This article first appeared in Music Mart in the Autumn of 2004. As far as I can make out, the situation hasn't substantially changed since. A version re-investigated and re-written for the computer industry magazine Computer Trade Only appeared a few months later, but didn't reach significantly different conclusions. Caveat emptor has rarely seemed more appropriate than when buying blank media (especially DVDs).



Just how safe is the material you've got burned onto CDs? The songs you've written, the alternative mixes, the live gigs, rehearsals, jams and assorted musical ideas? Depending on whom you believe, those discs could still be playable in a century from now - or, in just a couple of years time, all you might have left is a set of very nostalgic coasters.

The rumours about potentially unreliable CDs started circulating on the Internet sometime last year. They began when a Dutch computer magazine, PC Active, reported that it had conducted tests on blank CDs, burning data to them, then subjecting them to - well, not very much really. Certainly nothing like the abuse many domestic users subject their discs to. In fact all PC Active did was burn sample discs from 30 different brands, stick them in a cupboard for two years and then try using them. Astonishingly, the magazine found that some were completely unreadable and that ten per cent of their samples showed problems of one kind or another.

Written in Dutch, so exposed to a small audience, the article didn't cause the international consumer reaction that it might have done. But soon after the story was published, PC Active's readers are said to have started writing in with tales of their own problems with blank CDs and then digital photographers began complaining on technical websites about difficulties they'd encountered, burning irreplaceable images to CD, only to find them irrecoverable later on.

Ironically, one of the advantages we were promised if we moved to digital was that, once committed to ones and zeroes, our priceless works of art were fixed. Wasn't one of the major drawbacks with analogue tape supposed to be, that from the moment your tracks went onto tape, the quality began to erode? In theory, digital data, once stored, doesn't degrade in the way analogue magnetic tape recordings do. Providing all the ones and zeroes are there, you get back 100 per cent of what you recorded. But it's when they aren't that you've got problems.

On a commercially produced CD, the data storage is created by a series of 'pits' stamped into a polycarbonate disc, which is then coated with reflective aluminium material, over which are placed protective layers of acrylic. How permanent is that? It's a question to which the only honest answer is: nobody knows. From time to time, scare stories appear, suggesting that even commercially produced CDs will lose data after ten, twenty or thirty years - but the CD is getting pretty long in the tooth now and have you personally met anyone whose early copies are creating problems?

The home-grown CD, on the other hand, is a completely different creature. When you burn a CD in your writer, burning it is precisely what you are doing. A laser beam is fired through the protective coating on the surface of the blank disc and through the (usually aluminium) reflective layer, where it hits another layer, of light-sensitive chemical dyes. When it writes, it turns the dye dark (a "one") and where it doesn't burn, leaving the dye layer pristine, that absence represents a "zero".

You only need to watch a few washing powder adverts to see the potentially fatal flaw in this technology. Your prized Archdemons-Of-Mega-Death-Isle-Of-Wight-Tour-2003 T-shirt looks more, well, grey these days than the "Satan's Soul Black" it was when you bought it, simply because dyes fade when exposed to light. And that includes the ones in your CDs.

Which, at least in theory, could suggest that rewritable CD-RWs might be a better bet for archival purposes as they don't use dyes at all. Rewritable CDs use a metallic material in place of dyes. When it is hit by a laser it changes its state, reflecting differently and, thus, giving the on/off, zero/one pattern without the weakness to light. So does that make, paradoxically, the theoretically "temporary" storage medium actually better for long-term storage than conventional CD-Rs?

That was just one of the questions I wanted to ask blank CD makers, when I set out to write this article. And what a process that turned out to be. In fact it needs some explanation.

By and large, one of the fun things about being a journalist is that when you want to know something, you simply call XYZ Co. and ask them. Sometimes you get the run around from press officers but usually, in the end, you get to speak to The Man Who Knows.

Except that, this time, TMWK was awfully hard to get hold of. Some companies never responded at all (thanks, Verbatim), while others had to be repeatedly badgered before they would respond - and usually when they did, simply sent a set of e-mailed answers to specific questions. Only two actually agreed to be interviewed: take a bow Fuji and MPO - the French company that recently bought BASF's data storage division.

All the journalistic alarm bells were ringing. It seemed that many companies (for whatever reasons) didn't want to talk. But why?

One thing that emerged from this exercise was that there is no watertight consensus between blank CD manufacturers. You might, for example, have expected the major brand name manufacturers to close ranks and dismiss the problem as only applying to cheap Taiwanese/Chinese-produced no-brand discs which use unstable dyestuffs, prone to damage by exposure to heat and light. And in a sense they did - except that not everyone believes longevity problems are necessarily always due to problems with dyes. A case in point, being Ben Bruggink, the Quality Manager at Philips Recordable Media - the company that (with Sony) jointly invented the CD. He thinks there are other problems, too.

Asked if he thought premature data loss was a brand issue, or one inherent in very structure of all CDs, Mr Bruggink said: 'The dye is not involved in the process. The main culprit is the lacquer used to cover the metal mirror. There are some manufacturers that use very inexpensive types of lacquer or very thin layers that easily are damaged.'

You might be reluctant to take the word of Philips at face value, given that they co-invented the format and could therefore be expected to defend the it all costs, but Japanese rivals, TDK, at least to some degree, sing from the same hymn sheet. For example, Hartmut Kulessa, Product Marketing Manager for optical media, TDK says: 'PC Active reported a problem with some discs having a badly manufactured top coat, which allowed oxygen to get between the layers and made them corrode after a short amount of time. This was definitely only a brand/manufacturer problem and TDK discs were not affected.

'Regarding UV light, yes, cyanine dye as it is used in CD-R is sensitive towards UV light. Therefore TDK ships its CD-Rs in secured packaging and recommends storing discs in their packaging. Some CD-R products claim to have UV protection in the form of black coloured polycarbonate on the base layer, but TDK do not recommend these products because of reduced reflectivity and compatibility.'

For all that the specific problems encountered by PC Active may have been down to weak protective layers, there remains no doubt that the basic technology used in CD-R - the photo-sensitive dyes that make it work - are its Achilles heel. As Maxell's Harry Yoshimura says: 'This is inherent in organic dyes that are used as standard in CD-R manufacturing.'

Independent views tend to back the Maxell response, as far as I have been able to ascertain. While there is, clearly, an issue with the lacquers and other materials used to protect the integrity of blank CDs, most independent commentators seem to point to some quite significant differences between different dyes being used by the various manufacturers.

Imation, however, one of the most trusted media brands around, responded as follows. Make of it what you will: 'There are two different dye technologies used in CD-R manufacturing: green and blue. As there is no evidence to suggest that one has stronger advantages over the other, it is therefore not possible to draw general conclusions about the product quality of a CD-R brand based solely on the dye technology used.

'However, there are different quality standards amongst manufacturers and with lower quality CD-Rs it is possible that consumers may have a less than satisfactory experience when subsequently trying to retrieve archived or stored data from their CD-Rs.'

Fuji's Tim Husband (one of the few engineers who was willing to be interviewed), has his own thoughts on this issue and explained how Fuji reckons its products stand up.

'I don't know about other manufacturers, but we publish guidelines for the lifetime expectancies of our products and we have stated that we expect our CDs to last 100 years, our DVD +RW, -R and -RW also 100 years and our +R discs 35 years.

'Life expectancies, yes, are dependent on the dyes used and different CDs use different dyes, and there are conditions on these guidelines, for example those figures don't apply if discs are kept in direct sunlight. In fact the optimal conditions for storing CDs are a temperature of between 15-25C, 20-40% relative humidity, no direct sunlight and free from dust.'

It needs to be said that some experts are sceptical about claims of a 100 years lifespan for a CD. Given that the technology hasn't been around for very long, how can manufacturers be so sure? The answer is that they conduct wear and tear simulations to estimate the damage done over the years. Still, not everyone is convinced. For example, the UK's National Archives at Kew are reported to be using magnetic tape in preference to CD-Rs, and in the USA, experts working on behalf of the Library of Congress have gone on record, expressing their doubts about the value of manufacturer's tests. Indeed, one comment made was that when you buy a disc, you never quite know what you are getting, anyway. Manufacturers change the way they make their discs all the time - and don't tell the customer. More than that, when you buy Brand X, you can't be sure that the discs are even made by that company. Discs have become a commodity and marketing companies buy from a variety of manufacturers, switching sources as it suits them. It's reasonable to wonder if some of the major brand names you see in the High St actually make any discs at all - and the fact is, some of them don't.

A case in point was my first port of call when I started researching this article - a major Japanese computer manufacturer which sells its branded products through a very prominent UK retailer. Its discs are to be found in stores up and down the country and, in fact, I've used them very extensively myself - in CD-R, CD-RW, DVD+R and DVD+RW formats and never encountered a problem.

Only, they didn't want to comment, their PR man told me, as they buy-in their media products from a range of suppliers, so couldn't really offer much of an opinion. Which, if nothing else, was honest!

Ben Bruggink at Philips is equally frank: 'The issue is brand dependent, because every brand has its own agreements about the deliverables with manufacturers. We have a very specific set of specs to ensure the quality level we get from our suppliers. If some defect comes apparent we immediately correct and prevent it from ever happening again.'

All the same, if Philips admits it buys-in CDs from other manufacturers, why shouldn't consumers buy those makers' products direct and save a few pence per disc?

'The user should rely on the quality level of his brand,' he says. 'The real A-brands have a very active quality policy and all the colleagues I know are only releasing top grade material. Leave B-brands in the shop, many of them buy inferior B-grade products.'

His argument is pretty simple. Buy products from big name brand because they won't risk their reputation selling inferior quality products. Maybe.

For a different angle on the subject I next spoke with the French company, MPO. Though not very well known in the UK, it started life in the 1950s, pressing vinyl records, later was a pioneer with audiocassettes and, today, makes everything from CDs to Mini-Discs, which it sells under the Hi-Space brand (you can find them in PC World, among others). As mentioned earlier, MPO has recently bought BASF's EMTEC media business and Jean Huet, the company's Process Engineering Manager was not only very willing to be interviewed on this subject but, being an engineer, as distinct from a marketing man, was refreshingly frank.

'There are poor quality products on the market, for sure. But all CD-Rs are sensitive to light, are sensitive to humidity and temperature - in other words to the environment. Mostly, the problems we see when we get a complaint are down to handling. You can't believe what you see sometimes with badly scratched discs and so on. That is down to how they are used.

'Most CD-Rs will have a good life expectancy if you follow some simple rules and handle the disc carefully. Only touch by the edge. Do not apply labels. Do not use transfer printing systems - use inkjet printing instead. Avoid exposure to direct light - either sunlight or fluorescent. Avoid high temperature and high humidity - these are the same rules that all the manufacturers will tell you. If people follow these rules, they are likely to find their data is quite safe.'

Even so, how can one feel secure even buying a big-name brand, if its discs are produced in the same factories as those you find on a market stall selling for 20p each?

'Well, it's no secret that a lot of famous brands are no longer producing CD-Rs themselves. Something like 70-80 per cent of CDs are produced in India, China and Taiwan by three or four big companies - so famous brand names are all buying from the same sources - including us at MPO.

'So you're never sure when you buy a disc you were satisfied with that, when you buy that brand again, you're getting exactly the same product. Most of those big brands change suppliers from time to time and do not control what is produced for them.'

So what on earth is a musician supposed to do if he really needs to be sure that his material will be safe?

'I can only say, use our gold range,' says Jean Huet. 'Almost all CD-Rs use a silver reflective layer and silver is not a very stable material. If it's protected by a good layer and not exposed to light and excessive humidity, it should have a good lifetime, but if you use gold - gold cannot be oxidised and this gives a better lifetime.'

This isn't necessarily sales talk, by the way. There is a lot of independent testimony on the web suggesting that gold is the way to go if you are looking for the greatest possible reliability from CDs. Its cost, however, in a fiercely price-sensitive retail market, means that almost all companies that used to offer gold discs, have now stopped - except for MPO which, incidentally, guarantees its gold range for 100 years.

So what about CD-RW as an alternative? If CD-Rs are prone to problems, whether they are due to flaws in either dyes or coatings, why not use CD-RWs instead? Not reliant on dyes, surely that's cutting out one of the two main causes of trouble?

Yet again, it depends whom you ask. TDK's Hartmut Kulessa, for example, says: 'With regards to UV sensitivity, yes, CD-RWs are better. But given the problem of badly manufactured discs, it's most likely that you could find similar situations with CD-RW as well.'

Imation, meanwhile, says: 'It is true that the "phase change" technology used in CD-RW is not as sensitive to light exposure as the dye technology of CD-Rs. The CD-RW write process is based on the phase-change layer (crystalline/amorphous) being modified by applied laser head at a defined laser power during a defined timeframe. However, this does not mean that rewritable media is fully resistant to UV radiation. Therefore, the recommendation for storing CD-RW media is the same as for CD-Rs.'

Maxell's Harry Yoshimura, meanwhile, says: 'CD-RW is an excellent format, but we consider both CD-R and CD-RW are equally both good formats for archival applications. However, we believe the recording layer of CD-RW is more complex, so there can be some variation between brands or manufacturers depending on the equipment and quality control systems in place.'

Fuji, interestingly, quotes identical anticipated lifespans for both media types.

MPO's Jean Huet says: 'The materials used in that sensitive layer are inorganic, so they're not as sensitive to light, that's for sure, but they are very sensitive to oxidation and the properties of the recordable layer change each time you write and read it so I would say it's neither better nor worse, on the whole.'

More opinions? Memorex says: 'CD-RWs should last at least as long as CD-Rs, perhaps longer. CD-RW discs have the reputation of being unstable, but factors such as packet-writing complexities and incompatibilities, varying laser spot geometries from different drives, and file corruptions are more to blame for the difficulties than the design or materials of the discs. The recording layers are inherently more stable than organic dyes are immune to the effects of UV light and moderate heat. The alloy crystallizes at 200 C. (420 F.), well above temperatures that would damage CD-Rs. Chemical changes due to humidity (polycarbonate plastic can absorb water) can promote ion migration if the disc is not made well.'

On balance? It seems that CD-RW should perhaps be regarded as pretty much the equivalent of CD-R, at least in terms of its longevity.

Which brings us to DVD. Given the recent availability of cheap DVD burners, the relatively huge 4.7 gig capacity of the discs and the fact that they are a newer product, is the DVD our holy grail?

It seems not.

None of the manufacturers I spoke with was willing to suggest that DVD is any more reliable than CD-R or CD-RW, in terms of its longevity. These discs are made of, effectively, the same materials, though there are differences. Where a CD-R is protected by 10 microns of protective layer, a DVD has 600 microns of polycarbonate on its printed side - so it should be much more resistant to scratches and other surface damage. But it is just as vulnerable to corruption if exposed to light, high temperatures and humidity and it is specially subject to damage if flexed when getting one out of those ridiculous DVD cases which seem purpose-designed to encourage disc flexing.

So what are we to do? Most musicians have accepted that digital is the only way to go and that storage on (magnetic, never forget) hard drives, however cheap they have become in recent years, is just too inherently unreliable for archival purposes. We've been told that CD-RW is probably no better (though not necessarily any worse) than CD-R. That DVD in any of its variants isn't likely to offer any advantages other than the sheer amount of data we can store on it, so what do we do?

The moral seems to be to buy a decent brand. Taiwanese no-name spindles may be fine. They may even be exactly the same discs as you pay a premium for when they carry a well known brand name. But is it worth the risk to save a few pence per CD?

The next observation is, I admit, a personal one: buy a brand that seems to have confidence in its own products. Fuji, TDK, MPO and some others offer guarantees - I can't see any reason not to take advantage of them - though whether a box of replacement discs (usually the extent of any admitted liability) is really any consolation for the loss of your best-ever live recording, or priceless demos, seems questionable.

Perhaps above all, it seems that we, the users, may end up having the greatest influence over whether our precious data survives - given that we have bought good quality discs in the first place.

As, for example Tim Husband at Fuji says: 'There are some really simple golden rules. Do not touch the recording surface; avoid finger prints or smears on the recording surface. If cleaning is required, use a soft dry cloth or CD cleaning liquid / pure alcohol, if necessary. Do not use any solvents. If inscription is necessary, use a soft felt-tip pen on the label area. Do not use a sharp or hard-tipped pen. For labeling, use only a CD-R label with a labeling kit. Always ensure disks are stored in the case provided, avoiding direct sunlight, excessive heat and humidity.'

Imation offers much the same advice, with the following observation: ' Keep them dark, cool and dry. Optical media uses laser beamed light to read and/or write data, so optical discs shouldn't be exposed to sunlight. The UV-radiation of normal sunlight is quite aggressive and can damage the storage layer within the disk. Heat can also damage optical discs, impacting the stability of both the protective coating and the storage layer. It can cause warping of the disc, which will result in read/write errors and ultimately data loss. Moisture can also affect data integrity. Although the plastic (polycarbonate) itself is waterproof, high degrees of humidity accompanied by high temperatures can enable moisture to penetrate to the dye layer, which can lead to degradation. Where long-term storage of data is required, it is vital that the media is kept in a protective case or box, protected from direct UV-radiation, at a temperature range from min. 5C to max. +30C and an air humidity of 40 to 60 per cent.'

Memorex, too, has some very sound advice - not to mention some excellent information on its website. Mike Warden from Memorex Europe says: ' Our standard CD-R disc uses a light- and temperature-stable dye known as phthalocyanine that rigorous testing shows to be stable for approximately 70-100 years if properly stored. It is both the storage and handling that can be the greatest factors in determining longevity. Discs should be kept out of light, heat, humidity, and severe temperature swings. Discs should be kept in their cases and handled only on the edges with as little flex as possible to the disc when it is removed from or put into a drive or case.'

On the subject of its DVD's, however, he sounds a cautionary note. 'All the DVD discs we produce use the finest dyes recommended for recording, although by their nature these dyes are more susceptible to light deterioration than those used in CD-Rs. DVD-R uses azo-cyanine dye; the DVD+R uses a metal chelate dye.'

The fact that there are differences of opinion between experts suggests that we, the consumers, need to be careful here - in fact, downright conservative. I would really have liked to have been a lot less equivocal in this article but, even if we have no definitive answers, we do seem to have some guidelines.

Buy good quality, branded discs - ideally from companies that offer warranties. Handle and store discs as the manufacturers advise - keep them cool, dark and don't let your grubby fingers mark the playing surfaces. Don't label them unless you are certain the label system is safe. Use soft felt pens for marking or, maybe, use dedicated inkjet printers and suitable discs. And how about, if your recorded material really is irreplaceable, burning it onto different brands and types of discs? Say, a couple of different top brand CDs and a CD-RW? For the sake of a few pounds, the belt-and-braces approach has a lot of appeal.

And finally? Two often overlooked points. When a blank CD (or DVD) is written, your drive requires specific information about the blank medium being used - which means simply installing a new drive and forgetting it is not a good idea. Periodically updating your drive's firmware from the manufacturer's website is certainly worth doing.

The second is perhaps to change the very way we think about stored data. Just as the wax cylinders which so amazed the Victorians are obsolete today, so are the 5 1/4" floppy discs from the 1980s and, very nearly, the 3 1/2" discs that followed them. The pace of change is accelerating, which means once you have your data in digital form you have no option but to keep updating the medium on which it is stored. Being digital, you will lose nothing (given a perfect copy) as you move from, say, DAT digital tape to CD-R to DVD, but woe betide the songwriter who assumes that anything he records on today will still be usable in 50 years time. Even if the medium he uses will still be playable, will he be able to find the equipment required to play it? Updating your method of data storage may be a chore - but it is also an inevitable fact of life now that we are, however insecurely, in the digital era.


2005 Gary Cooper