TDK) are still OK.

I think, too, that phase-change media must be more stable.

"Arno Wagner" wrote in message
...
There is a current article with interesting links about
this on /.:


http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus

In article ,
says...
There is a current article with interesting links about
this on /.:

http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.


I've said it before, but it's also a smart idea to add
recovery data (e.g. PAR2s created by QuickPar) to any
sort of archival media. Gives you a bit more safety net
then the built-in error correction on the media, allows
you to verify the contents of the media, and you can
pick how paranoid you want to be. (For low-paranoia,
only create 5% recovery data, for high-paranoia create
100% recovery data and burn multiple copies.)

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.

Basically, use Mitsui Gold CD-R discs for the longest possible shelf life.

Otherwise, even the silver discs I have from 10 years ago are still
good. Only two out of thousands have gone bad in that time and the rest
of the discs are still readable w/o errors -- all stored in cold storage
w/o light exposure in a box.

---

http://www.silverace.com/dottyspotty/issue12.html

This issue contains advice on selecting quality CD-R discs to use when
recording/storing imporant data and photos for long-term storage.

0. Don't use CD-RW discs! They're only designed for short-term storage,
and are erasable. Don't go putting your imporant photos on these!

1. The dye used is probably the most important for long-term stability
and lifespan.

There are three dyes used today in CD-Rs, in order of preferece.

A. Phthalocyanine - inherently stable unlike cyanine dye which must
be stabilized by the addition of other metals, etc. Longest lasting
under testing. Expected lifespans exceeding 200+ years when coupled
with gold reflective layers in cold storage.

(color, light green on silver reflective surface; light
yellow/green on gold reflective surface)

B. Azo. Only Mitsubishi/Verbatim makes this. Up to 100 years,
less stable than Phthalocyanine dye, but more so than cyanine dye.

(color, very deep blue on silver reflective surface)

C. Cyanine dye. Chemically unstable alone, and must be stabilized.
Still, less stable than the other dyes in long-term simulation tests.
However, the first made consumer CD-Rs used this and is part of the
'standard' all CD-RW drives must be compatible with.

(color, green-blue/blue on silver)

2. You can use CDRIdentifier to read the dye information stored on the
CD-Rs you buy as well, which is more reliable if you don't know which
bottom colors represent which combinations.

CDRIdentifier: http://www.gum.de/it/download/english.htm

3. Besides dye, the reflective layer used affects long-term storage
lifespans.

There are three known types used today, in order of preference:
A. Gold - yep, expensive, but from the bottom, it'll look like real
gold.

B. Gold + Silver - Only Kodak Ultima Silver+Gold CD-Rs use this.
Their tests suggest this combination lasts longer than silver-looking
only discs.

C. silver. - not true silver, but silver looking. Most discs
produced today use this. However, given that metals except gold
corrode, corrosion of such surfaces (when the top laquer layer has been
removed/scratched) can and will occur.

4. Do not look at the top when trying to determing reflective layer
type! Look at the bottom in particular, the exposed areas around the
rim and inner hub. The fake 'gold' layer they put on top of some CD-Rs
are not true gold at all, and only there to 'fake' the customers.

5. The longest lasting CD-Rs, based on longevity testing, use
Phthalocyanine dye and Gold reflective layers. Phthalocyanine with
Gold&Silver reflective layers next, followed by Phthalocyanine and
Silver and Azo and Silver. Cyanine on anything (only silver today) is
the worst performing disc.

6. Typically, Japanese made CD-Rs tend have better quality control, IMO.
Once they move production to Taiwan/Mexico, forget em.

Top Quality Name-Brands, in order of preference:

A. Mitsui Gold CD-R & Kodak Gold Ultima - Phthalocyanine dye + Gold.
B. Kodak Gold Silver+Gold - Phthalocyanine dye + Silver & Gold.
C. Mitsui Silver, Ricoh Platinum - Phthalocyanine dye + Silver.
D. Mitsubishi/Verbatim - Azo + Silver.
E. Taiyo Yuden - Cyanine + Silver.

7. You can get these at www.cdrexpress.com and www.memorymedia.com.

Use www.silverace.com/smartpig/ to locate more places to buy, along with
online coupons, rebates, and ways to save money online.

8. All the other 'cheapies' not mentioned above are generally of poorer
& more variable quality. eg. those $15 / 100pk of no-name CD-Rs at
www.microcenter.com and www.compusa.com are the worst in long-term
longevity.

However, even the quality of cheapies today will last a few years
before deteriorating (assuming no long sunlight exposure; a few hours in
the sun will kill discs) and will work fine. Great for disposable
burns, music, and stuff you don't care if they fade away soon.

The poorest performing brand-name disc above, 6-#E, will still last
10-20 years in cold storage w/o any problems at all!

Remember to keep them away from sunlight/UV, and that most likely,
you'll be moving all of the data off CD-Rs onto newer storage medium
(maybe DVD-R? Holographic?) in 10-20 years. Why? Not one storage
medium has lived that long, and realistically, you'll want to move the
data off obsolete storage media =before= they stop making the drives!!

Just like nobody has 5 1/4" (or 8/12") floppy drives anymore,
expect only to need at the minimum discs that'll last until you migrate
the data off them in 10-20 years from now. Of course, higher quality,
longer lasting discs 6-#A-#D will only help your peace-of-mind.

9. Burn at least two copies of every important piece of data.
Preferably, to two different brands of discs, and preferably two
difference types (eg. Picking #A always is a good #1 pick, anything from
6-#B-#E as your #2). Although simulated longevity tests suggest #A has
the best lifepsan, nobody has ever sat around long enough yet to see if
that's true (in fact, CD-Rs have been out only about 10-15 years!).
Best to make sure you use two different brands and types to keep
bad-batches and combos of dyes/layers from ruining your data years from now.

10. Double-check all imporant discs after each burn - make sure they
match the original files 100% before putting them away and assuming
they're good burns. Use the CD-R program's 'verify-after-write'
feature, available in only some programs, or CDCheck:
http://fusion.zejn.si/ to do this.

Arno Wagner wrote in message
...

There is a current article with interesting links about this on /.:

http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it.

Stupid summary then.

Some testing with reasonably conditions (dark cupboard)
had unreadable media as soon as 2 years after writing.

And I havent lost a single one over a much longer period
than that, with no care taken at all, in fact some stored
in jewel cases where they get full sunlight in winter too.

The problem seems to be that durability varies widely and there
is no way for ordinary users to asess media and writing quality

So the obvious thing to do with data that matters is to write
it on more than one brand of media and do some basic error
checks to ensure that it is at least written properly.

(yes, a good CD-R medium can be badly written and die early...).

And its completely trivial to check if its been badly written.

Mentioned alternatives are CD-RW (not dye-based,
lifetime unknowen), DVD+/-R (wishful thinking IMO,
it has the same problems as CD-R),

And you've always had silly ideas about all these media.

MOD (phase-change, lifetime 50 years,

We dont know that yet.

but expensive),

Makes more sense to use multiple cheap approaches instead.

tape designed for long-term storage (lifetime ~30 years,

We dont know that yet either.

but this is not any of the ''cheap'' tapes like DAT) and
keeping multiple copies with constant checking and copying.

Which is much more viable with CD/DVD formats, stupid.

Arno Wagner wrote:

There is a current article with interesting links about
this on /.:

http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus


Hello, Arno:

Don't forget DVD-RAM, please. It's on a par with your precious MOD,
after all. :-P


Cordially,
John Turco

John Turco wrote:

Arno Wagner wrote:

There is a current article with interesting links about
this on /.:


http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus


Hello, Arno:

Don't forget DVD-RAM, please. It's on a par with your precious MOD,
after all. :-P

Uh, why would that be?



Cordially,
John Turco

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)

DVD-RAM is phase-change technology, and is likely equal to DVD-RW in
longevity.

"John Turco" wrote in message
...


Hello, Arno:

Don't forget DVD-RAM, please. It's on a par with your precious MOD,
after all. :-P


Cordially,
John Turco

Previously John Turco wrote:
Arno Wagner wrote:

There is a current article with interesting links about
this on /.:

http://slashdot.org/article.pl?sid=0...id=137&tid=198

The summary is essentially to forget it. Some testing with reasonably
conditions (dark cupboard) had unreadable media as soon as 2 years
after writing.

The problem seems to be that durability varies widely and there is no
way for ordinary users to asess media and writing quality (yes, a good
CD-R medium can be badly written and die early...). Mentioned
alternatives are CD-RW (not dye-based, lifetime unknowen), DVD+/-R
(wishful thinking IMO, it has the same problems as CD-R), MOD
(phase-change, lifetime 50 years, but expensive), tape designed for
long-term storage (lifetime ~30 years, but this is not any of the
''cheap'' tapes like DAT) and keeping multiple copies with constant
checking and copying.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus


Hello, Arno:

Don't forget DVD-RAM, please. It's on a par with your precious MOD,
after all. :-P

Well, yes. If you can get a drive that accepts the version with
cartridge... (Yes, I know. They are out there. Just difficult to
find. And my critical system backups still fit on a pair of 640MB
MOD's each.)

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus

Toshi1873 wrote:


I've said it before, but it's also a smart idea to add
recovery data (e.g. PAR2s created by QuickPar) to any
sort of archival media. Gives you a bit more safety net
then the built-in error correction on the media, allows
you to verify the contents of the media, and you can
pick how paranoid you want to be. (For low-paranoia,
only create 5% recovery data, for high-paranoia create
100% recovery data and burn multiple copies.)


Can QuickPar be really be that helpful for imaging software, though?
To me, it appears to be helpful when you have something that is split
into many smaller files. So if one file is missing or cannot be read,
QuickPar can reconstruct the missing data.

But what about the case where you have a single image file on a DVD-RW.
And part of the disc got a scratch on it. Can QuickPar help in such
a case?


-WD