CD-Roms: On The CD Side

Sounding like something from Reservoir Dogs, the history behind standards for CD-Rom is almost as messy. But, writes Chris Long, the technology is still alive and kicking

If there is anything you can be sure of in this industry, it?s that you can?t be sure of anything. We only have to cast our minds back to that joyous day when our very own Bill Gates said ?640 kilobytes of memory will be more than enough? to understand that these things are never as simple as they look.

The list of things that, when they arrived, were going to change the world, and in the end weren?t around long enough to change their underwear, is endless. Similarly there is the list of things that ?the people in the know? said would go away but didn?t. We only have to cast our mind back to that other joyous day when our very own Bill Gates said ?Dos will be dead by 1993? to see that. Still, you would have thought a man so rich would have a higher batting average wouldn?t you?

Well, the industry has been wearing black armbands for CD-Roms for a while. Grave predictions of their passing have been made for several years. And as we see the latest batch of 16-speed drives arrive, we may ponder the words of a certain CD-Rom vendor who said: ?I don?t see the market going much beyond four-speed.?

All the same, the evolution of the CD-Rom has seen many stages, usually defined by a colour. CD-Rom standards are nominally named after the folder or book that standard was distributed in. It is easy to get the impression that this is helpful. It isn?t.

Philips and Sony defined the original CD format in 1980. The CD DA (compact disk digital audio) Red Book standard demands a 44.1KHz sampling rate and up to 74 minutes of recording. A lot of people blame Dire Straits, arguing that if it wasn?t for the success of their Brothers In Arms CD, we wouldn?t be looking at a CD-Rom market at all.

Certainly it didn?t take long for the computer industry to see the benefit of storing 650Mb of computer data on a 12cm optical disk. The Yellow Book standard arrived in 1983 and set out the methods for working with data ? the bog standard CD-Rom.

CD-I (compact disk-interactive), or Green Book, came next. It defines a hardware and software standard developed by Philips for use in its CD-I players. It was a consumer format designed as a home entertainment system delivering music, graphics, text, animation, and video. Unlike the CD-Rom drive, a CD-I player is a standalone system that requires no external com- puter. It plugs directly into a TV and stereo system, and it is worth noting that all of your standard music CDs will play on a CD-I player.

While Green was aimed at home entertainment, Orange Book defines CD-R or CD-recordable, a standard that originally came from the magneto optical drive arena.

Next came CD-Rom/XA (extended architecture), which was another form of CD-I combining elements from several books, mostly Green and Yellow. It permitted the interleaving of audio, video, and arbit- rary data like text, so an XA drive can play audio and display pictures at the same time.

One of the problems with the CD-Rom format was that it was a write once system. Once the table of contents (TOC) had been written on the disk that was it ? the hardware and software didn?t look any further. This was fine for all formats so far, but the next development had some problems with it.

See, Kodak had this idea for storing photographs on a CD, which in it self was all fine and dandy. But this format rather demanded that you could ?update? your CD as you got more photos. Essentially the need was for a multisession disk where the user could add data to the end of the previously written data in different sessions, which meant the CD-Rom?s TOC must be ?updatable?.

So, enter another format: Photo CD ? readable on CD-Rom/XA and CD-I. Of course, this improvement of CD-Roms to the multisession format wasn?t a magic wand process. Oh no. A single-session drive can read a multisession disk fine, but it can only read the contents of the first session and that is all. All the other sessions are ignored, as the system doesn?t understand the ?new? multisession TOC format.

Thus the new TOC required hardware changes to CD-Rom/ XA drives ? applied after the format had been shipping, which is why there are a fair number of single-session and multisession XA drives on the market. Ironically, despite CD-I disks being single session, the drives themselves are multisession so they didn?t need to be upgraded.

And finally in this whistle-stop tour: White Book. Set up in 1993 is effectively a round up of the Video CD standards providing up to 72 minutes of linear video on a single CD.

During this time it will come as little surprise that users started demanding more from their hardware, and it soon became apparent that the original CD-Rom transfer rate of about 150Kbps, and access times measured in hundreds of milliseconds, was not enough.

Soon CDs were carrying digital pictures and other data that needed drives with better specifications. So manufacturers came up with the pleasantly simple solution of spinning the disk twice as fast, which meant the data arrived twice as fast at the laser pickup. These became known as double speed drives (or two-speed), then it was quad drives (four-speed) with transfer rates of about 600Kbps. When 1996 arrived, so did four-speed drives, but these were soon to be eclipsed by six-speed models. They, in turn, were supplanted by eight- speed models. And as if caught in a mad rush, 10-speed drives made their appearance, quickly followed by 12-speed and 16-speed versions.

But like a lot of things in this business, just because it says something on the box, it doesn?t mean it?s true. So, for example, saying the disks run at 16-speed doesn?t actually mean they do run at 16-speed. Which brings us down to yet another computer industry fudge. It is all down to the spin of the disk.

Conventional CD-Rom drives have grown up using a technique called constant linear velocity (CLV). This aims to keep the transfer rate constant by fiddling with the speed of the spinning disk. This is because there is more data around the outside edge of the disk than closer to the centre. For example, reading a disk spinning at 4,240rpm at the inner tracks would give a data transfer rate about 1,200Kbps, which is an eight-speed rate, but at that speed reading the outer tracks would give data at something like 3,000Kbps ? equivalent to a 20-speed disk. This is close to the limit of most current optical pick-up systems.

To put this into perspective, on an audio CD player, each sector of audio data must pass under the laser pick-up in exactly 1/75 of a second for the music to be played at the correct speed. This 1/75 of a second per sector velocity is the constant linear velocity associated with a one-speed CD-Rom drive.

So when referring to a CD-Rom drive as a two-speed, four-speed, eight-speed or what- ever, it isn?t the actual spindle speed being described but the data transfer rate. This means that there are times when a four-speed disk is spinning faster than an eight-speed disk ? although the eight-speed drive will always move data from the disk twice as fast as the four-speed drive.

But there is a cut off point here, simply running the disks faster isn?t the answer. To achieve its true data rate, a 20-speed CLV drive would have to spin the disk at 4,000rpm on the outer edge, but at something like 10,600rpm on the inner portion ? a speed at which the disk, if it ever got loose, would have no problem imbedding itself into the soft thing that sits in front of a CD-Rom drive: the user.

Fundamentally this is too fast for the drive, the disk is not balanced ? the printing on the top of the disk would unbalance it, and the hole in the centre isn?t always in the centre ? and the drive mechanism itself wasn?t designed to cope with such speeds.

The answer seems to lie in CAV (constant angular velocity) a system where the disk rotates at a constant speed and doesn?t change speed when the data density changes. This method is easier on the spindle motor and relies on improved pickup technology to allow the laser to read the disk?s outer tracks at ?full speed? (where the disk would have slowed down to read the outer tracks it now won?t ? giving the increased transfer rate). But the pay-off means that data transfer rate varies between the outer and inner tracks.

And let us not forget that a CD is written from the inside edge out, making it more likely that the data will be on the inner tracks. So if the CD is three-quarters full, you?ll won?t see it running at full speed because that is only achieved from reading the outermost tracks ? even though that is the speed the drive will be labelled at.

This has led to complaints that the labelling of the CAV drives is misleading, because they say 16-speed drives will only be achieving 16-speed throughput when taking data from the outer edge of the disk.

Some drive makers have decided on a hybrid approach called P-CAV (partial CAV), keeping the data rates up by spinning the disk faster when reading the inner portions. The actual figures are different for each manufacturer, but it goes something like this: a 16-speed drive spins at a constant CAV rate of 4,200rpm over 70 to 80 per cent of the disk. This is much the same speed ? and so gives the same data transfer rate ? as a standard eight-speed disk. This then drops to about 3,200rpm for the outer 20 to 30 per cent of the disk, giving a data transfer rate equivalent to a 16-speed drive. In effect, despite the drive being labelled as 16-speed, it actually performs somewhere between eight-speed and 16-speed depending on where it is reading the data.

Yet again the industry has come up with a partial solution to an insoluble problem. The irony is that while DVD ? the supposed CD-Rom killer ? is still waiting to take off, the CD-Rom market has bought itself time and technological status. It is unlikely that an industry so agile will go away in a hurry.