Burned CDs Last 5 years Max -- Use Tape?

Lam1969 writes "Computerworld has interviewed Kurt Gerecke, an IBM storage expert and physicist who claims burned CDs only have a two to five-year lifespan, depending on the quality of the CD. From the article: "The problem is material degradation. Optical discs commonly used for burning, such as CD-R and CD-RW, have a recording surface consisting of a layer of dye that can be modified by heat to store data. The degradation process can result in the data 'shifting' on the surface and thus becoming unreadable to the laser beam." Gerecke recommends magnetic tapes to store pictures, videos and songs."

Conflicts with other studies

[Nom du Keyboard]

Haven't other studies confirmed much longer lifetimes in the past for CD-R? After all, we've had CD burners for longer than 2-5 years. Is this only a surprise because absolutely nobody has ever gone back and tried to read an old disc? Somehow I'm still doubtful of his conclusions.

5 years max?

[blanktek]

I have CDs that have lasted 10 years with no errors. Obviously 5 years is not the maximum life. Perhaps the maximum EXPECTED life.

Re:I guess it depends on how you treat them

[Jaysyn]

Ditto, but they are data CDs & pushing 7 years old. Only read problems are ones I've inflicted (scratches, etc)..

Jaysyn

CD-Rs with a 100 year warranty

[OverDrive33]

We've known that CD-Rs will degrade for a long time. Hispace [hispace.com] have recently launched a new range of CD-Rs aimed at digital photographers. These disks use 24 caret gold to help add stability to the disks. As a result, they come with a 100 year warranty.

Your porn will be around for decades after all!!

Re:5 years max?

[BushCheney08]

I have a few data CDs that I burned in 97/98 that I recently pulled out. Worked fine for me. Then again, they were burned with a good quality Yamaha drive. I'm sure some of the stuff that I've burned more recently with a cheapo LiteOn drive (which died just after a year of use) won't fare so well.

Old and incorrect news

[digidave]

Not only have the early deaths of CD-Rs been greatly exaggerated before, but even the lifespan of pressed CDs were (are?) hotly disputed. In the early 80's I heard all about how CDs would last forever because each play didn't degrade the quality ever so slightly like it did with cassette tapes and vinyl records. Then in the late 80's a group of researchers determined that CDs would probably only last ten years, for whatever reason.

I got my first CD-RW drive when it was a $700 2x model well over ten years ago. The first things I burned were a bootleg Tragically Hip CD and a few rented Playstation games. I still play that Hip CD and recently I dugg out my Playstation collection to use with the epsxe emulator and they all still work great, though I can't remember which of my burned games were copied when.

I have had a few CDs and DVDs go bad, but they've always been really cheap media. Even cheap CD-Rs have been ok, but I have noticed that cheap DVD-Rs can be very poor quality and sometimes the data won't last through the night. These are usually identifiable because at least half the time the data will be corrupt straight out of the burner. You don't have to spend a lot to get good media, just don't get the cheapest media you can find.

Re:I guess it depends on how you treat them

[letxa2000]

I have both audio and data CDs I burned way back in, what, 1996? They still work perfectly. The first audio CD I burned at that time has spent the last 10 years in my car in the heat of Mexico. Still works perfectly.

Comment removed

[account_deleted]

Comment removed based on user account deletion

NPR reported the solution nearly three years ago..

[Anonymous Coward]

From a story three years back on National Public Radio:

"The whole history of recorded sound has been a case of one technology leapfrogging over a previous one," Karr says. "But in the last few decades, the changes from vinyl to tape cassette to CD to MP3 have shortened the life span of most music collections."

But thanks to a grant from the Smolian-Giovannoni Foundation, all of these audio formats are being transferred onto 10-inch wide, 78 rpm shellac disks -- the one rock-solid format archivists have identified that works every time.

See complete article at Shellac, the Sound of the Future [npr.org]

Depends how you define lifetime

[Solandri]

Keep in mind that CDs have a ton of error-correction coding on them. You could lose probably 20%-30% of the bits and still have a working CD. I suspect by "lifetime" the guy means when dye layer starts to deteriorate. Error correction can get you past that point, but you end up with a CD that reads fine one month, then "suddenly" develops dozens of bad sectors.

Most serious photographers I know re-burn their archives every one or two years.

Re:I can attest to that...

[jridley]

I have had many problems with DVDs. I've had media that degrade quickly, and also writers that cause discs to degrade quickly. Every disc I wrote in late 2003 is bad. They started going bad after only a few months (in some cases, days).

I switched both burners and media and now have no problems. However, I still do a 100% verify, and don't totally trust DVD-R. For stuff I *really* want backed up, I put a PAR2 set on the disc, and I burn both DVD and at least one CD copy for offsite.

BTW I found that some really crappy DVD-ROM drives will read almost anything. All of those hundreds of bad discs that I have? I bought a shitty CompUSA DVD-ROM drive for $35, and it will read them all, even though NO other drive I own will read them (I tried Sony, 2 NEC, 1 Pioneer and 1 Lite-On DVD-R drives, plus Teac, Pioneer DVD-ROM drives). I have NO reasonable theory why this is, but the damn thing just reads anything. I'm glad of it too. I discovered this when I realized that my shitty $40 mintek set-top DVD player would play the discs and "better" players choked, so I decided to try a crappy DVD-ROM drive. So I can now make new copies of the messed-up discs.

Been there...

[Life700MB]

That's a well known idea, I was going to put here some samples of the distributed backup in action but only can find when Cringely talked about the very same concept [pbs.org].


--
Superb hosting [tinyurl.com] 20GB Storage, 1_TB_ bandwidth, ssh, $7.95

Re:Instead of tape why not drives for long term?

[Beryllium Sphere(tm)]

Experience with other mechanical objects stands up and asks a question.

Will the lubricant in the bearings go sticky if the drive is on a shelf and never spun up? Someone out there must have direct knowledge.

The issue is that most (most complicated, powered) machines with moving parts need occasional mild exercise.

Buy OEM

[jridley]

Don't buy name brand media. They're just rebranding, and they can switch suppliers on you.

Go right to the asian source, buy from a reputable importer. I use supermediastore.com, and I buy nothing but Taiyo Yuiden media.

The place where I work has a high-speed multispindle CD-R duplication station, and goes through CD-Rs by the thousands per month. I asked a while ago and they have tried everything, and now use NOTHING but Taiyo Yuiden media also. If they have a failure, we have to ship a replacement overnight; so a single disc failure, by the time you count all the people who have to handle a complaint and the postage, can easily cost $50, so they buy what WORKS.

Re:5 years max?

[gihan_ripper]

What does that mean? By 'maximum expected life' you surely mean the expected life of the medium, that is, the mean of the lifetimes of a good sample set of CDs. When a lifetime is quoted, e.g. for lightbulbs, the manufacturer doesn't guarantee that the product will fail when its expected lifetime expires!

Re:Backup media

[Anonymous Coward]

Anybody hear about MO (magneto optical) media? Pretty good technology. A 3.5 inch disk can store 2 GB and 5.25 inch one 9 GB. Plus their "Archival life" is 50 years. Says so on Fujitsu's web site. I think they are cool and should be more popular (would probably make it cheaper). Anybody interested look here: http://www.fcpa.com/products/mo-drives [fcpa.com]

Re:5 years max?

[drinkypoo]

Just to make your comment more explicit, hard drives are cheaper not only than the tape drives, but are also often cheaper than the tapes. You can get hard drives of fairly large capacity around fifty cents a gigabyte. A 200GB Sony AIT4 tape is at least $50. Granted, that's half the price of the hard drive, not a very good example! However, the drives start at $2400. That's a lot of hard drives! A 7-pack of 40GB DLT tapes (kind of useless) is $150 or so. That's around hard drive prices...

Re:Depends how you define lifetime

[Hatta]

Keep in mind that CDs have a ton of error-correction coding on them. You could lose probably 20%-30% of the bits and still have a working CD. I suspect by "lifetime" the guy means when dye layer starts to deteriorate. Error correction can get you past that point, but you end up with a CD that reads fine one month, then "suddenly" develops dozens of bad sectors.

Is there any software that allows you to check on the status of the dye layer? It would be good to know before hand that you're using nearly all the error correction on a disk so that you can replace it when you have the chance.

Re:CD-Rs with a 100 year warranty

[_Shorty-dammit]

so, how exactly do the sectors get remapped etc when they're sitting on the shelf nowhere near a computer? ;)

Ziplock bags don't breathe,...

[Futurepower(R)]

Tapes? Tape drives? Someone is trying to sell a stock of old drives, I'll bet. My $1,300 Seagate tape drive and expensive tapes were very unreliable compared to CDs and DVDs.

CDs and DVDs stored in ziplock bags seem to last a long time.

Changes in atmosperic pressure cause other methods of storage to breathe. Eventually pollution enters. Ziplock bags don't breathe, they just expand and contract as the weather changes.

Re:Conflicts with other studies

[sionki]

Problem with this theory is that the oxide on magnetic tape will degrade withing 3-5 years also. We have tried reloading data from tapes that old and sometimes it works, sometimes it doesn't. I always have to clean the drive afterwords though.. :-)

Re:I guess it depends on how you treat them

[Sen.NullProcPntr]

As an experiment I took a freshly burned brand name (is it live or...) CD-R in a thinline case and put it in an environmental chamber [tenney.com] (as a piggyback to normal testing) at work. The chamber routinely cycles between -10 and +70C with extended periods (>24 hours) at the extremes. After about three months the data was still good (md5sum /dev/cdrom matched and no read errors in /var/log/messages).

This was not a scientific test but it did give me more confidence in the media.

Re:If you say so...

[zoeblade]

Anyone know where I can download an MP3 jukebox for my Vic 20?

No, but there was a program listed in Zzap! 64 once that let you play audio tapes using your Commodore 64. Type in the program, press play on tape, turn your TV's volume up, and listen to something with slightly more signal than noise.

Re:I can attest to that...

[aonaran]

Usually the reason why crappy DVD-Video players play back things that don't play in other players isn't because the pickup is better, but because the decoder is more tolerant of non-standard discs. Some discs that play on PCs and these cheap players just won't play on the better ones because the disc itself is not made to the proper DVD spec. Most often I've seen either improperly encoded video or missing AUDIO_TS folders. Next to that is not having the files organized properly on the disc, or the wrong file system (ISO vs UDF)

Re:Use intelligent distributed databases

[burris]

> Peers download it and keep it on their machines.

No , they don't. Not unless they really want it.

> The P2P network automatically polls for chunks and ensure redundancy by
> pushing rare pieces to clients.

Like UseNet?

> Some sort of bittorrent expect it's rather a bitpool.

Good one! You realize that Bram was working on just such a distributed file store before he decided it was a rat-hole and quit? Then a bit later he wrote BitTorrent.

Re:Nothing lasts forever

[whitelabrat]

That's right. I've done quite a bit of research on this, and the fact is that the dye based CD should last about 70 years if care is taken with the storage. Interestingly enough the aluminum based cd's can run into corrosion problems that the dye based cd-r aren't subject to. Most DVD-R's seem to use a high quality dye too.

I think the main issues with cd/dvd recordable media are poor burning and storage. I don't get the part about magnetic media being better, when you consider that an unused VHS tape only has a 20year shelf life before going kaput. I've see a lot of tape based media where after aging a couple years the ferrous material starts flaking off. Hmmm.

Re:CD-Rs with a 100 year warranty

[Anonymous Coward]

Actually... from a forum I was reading recently... among others...

Tweertinelle
06-18-2002, 02:07 AM
As I understand it, Quantegy discs are reliable not only because they can be burned on computers as well as dedicated recorders, but also because they're said to outlast other brands (hence their common use by music professionals). The problem I've had with Sony, Maxell, Memorex and TDK is that discs can become unreadable after a few years. If you're archiving original material, that can be a problem.

I should also mention that Hi-Space gold CDs are absolutely worthless. Beautiful packaging, but absolutely unreliable media -- the ones I bought had visible imperfections on the CDs' surfaces. I'm glad those things aren't about any more to confuse people and pollute the market.

So you see, people have been complaining about the POOR quality of these discs for years.

So please, MOD PARENT DOWN.

Re:Depends how you define lifetime

[Jherek Carnelian]

Is there any software that allows you to check on the status of the dye layer? It would be good to know before hand that you're using nearly all the error correction on a disk so that you can replace it when you have the chance.

Yes. At least for DVDs. But you have to have a drive that supports it. Plextor burners typically come with software that can report error rates. Then there is KProbe [cdrlabs.com] for some other drives, read the docs for the details.

As a rule of thumb, when purchasing blank media, prefer "made in japan" over the others. But be careful, two otherwise identical looking spindles of blank media from the same brand may differ solely in the "made in japan" / "made in china" fine print.

Re:CD Presses

[Ced_Ex]

The problem is that pressing your own CDs is not cost effective in any way. It only becomes cost effective once you print many (1000 or more) identical discs. The costs per CD go down as you do more identical discs. Setting up the platter to stamp out just one disc would be extremely expensive, as well as very time consuming.

I was thinking more along the lines of someone inventing, or streamlining the process to make it cheaper.

For instance. Currently we're using lasers to burn out reflective pits on metallic dyes adhered to plastic discs.

For pressed CDs, we're not using dyes, but some metallic compound that is physically pressed into pits.

My idea would be to have the same or similar metallic compound as pressed disc, only use a stronger laser to burn out physical pits. This could be a commercial grade device used in professional shops and not for the average home consumer. Something like this could be viable could it not?

Re:Photography's loss

[timeOday]

I'll agree that the .jpg format may be replaced at somepoint, but how long would it take an automated tool to batch convert a few million .jpgs?

And I will disagree. Worrying about jpeg disappearing is no different than saying "don't waste your time writing in English, it'll go obsolete and humankind won't understand it anyways." There are billions of jpegs out there, and unlike human language, they are documented unambiguously and in source code. And JPEG is used globally, unlike risky region-specific encodings such as English, Chinese, and Spanish (I'm only half kidding).

ASCII is safe. JPEG is safe. Basic HTML is safe. The only problem a thousand years from now will be finding the good stuff among all the boring crap our descendents will wish we'd deleted. (Then again search engines may be smarter than we are by 2106).

Unfortunately I'm not as confident in sound and video. MPEG is pretty safe due to DVDs, other codecs I wouldn't trust for archival in the slightest.

Re:CD Presses

[marcosdumay]

I was thinking about this recently. The problem is that you need a really higth power laser to create a medium that lasts. If the material is inert (the nearer that I can think about are gold and glass, preferably glass), the degradation time will decrease exponentialy with the laser power*. That is good, but even then, you'll need a very powerful laser.

Things can go better if you use some kind of revelation process, like films. Then you can create higth energy recordings with a low energy laser and a chemical reaction. The problem is that inert media don't sufer chemical reactions easily, that is by definition. Other alternative is using a mechanical process. Mechanical processes with the power required to make your media last a very long time are easy to engineer, but they are not very precise. So, we end up with a low density medium.

The only thing I can see how to create a long lasting medium with is eletrical fuses. I think it is possible to design a hight density PROM chip with this characteristic, but I never saw one.

* The odds of any process happenning that will destroy your data decreases exponentialy with the energy that this process needs to happen. That is, if you use an inert material.