Langa Letter: Time To Check Your CD-R's

Almost three years ago, in "Is Your Data Disappearing?" we discussed the likely lifespan of burn-it-yourself CD-Rs. It's an important consideration when you're using CDs to archive data for long-term storage. At that time, even the cheapest CD blanks were thought to be good for at least 10 years after burning, with some premium rewriteable CR-ROMs possibly good for as long as a century.

Those estimates were based on accelerated aging tests performed on the various dyes used in CD-Rs or CD-RWs: The dye layer is what actually contains the data in a CD, and also is what gives the disk its characteristic color--blue, silver, green, etc. You can find abundant technical details in the original article or in Andy McFadden's ever-excellent CD-Recordable FAQ, especially in the section called "How long do CD-Rs and CD-RWs last?"

But the conventional wisdom about CD-R longevity was called into question several months ago with a series of tests originally presented in the Dutch PC-Active magazine and widely recirculated on English-language Web sites, including the Register and Slashdot. Those tests suggest that many CDs may fail in as little as two years! Of course, accelerated aging tests--while normally quite reliable--are only a simulation of aging. There's no way that a carefully controlled laboratory test will precisely reproduce every random aspect of actual aging in real-life conditions. But two years seems very, very short, unless the storage conditions were poor. And some environmental conditions can be quite unexpectedly harsh on CDs. For example, way back in 2001, we reported on the existence of certain special fungi that actually can consume the dye, foil and substrate in some CD types! An article in the science journal Nature had said:

After visiting Belize in Central America, Victor Cardenes of Madrid's National Museum of Natural Sciences, found one of his CDs discoloured, transparent and unreadable.... The disk's aluminium and polycarbonate layers were riddled with fungus, Cardenes and his colleagues have discovered ... Burrowing in like worms from the side of the disk, "the fungus destroyed crucial information pits", says team-member Javier Garcia-Guinea. Pits in a CD's aluminium and polycarbonate sandwich store binary data, which is read by a laser ..." (Full story: Fungus Eats CD)

But premature failures like that seemed like a fairly low-probability issue--how many people expose their CDs to rain forest fungi?--until the PC-Active article. That article sent many of your fellow readers (and me!) back into our libraries of old CDs to see how they're holding up.

Failure Patterns

This letter serves as an excellent example of what many readers found. More than that, it also offers a great tip for an easy way to test old CDs you may have:

Fred: Just had an eye-opener recently. We tried to restore some files from a [Norton] Ghost image that had been burned to CD-R disks about 2 years ago. The CD-R disks were unreadable in every drive we tried them in. I searched for some way of testing old CD-R disks short of copying them to the hard drive and waiting for an error message and found that the utility included with Nero is fairly good. It is part of the "Nero Toolkit" and is called CDSpeed.exe. I also found that I can run this on a machine without installing Nero, but I suspect that wnaspi32.dll must be present somewhere for this to work. What I have found is that a lot of our old "backup" ghost images burned to no-name CD-R blanks are now virtually unreadable. The really old gold disks with the green chemical have so far been OK, and the deep blue Verbatim Datalife Plus CD's are also OK. The ones that are failing have been 2- to 3-years old and have a silver reflective layer with a light green chemical coating. You might want to alert your readers to test some of their old backups to make sure they are still good. All of these ghost images were recorded with Nero with the verify option on, so they were all fine the day they were recorded. -- Tom Alverson

Thanks, Tom. You can get CDSpeed here here, and an ASPI driver, if needed, from the Nero Support/Utilities page.

The only variable that Tom mentions is dye type, and as discussed previously, different dyes do yield different longevities--though none is known to last only two years in normal storage. Might there be other factors at play as well?

That was certainly the case in my tests: I did discover bad CDs, but not one--zero--was traceable to a specific brand or dye type. Instead, for me, the only disks that failed were those that I had covered with a glue-on paper label from a print-it-yourself CD label-maker kit. Several of these labeled CDs were, in fact, totally unreadable by every means I tried (including special data-recovery software); all the data on those CDs is just plain gone. My best guess is the glue on these do-it-yourself labels interacted with either the foil or the dye, rendering the CDs useless.

Fortunately, although I have more than 1,000 data CDs in storage, only a handful have these labels. All my other CDs were hand labeled using an ordinary permanent-ink felt-tip marker, and all of these that I tested--all of them, including the very first CD-R in my collection--remain intact and fully readable. Plus, because I make multiple backups of important files http://www.langa.com/backups/backups.htm , no essential data was lost (what was missing on one CD was available on another.)

I'd used two different brands of CD label-makers in the past, and (alas) have no way of knowing whether only one or both caused the problem. But the bottom line is that while some CDs with glue-on labels had problems, none--not one--marked with a permanent ink felt-tip pen did.

Recommendations

To me, the above test suggests a two-step strategy to achieve long life for your CDRs and CD-RWs. First, pick a brand or dye type known for durability. (See our original article Is Your Data Disappearing? for a quick summary, and follow the links in the article for detailed information.)

Second, keep your CD-Rs clean and unmodified, and don't do anything that could damage or alter the top side of the disk, where the data is stored. As I painfully discovered, this probably includes not gluing labels to the CDs!

But there's more: In inexpensive blank CDs, the foil layer atop the CD is left totally exposed and vulnerable to physical damage. Tossing recorded CDs into a drawer where they'll slide around on top of each other or commingle with pens, letter openers, or other sharp implements, is asking for trouble. If you need repeated access to a CD you've burned, put it in a protective sleeve or jewel case. If you don't need ready access to the CD, put it on a spindle with other CDs. (This is how I store my backup CDs: I put them back on the same spindles the manufacturers use to ship the blanks.)

Beyond that, it's also smart to watch other environmental variables. For example, many accelerated aging tests use high heat, light, and humidity to speed oxidation and other chemical reactions. It stands to reason that you can retard aging by storing your CDs in a relatively cool, dark, and dry location. A closet or cabinet in an air-conditioned space is probably close to ideal.

If full climate control isn't possible, at least try to control humidity. All organisms (including fungi) need water to survive. Keeping your CDs dry will help prevent biological agents from causing trouble, and also will prevent unwanted interactions between water and the dyes and bonding agents used in CDs.

My guess is that by protecting your CDs from physical damage and environmental extremes, you'll probably achieve a very long data life indeed.

Optical Storage Technology Association

In researching this same issue--CD-R longevity--another reader found a wonderful resource:

Fred, I just happened onto a site that goes into great detail about CD-R and CD-RW technology and thought it might be of interest to you and your readers. It's the Optical Storage Technology Association, and it covers labeling, writing, longevity and much more. -- Bill Tone

Thanks, Bill. There's lots of good info on that site--including recommendations on print-it-yourself CD-R labels I wish I'd known about several years ago! Highly recommended.

Armed with the above info, I think most of us will be able to avoid premature failures in CD-R/CD-RW media. But don't make assumptions: It's wise to check your old CDs from time to time to make sure they're aging gracefully. With regular spot-checks, you may be able to catch degradation before it's gone too far, and then recopy the deteriorating CDs to fresh media.

There may be some cases where unusual circumstances may cause the disks to fail sooner than they otherwise would. But with a little care in selecting the right medium to begin with, and then further care in storage and handling, we should be able to rest easy about our archived data!

SHARE THIS ARTICLE

More