1/27/06 How long do burned CD-Rs and CD-RWs last?

Question:

I recently read an article by a data storage expert who claimed that burned CD-Rs and CD-RWs can be expected to last only two to five years and not a whole lot more. I personally have commercially pressed CDs from the 1980s that still play fine, but I have begun to notice that some of my burned CD-Rs are beginning to skip, or not start (player shows "no disc"), or have a strange echo that was not on the original. This sounds serious! The expert suggests that for secure long-term storage, high -quality magnetic tape is the way to go. Are any of your readers beginning to notice this problem with their burned CDs, and are there other opinions about how to combat it? Are some burnable CDs of higher quality than others? What are the best storage methods for the discs that will make them last longest?

Submitted by: Carl N. of Cottonwood, Arizona

***********************************************************************

Answer:

Factory-pressed CDs are totally different from recordable CDs. In a pressed CD, the data is literally "molded into" (actually pressed into) the media and will not disappear unless the CD is physically damaged. Recordable CDs use a dye that changes color or reflectivity when heated. There are different dye types commonly used in recordable CDs--phthalocyanine, azo, and cyanine, in particular--and they do not all have the same life expectancy and stability.

All of the studies that I have seen except one suggest that properly burned one-time media (-R media, but not -RW media; see below) has an expected life of decades to possibly even centuries. There was a study by NIST (a U.S. government agency, used to be the National Bureau of Standards) on the relative stability of different media here:

http://www.itl.nist.gov/div895/gipwog/StabilityStudy.pdf

You can see some comparisons in the NIST study of the different dye types. But this study did not attempt to extrapolate the data to a life expectancy, although it did provide data about the relative stability of the different dyes and reflection layers behind them.

However, opinions still differ as to how long such media will last. The OSTA (Optical Storage Technology Association), in a report here:

http://www.osta.org/technology/cdqa13.htm

suggests that optical recordable media will last 50 to 200 years. This observation is backed by quite a number of studies that I have seen done both by the media makers and others. However, some storage experts suggest numbers more in line with your question, for example the expert in this report suggests a life of only 2 to 5 years:

http://blog.eogn.com/eastmans_online_genealogy/2006/01/life_expectancy.html (I have a suspicion that this is the article that you read).

The bottom line is that you are not going to get one single answer that everyone agrees on, although I personally am confident that properly recorded CD-R media can last decades if not a century or two. These 3 articles provide a good starting point for understanding some of the variables involved, which include:

-Dye type
-Physical construction of the media
-Storage conditions (temperature, humidity, light exposure, mechanical stress, chemical exposure and air quality)
-Manufacturing conditions (can vary from batch to batch in otherwise identical media of the same brand)

Now let’s mention some other things that are relevant and important:

-The quality of the burner. A borderline defective burner can "under expose" the media to the laser beam, producing a seemingly good recording (at the time of burning) that will "fade" over time (failing weeks, months, years or decades sooner than it should have had the laser beam intensity been correct)
-Recording speed. Fast burns (52X) are probably less stable than somewhat slower burns (say 16x to 32x), but you can burn media too slowly also. There is a very good analogy here to photographic film and exposure levels. The dyes on a given media have a certain range of acceptable "exposures" and outside of that range, you can either under or over expose the media to the laser beam. However, mechanical jitter and certain other variables (largely a function of the quality of the drive) generally will be unconditionally worse at faster speeds.
-Your own handling and storage practices. On a CD, the data "exists" in a dye layer on the label side of the media. This can be scratched from the back (from the label side), which will literally and directly destroy the data. The front side is clear plastic but can also be scratched. While front side damage may make the data less readable or completely unreadable, the data is still intact and undamaged on the label side, and the scratches on the front can normally be removed by polishing the plastic. On recordable DVDs, the data is on a layer "inside" the media, but the media is a laminate of several layers and can delaminate, destroying the data. Flexing – even VERY minor flexing – is particularly bad at causing such damage. And, also, recordable DVDs tend to fail from the outside in, so you can increase your success rate and decrease the incidence of failures by not recording such media beyond 80% to 90% of capacity, leaving the outside edge, where the failure rate is greatest and failure occurs first, blank anyway.

-Labeling: The glues in adhesive labels, or the solvents in pen-type markers, both applied to the label side (the side containing the data) can SLOWLY penetrate the reflective backing and dye layers and destroy the data. Therefore, for archival media, the safest policy is to not label the CD or DVD itself at all. If you do label it, with either a label or a pen, you are, at best, taking a chance with your data (hint: it is safe to write on the clear inner hub (where there is no data at all) with a suitable pen that won’t rub off).

And, finally, I would be remiss if I did not mention one other factor which is really huge: Eraseable "RW" media is FAR less stable than one-time ("R") media and should absolutely not be used for any permanent recordings of any kind whatsoever. There is no question that RW media can and does "fade". Although I’ve never seen failure of "R" media that I could attribute with absolute certainty to dye instability, I routinely see "RW" recordings that are unreadable after periods of months to a year or two when there is really no other explanation for the failure. I see this both on CD-RW and DVD+/-RW media, and I advise people in the strongest possible terms not to use "RW" media for anything that they want to consider permanent. Since RW media is also both more expensive (a lot more expensive) and slower, from my perspective the decision to never even buy RW media at all is an easy one from my perspective.

Submitted by: Barry W. of North Canton, OH