Home Page | About Me | Home Entertainment | Home Entertainment Blog | Politics | Australian Libertarian Society Blog | Disclosures

Killer CDs?

Previously published in Australian HI-FI


December last year saw The Age warning of the dangers of CDs.1 Dangers? What dangers? Can you play your old Black Sabbath CDs backwards to unleash spiritually destructive Satanic messages? Will excessive exposure to John Denver re-releases propel one down a path leading inevitably to the bottle?

No, the fear is that your CD may well explode, impaling you, the innocent listener, with shards of aluminised plastic. Fanciful? Well, yes and no.

Forget the hype, you will not die, shredded by shrapnel that formerly constituted a treasured Schubert String Quartet CD, as long as it is playing in a regular CD player. Likewise, you will survive your encounter even with an Arnold Schwarzenegger DVD, so long as it is spun by a regular DVD player. By regular, I mean a CD or DVD player ranging from a no-name unit picked up cheap at the local Two Dollar shop, to a high end unit with its wiring personally soldered by an audio/video-phile company's CEO.

However there are things to keep in mind with computer drives.

Speed kills

One of the problems with LPs that was overcome by CDs was the variation in quality depending on where the particular segment of music was located on the disc. The reason is that LPs are Constant Angular Velocity (CAV) devices, whereas CDs are Constant Linear Velocity (CLV).

An LP rotates at 33 and one third revolutions per minute. That's all there is to it. Data density, however, is determined by the speed at which the recorded surface passes by the reading device. In the case of an LP, this speed was pretty high at the edge of the disc, and rather low towards the middle. To ensure that the greatest amount of data (for the time) could be packed onto a CD, it was designed to offer its tracks to the reading laser at a constant speed. So it had to spin the disc faster as the inner tracks (which are also the first ones) were being read, and slow down as the disc progressed. All this was made possible by the feedback system in which the motor spinning the disc was controlled, in part, by the structure of the data being read from the disc.

In the case of a CD, the read speed maintained across the whole disc is 1.21m/s (metres per second), while the inner track is located at a radius of 25mm and the outer at 58mm (the radius of the disc is 60mm).2 To achieve this linear velocity the disc must be spun at 462rpm for the inner track, with the speed gradually diminishing to 199rpm for a track right at the edge of the disc.

DVDs also use CLV reading techniques, and the comparable rotational speeds are from 1,400rpm down to 550rpm.

These are modest rotational velocities and we seem to have survived the last two decades since the introduction of the CD without government-mandated health warnings as to their physical danger.

The concerns about killer CDs stem from destruction testing on CDs and DVDs performed by a Swedish technical man, Jörgen Städje. In 2001 he published his results.3 Spinning a series of CDs, he found that when they reached rotational velocities of between 23,000 and 30,000 rpm (revolutions per minute), they shattered explosively. Note: not might shatter, could shatter, nor any other equivocations. He tested eleven CDs, and all eleven shattered. Even the one to which the experimenter added Kevlar binding.

In each case, the result was thousands of tiny shards of plastic. The aluminium protective casing around the rotating spindle was heavily bent, in one place pierced. Sometimes the destruction was instantaneous. In the case of the Kevlar-bound CD, it took twenty seconds (during which time the plastic stretched around the Kevlar) before letting go.

He later repeated the experiment with DVDs and found that they tended to let go at around 27,000 or more rpm, thanks to a somewhat stronger structure. But for the same reasons DVDs tended towards longer, sharper shards of plastic.

Big numbers

So what has all this to do with the real world of CDs and DVDs?

There is almost no point in driving a consumer CD or DVD player faster than the specified speeds (the exception is anti-shock portable CD players which need to read ahead somewhat to buffer data to fill in shock-induced holes in playback). Computer CD-ROM drives, though, have changed all the rules.

The principle use of CD-ROM drives is not playing CDs, but running programs or extracting data from CD-ROMs. Increasingly, a major use is also extracting audio data from audio CDs. Note that this last is not the same as 'playing' the CD, since it bypasses the audio-specific circuitry of the CD-ROM drive. Instead it 'rips' the audio data using the computer's IDE (or SCSI) interface, just as though it were a program being read, or some spreadsheet data.

In these applications, speed is a virtue. If a CD-ROM only operated at the speed of a regular CD player, a ten megabyte program would take a full minute just to start! So soon after CD-ROM drives were introduced, there were 2x, then 4x, 8x and so on. Nowadays a standard CD-ROM drive is 52x.

What this 'x' terminology means is that the drive is specified to read at up to that many times the regular playback speed of a CD. A 52x CD-ROM will run at up to 52 times the speed of a standard CD player. And here is where some enormous numbers start to appear.

Think of a CD playing at 52 times normal speed. Not only would it be musically unattractive, but boy would that disc be spinning fast. How fast?

Well, if the inner track is at a 25mm radius then the edge of the disc is going at 60/25 times the linear speed, or 2.4 times. At 52 times the 1.21m/s normal reading speed for the inner track, the outer edge of the disc would be running at:

1.21m/s x 52 x 2.4 = 151m/s

That speed is an incredible 46% of the speed of sound!

It would also be spinning very fast. How fast?

462rpm x 52 = 24,000rpm

Now clearly there are millions of 52x CD-ROM drives out there. Given that Mr Städje's test showed all discs blowing up at only slightly more than this angular velocity, clearly there must be thousands of people whose drives have been damaged and discs have been destroyed by these extremes of rotation! Especially as not all CDs are perfectly formed, and thus have weak areas which would bring them to an even earlier demise!

CD-ROM makers aren't as dumb as feared

But there is a problem with this analysis. Despite his work being reported around the world, he has in turn received reports of a mere score of damaged CDs (he suffered one himself, although the damage was less than explosive).

So why this discrepancy? This, it turns out, is the discrepancy between CLV and CAV. A standard CD or DVD player is a CLV device, operating at the appropriate 1x speed for the format. A CD-ROM or DVD-ROM drive can operate in both CLV and CAV modes. CLV mode is for playing back a CD or DVD. CAV mode is for data extraction or CD ripping.

That's why a 52x CD-ROM drive is expressed as running at 'up to' 52x. It achieves that speed only on a CD (or CD-ROM), that's in very good condition, is completely full (so that there is data on the extreme edge tracks) and, even then, only at that extreme edge. For the inner tracks, a 52x CD-ROM will read at no faster than 52/2.4 = 21.7x (remember that 2.4 is the ratio between the inner and outer tracks).

That is also why you may have noticed, if you have ever ripped a CD, that the first tracks rip very much more slowly than the later tracks.

Most importantly, this also means that a 52x CD will run very much slower than feared. A 1x CD runs at 199rpm when reading the longer outer tracks, so the maximum speed the 52x CD will achieve is:

199rpm x 52 = 10,348rpm

This is much less fearsome number, being less than half of the normal self-destruct angular velocity. But if you spin a disc with some kind of flaw at this speed it could, still, break up, or perhaps crack and become unplayable.

What about some of the new high speed CLV CD-ROM drives that are occasionally touted? Well, as it happens they are high speed, but are still built to limit the maximum rotational velocity of the disc. For example, a Google search of the Web reveals one of the fastest of this breed to be the Philips CDD5101, a Combo DVD/CDRW drive which offers a 16x speed and is specified to run at between 3,312 and 8,640 rpm. Note, that maximum speed is rather less than a 52x CAV drive.

Conclusion

Mr Städje's tests are instructive, and do make useful cautionary points. But his recommendations regarding the wearing of protective clothing seem to be somewhat tongue in cheek. Regardless of what you choose to wear while spinning CDs in your computer, though, you can rest assured that the modest spin speeds achieved in consumer equipment in good order represent no danger whatsoever to consumers!


NOTES

(1) 'Experts in a spin over killer CDs' by Nathan Cochrane, The Age, 10 December 2002 -- http://www.theage.com.au/articles/2002/12/09/1039379784015.html

(2) http://www.daisy-laser.com/tech3c.htm

(3) http://www.qedata.se/eindex.htm

© 2003 by Stephen Dawson