Some Observations On The Longevity of CD-Rs

There is often discussion and speculation about how reliable a storage medium CD-Rs are and whether they have anywhere near the longevity of tape, hopefully I can provide some evidence either way, at least of their short term reliability.

In the late nineties I had a tiny label and put out half a dozen CDs, as well as having various CD-R masters for the releases I also have discs containing lots of other music I was interested in releasing, I have just retrieved around forty or so CD-Rs burnt 1998-99 on various drives and media and thought that it would be interesting to see how well they have aged as I back them all up.

The discs fall into two categories, those I burned using a Marantz CDR-630 and those the artists burned on unknown burners, the media ranges from generic no brand blanks to TDKs, Kodaks and Sony Pros.

All the discs have been stored, upright in their cases, inside a wooden crate at room temperature for the last ten years.

I will use EAC and my Asus rom drive to rip them and report any errors, hopefully some of you may find this interesting.

The first disc is a gold coloured Sony Pro, it is 74 minutes long and was burned on an unknown computer CD burner in 1998, the disc ripped flawlessly reaching 26x with zero errors, so 12 years and counting for Sony Pro discs.

The second disc is a TDK Reflex CDR 74, (I think this was either the bottom or middle of the 1998 TDK line), it was burned on January 17th 1999 on the Marantz CDR-630 and again EAC ripped it flawlessly reaching the same 26x speed, so over 11 years for the TDK.

I'll post some more results as I find the time to rip the discs.

 

I have found that the quality of the burner, and the speed at which the burn was done, make all the difference in the world. The quality of the CD-R may help, but I have not noticed anything definite. I will say that virtually every single Kodak and Mitsui have held up.

 

I have dozens of CD-R's (copies of box sets I made when I was worried that taking the original discs in and out of paper sleeves might scratch them), most made 8-10 years ago, all different brands of CD-R, and for most of the last four years stored in boxes in a basement, and I've noticed hardly any problems with them. Some would not play back perfectly on my PS3, but I suspect that had something to do with compatibility with the player, as they still all play fine on my Pioneer ca. 2004 "universal" player, the DV 47ai. Almost all of these CD-R's were made with an HK CDR-2 burned at 4x.

I did recently notice that the CD-R's will not rip accurately with dBpoweramp; I suspect it is because there is some virtually indistinguishable difference between original and copy that has nothing to do with potential degradation but rather how they were made.

 

Here's the next couple, I would do it faster, but I'm supposed to be working and I'd get confused doing more than a couple at a time.

BASF Extra 80 Min., burned 1998, unknown burner, ripped flawlessly, got up to 24x and would have probably gone higher, but was shorter than the previous two discs.

Kodak ds (Digital Science), 74 min 1x-6x, recorded at the end of 1998 on the Marantz CDR-630 this ripped without errors, but was very slow reaching 8.8x maximum, I'm not sure why it was so slow and was expecting EAC to detect errors, so the Kodak appears to be perfect, but I have doubts about it.

 

The next two I had low expectations of.

The first is a Jessops own brand 74 minute disc, (Jessops are the leading UK photographic chain), the date is '98 or '99 and the burner is unknown, surprisingly this ripped flawlessly and reached a speed of 22 x which considering it only contained 5 songs was very good, whoever made these discs for Jessops must have been a good company, the dye reminds me of TDK.

The next disc is a plain silver disc, I think it's probably from Lead Data, a Taiwanese supplier, it was burned in September '99, I'm not sure about the burner. This disc behaved like the Kodak, no errors, but a very slow rip reaching a maximum of 8.8 x, other than the slow rip it appears to have nothing in common with the Kodak and uses a different colour dye, being a cheap brand, the slow rip didn't really surprise me and it's reassuring that there were no errors.

 

I've ripped another dozen TDKs and all of them performed as well as the first, interestingly one of them had a paper label stuck over the entire surface and it appears to have had no ill effects whatsoever.

So far I'm really pleased, and slightly relieved, that a random selection of 11 or 12 year old CD-Rs burned on different machines have all been perfect, perhaps the naysayers are wrong, I'll rip the rest tomorrow, but it's looking like all the discs may be 100% error free which will be very impressive and also slightly surprising.

 

Any idea as to whether or not consumer grade DVD-R's would be prone to the same kind of longevity? I've wanted to archive some family movies and, seeing as I lack a mini-DV camcorder to use as a data storage system, DVD-R's would be the best "solid state" storage system at my disposal for the uncompressed video files. Despite not having a real option, I don't want to use them if they will degrade in a five-year-or-less time span.

 

I've now finished this particular batch of "old" CDRs, I've done some more TDKs, they performed as the others, several NX'R' brand 74 minute CDRs, they performed as well or better than the TDKs, some Nugen Media and a few other non branded discs, of the forty or so discs there was only one that has caused any problems, (it was full of errors and ripped at less than 1x), it was an unbranded silver topped 74 minute disc, I guess the artist who gave it to me used the cheapest media and burnt it as fast as possible in a computer. The disc is from 1998 and interestingly has a discoloured surface, the silver has darkened, I guess that this is not coincidental with regard it's performance.

I'm rather pleased with these results, if a random batch of 11 or 12 year old CDRs have a failure rate of around 2%, and even that disc was recoverable, then I have every faith that the mainly TY blanks I have been using for the last ten years will be equally as good, if not better, so personally I have no fears for my music.

 

I've found DVD-R/DVD+Rs to be much less reliable, especially non TY based ones, for home videos I burn at least two copies to disc, (use different brands), copy the external hard drives and also keep the original tape, I also keep a minimum of one copy off site, you can't be too paranoid about home movies or photos come to that.

 

I've never had any problems with my CDRs yet, all of them still play perfectly. The oldest are 8 years old, I came to CDRs quite late. DVDRs have a lot more problems in my DVD player.

 

Most of my older CD-R's, burned before 2000, still play perfectly fine today. Some of these were mixtapes, and some were backup copies of CDs. Saying that, most of them were basically stored away from anything that hurts CDs. They were never exposed to sunlight, always properly stored in a jewel case, and played sparingly since first made.

Here is an interesting, if outdated, link on CD-R longevity:

http://www.mscience.com/longev.html

 

This was my first thought!

 

I guess I was lucky that just by chance the first blank CDRs I came across were Taiyo Yudens. Back in the '90s I didn't really know much about the actual CDR manufacturers so I just went on brand name and fortunately the TDK and Maxell CDRs I used were TYs. Needless to say we went through a lot of them at our studio and I still have masters, rough mixes, tracking sessions, etc. of many of the better sessions. Of the CDRs made in 1994-1996 the majority of them still play and rip fine although some show C2 errors. I also have a bunch of what I would consider "no-name" brand CDRs (Datasafe Media) from the late '90s and those also still play fine. I actually find more audible errors with some of the more recent batches of midline CDRs.

 

I got my first computer with a CD burner in Aug of 2000. I started putting photo's on CD soon after. So far every disc I have gone back to has been as far as I can tell perfect.

I also store my CD's in jewel cases vertically.

 

I have many Sony ones that are DEAD.
I could salvage some data with CD Roller, but, only so much of it...
I am not a believer of cd-rs being a good long term storage solution...sorry, have had too many of them go "croak"...

 

I think I got ya all beat. I have at least a dozen CD-R's from 1994, when I first bought a burner (I think it was a Pioneer DVR-102, for $2000). Most of them work, but a few of the Mitsuis and Taiyo-Yudens have evaporated. The rest are hit and miss -- some survive fine after 16 years, but others have moderate-to-severe data problems. All were stored in jewel boxes, in drawers, in an air-conditioned house.

I am absolutely unconvinced that optical media is a reliable format. Note, BTW, that Nero can tell you the actual name of the blank disk manufacturer. TDK, Sony, and most of the other name brands have not actually made their own blanks for many years now. Mitsui/MAM and Taiyo-Yuden are still legit.

 

Other than proper storage, in cases, away from sunlight, etc., I'm wondering if the biggest threat is climate, we don't exactly get extreme heat or cold here, or humidity come to that, I know you have air conditioning, but what's the climate like where you are?

I rarely use Nero and had forgotten it would ID the disc manufacturer, I should have checked, maybe next time.

 

Interesting thread. I used to put adhesive labels on the CD-Rs i burned and most of them started to skip on my BLU Ray player. Alternatviely, the CD-Rs that I didn't use the labels but labelled them with a sharpy are still going strong. Needless to say, I had to reburn all of the labelled CD-Rs being grateful that the LG burners I was using was able to retrieve almost 99.99% of all songs ripped. Lesson learned. Stay away from adhesive labels.

 

Are you using secure mode, test and copy, those speeds seem fast

 

I'm using secure mode to rip, but just copying, not testing, I was surprised by how fast they ripped, actually faster than most commercial CDs, but I am using a pretty decent read only drive to rip them.

 

On that basis I wouldn't be convinced I think you need the CRC's to show any possible data error

 

I have about 12 Kodak CD-Rs that were burned in 1996 that still play fine today. I've long since backed them up to hard disk, but I test them every so often just to see.

 

As I understand things, and I wouldn't claim to be an EAC expert, Test & Copy mode is not as accurate as Secure Mode and EAC suggest you use Secure Mode if you get any problems using T & C mode, or am I misunderstanding the fundamental differences between the two approaches to ripping?

 

I'd just add that the discs ripped without registering a single data or sync error, maybe there is some truth in the old story about CDRs being more accurate than pressed CDs and easier for drives to read?