The rise of optical media for data storage has provided answers to many of the problems thrown up by ever more powerful samplers and digital audio systems. Yasmin Hashmi takes a look at where optical technology has been, and where it's going in the future.

Even before its introduction, the optical disk was hailed as the potential panacea to the problems of archiving. Like hard disk it has a relatively large storage capacity, but like the floppy disk it is removable. Furthermore, unlike floppy or hard disk, it is very robust storage medium, and is not affected by strong magnetic fields or X-rays.

When optical drives became available many realised that it was in fact too slow for some applications, its potential was nonetheless recognised and many foresaw that it would inevitably begin making serious inroads into the professional audio industry. In fact, it could be argued that this has already happened. The WORM (write once read many) optical disk offers mass storage capacity and has been used by the higher end of the market for library purposes for some time. However, it was not until the introduction of the erasable MO (magneto optical) disk that interest in the technology began in earnest.

As with all computer-related products, development of optical technology has not stood still. Access times (a measure of how rapidly data can be located and retrieved from a storage medium) are constantly improving, so much so that optical drives can now compete with hard disks for a variety of uses. In the tapeless recording/editing market for example, at least half a dozen systems are already using MO as the primary recording medium — although this still only represents less than 10% of the market, it is a clear indication of the future role of the medium.

For applications where access times are less critical, the adoption of optical is more likely to be dependent on costs — not only of the drive, but of the disks themselves. Furthermore, potential purchasers need to feel confident that their investment will not be immediately superseded by better and cheaper technology, or new formats which are incompatible.

Nonetheless, the advantages of optical are beyond doubt, and the recent introduction of the low-cost 3.5" MO disk promises to significantly increase the use of optical for a wide range of applications.

BACKGROUND

Optical technology first came to the fore in the audio industry with the introduction of the compact disc. Designed as a read-only consumer product, its audio quality, convenience and robustness ensured its victory over the vinyl disc. Its success also had a major impact on the professional audio market, heralding the end initially of analogue mastering, and eventually also of analogue recording and editing techniques. Apart from the compact disc, all other disk-based technologies which are used by the professional audio industry (such as floppy disk, hard disk and optical) are borrowed from the computing industry, which has significantly greater resources for research and development.

The first application of optical technology for recording purposes was the WORM disk, which could be recorded to but not overwritten. It was introduced to the audio market as a supplement to hard disk, providing mass storage and on-line access to sound libraries. The WORM is double-sided, available in 14", 12" and 5.25" sizes, and can have a total storage capacity of up to 4GBytes. The cost of the drive and media, however, put the technology out of reach for the majority of potential users.

The next major development was the 5.25" magneto-optical (MO) disk. This is double-sided with a total storage capacity of 600-650MB. It is erasable and could therefore be used for real-time recording as well as for library purposes. However, the 5.25" disk is still prohibitively expensive for the 'mass market' of mid to lower range studios — around £3,000 for the drive and around £150 per disk.

More recently 3.5" drives have become available. The disks are single-sided, with a storage capacity of 128MB, and in terms of pricing the medium is aimed at the mass market. The cost of the drive is currently around £1,400, and the cost per disk around £60.

It should be noted that the cost of optical technology generally follows the same trends as for any other computing products, falling as production quantities increase. The reduction in price for media, however, may be an effective one rather than a direct one — for example, the capacity of a particular type of disk may go up whilst the price stays constant. The cost per MB has therefore dropped.

Most optical drives are currently manufactured in Japan, and some in the USA. Manufacturers of 5.25" drives include Sony, Ricoh, Maxoptix, Panasonic, Sharp and Pinnacle. The biggest manufacturers of the 3.5" drive are IBM, Panasonic and Sony, all of whom anticipate a growing demand for the technology, and all of whom have recently significantly increased their production capacity.

NEW DEVELOPMENTS

The main manufacturers of optical media are Verbatim, Sony and 3M. Verbatim is a subsidiary of Mitsubishi Kasei, and is perhaps best known for floppy disks. They are, however, the world's leading suppliers of optical media and are currently the sole approved suppliers to IBM worldwide. Brian Finch, Optical Programme Manager for the company's UK headquarters, stresses that although Verbatim's largest market by far is the computing industry, they are keen to develop new markets by working closely with OEMs such as pro audio manufacturer DAC, who specialise in providing mass storage devices for sampling and digital audio applications. Finch sees the audio industry as a growing market for optical media, particularly the 3.5" disk, which he describes as 'the diskette of the the future'.

Verbatim's European Optical Product Manager Gerry Kelly adds that with the 5.25" optical, the emphasis is on developing faster access and more capacity, making the disk more suitable for the high-end market. The 3.5" disk, on the other hand, has been designed for the mass market and the emphasis is on reducing its cost.

As far as reducing the time required for preparing and completing the recording process is concerned, Kelly suggests that there are a number of ways in which this can be achieved. By eliminating the verify pass, for example, a significant saving can be made. A more likely solution would be to employ a method which allows the information on the disk to be directly overwritten, but Kelly points out that this is still some way off and will require close cooperation between drive and media manufacturers. Furthermore, he acknowledges the need for backward compatibility such that future drives would need to be capable of at least reading existing disks.

The other drawback of the optical drive, namely the relatively slow access time, is mainly due to the weight of the head mechanism, which limits the speed with which the head can travel across the disk. Kelly suggests that there are ways in which the access time can be improved. One solution, which is currently being developed, is to make the head lighter by splitting the optics such that one or more of the sensing components becomes a fixed part of the drive and is removed from the head.

Once the head has found the correct position above the disk the next delay, in finding the exact location of data, is introduced by the rotational speed of the disk, ie. the time it takes for the particular location to pass underneath the head. In order to decrease this delay, the solution would be to increase the rotational speed of the disks — in fact, the speed of the 5.25" and 3.5" drives has doubled over the past two years and 18 months respectively.

APPLICATIONS

Mark Young of DAC agrees that the main role of optical may not yet be to replace hard disk multi-channel applications, and does not anticipate it being capable of supporting more than four simultaneous record channels for some time to come. However, he is quick to point out the medium's potential for other applications and maintains that access times should be put into perspective. The access time for a hard disk is less than 20ms compared with around 40ms for optical — floppy disk access times, however, can be as much as 20 times slower.

Furthermore, for many applications such as sound library and sequence file storage, the difference in access time between optical and hard disk is not the issue. Once a hard disk is full, for example, its data can be loaded onto tape streamer, which is relatively cheap but inconvenient for retrieval purposes. An optical disk, on the other hand, can be removed and replaced at any time.

An alternative to tape streamer is to use floppies, which certainly make retrieval easier, but have a limited capacity. When compared with the 3.5" optical, for example, a floppy disk costs around £1.50, and can have a capacity of 2MB, whereas the optical costs around £60 per disk and has a capacity of 128MB. It can therefore cost almost twice as much to store the same amount of information on floppy as on optical.

In addition, the optical's robustness must also be taken into consideration — a factor which could have an impact on both time and cost savings. In terms of reliability, Verbatim's Brian Finch also has some points to add. Firstly, an industry standard has been agreed for formatting optical disks, so they should be compatible with whichever drive you care to use. Secondly, he stresses there is no such thing as a head crash with optical — the medium has a very long life, and Verbatim are backing this up with a 40 year guarantee.

DAC supply a variety of optical drives and are currently the only manufacturer to produce a 3.5" drive, the DAC RW4000, for professional audio applications. Mark Young maintains that his customers are already using the system to replace both the hard disk and the floppy for sampling applications and, to a certain extent, disk-based recording. One such user is freelance keyboard technician Tony Smith. Smith's credits include INXS and Spandau Ballet, and he is currently working with Elton John. His equipment includes a Mac Powerbook and SE30, and he maintains that since he works with many different keyboard players he must be able to deal with a range of media such as floppy, 44MB Syquest (removable single-platter hard disk cartridges), hard disk and optical.

Smith's job involves a great deal of touring and it is therefore essential for him to have storage which is robust and portable. He decided to use optical because of its reliability, and chose the 3.5" format because of its compact size. "It's just perfect for me," he says. The drive is cheaper to ship than larger drives, and is fairly easy to carry by hand if necessary. For security purposes Smith can fit the disks in his pocket and does not have to worry about magnetic fields or passing through detection machines.

At the time of writing, Smith had been using the DAC RW4000 alongside an IBM drive for the past six months, and had so far experienced no problems whatsoever. He uses the optical mainly for mass storage of samples, transferring to and from hard disk via SCSI which he uses as his working medium. He is often required to work with different makes of samplers, such as Akai, Casio or Kurzweil, and uses Alchemy software on his SE30 to convert between different file formats. Sounds are then transferred to the optical which can be used by the sampler via SCSI.

Smith acknowledges that the optical is not as fast as the Syquest, but weighs this up against a greater storage capacity, commenting that "the good thing about mass storage is you can store everything on it and find it quickly, provided you have a good disk librarian system". Furthermore, he is convinced that the medium will increase in popularity once computer software becomes available on the 3.5" format.

Another user is drum programmer Paul Brook. He runs the Real Drum Company (RDC), whose Masterkits CD-ROM of drum and percussion sounds has just been launched. His initial interest in optical storage arose out of his requirements for vast amounts of storage space for his drum sample library, on which the Masterkits library is based. He claims that Masterkits provides "the first realistic representation of a drum kit", and some of the kits require as much as 32MB of sample memory. Brook chose the 3.5" optical format for his own use because the capacity was convenient — he organises his kits in two volumes of 60MB each. He maintains that the 5.25" medium, on the other hand, is somewhat unwieldy, adding that it can take longer to find a particular sound. Furthermore, he finds that the 3.5" drive is a natural progression from using floppies.

On the practical side, Brook notes that the disks can be stored in standard floppy boxes, and maintains that it is easier to manage the library with a number of disks, "rather than having all your eggs in one basket" as you would with a larger format. In addition, with both formats there is limited space on the label, thus with a smaller capacity it is easier to list the contents than with a larger one.

Brook uses the 3.5" drive in conjunction with an Akai S1000 sampler to compile, test and demonstrate the library. He finds the robustness of optical media a great advantage: "they're so reliable — every time we know it's going to work, which is important for demos". CD-ROM was chosen as the medium to launch the library on, however, because mass-producing the disks works out cheaper than recording individual disks. However, Brook maintains that for users it would be ideal to run the 3.5" optical in conjunction with the CD-ROM player in order to store customisations. Brook acknowledges that the technology may not be as cheap as everyone would hope, stressing "it's quite a large investment initially, but it pays back and is open-ended for the future".

THE FUTURE

Both Verbatim and DAC are confident of the medium's future success. From past total worldwide sales of drives. Verbatim project that the 5.25" and 3.5" inch markets will respectively double and triple in size over the next year. Furthermore, the new 2X products (announced at the ICC show in January 1993) effectively double the capacity of both 5.25" and 3.5" disks.

DAC are also confident of the medium's future, and are keen to develop the 3.5" market. Their drives will not only support the standard 3.5" MO, but also O-ROM and P-ROM produced by Verbatim as part of a mastering service. The former is similar to CD-ROM in that it is a pre-stamped disc which is read-only and cannot be erased. However, it operates at the same speed at the MO and therefore has a faster access time than CD-ROM. The latter is similar to the O-ROM, but the stands for 'partial'. This means that only part of the disk is pre-stamped — the amount is defined by the customer, with the remainder of the disk being erasable. Such disks seem ideal for commercial and custom sound library purposes. A large library could be offered on O-ROM, whereas a smaller one could be offered on P-ROM with room to spare on the same disk for storage of sounds which have been edited.

With such flexibility, convenience, reliability and relatively low cost, the 3.5" disk looks set to become the standard for library applications, whereas the 5.25" inch, with its emphasis on speed and capacity, looks set to continue meeting the challenge of professional recording/editing applications.

Ultimately however, the success of optical technology will depend on its adoption by the wider market. As Gerry Kelly puts it "everyone wants a multimedia PC — at the moment it doesn't exist, but when it does they'll need optical". Brian Finch agrees and adds that a significant growth in the market will begin when IBM and Apple include a 3.5" drive in their computers as standard. In fact the drive is already optionally available with IBM's PS2 Mode 90 PC — perhaps an indication of the major role optical disk is destined to play in the future.

Further information

DAC, (Contact Details).

Verbatim, (Contact Details).

Optical disks operate in a similar way to hard disks, in that the disk rotates beneath a head which can move back and forth between the centre and the outer edge of the disk. These combined movements mean that the head can locate any position on the disk, and the speed with which it does this will depend on how fast the head can move and how fast the disk rotates. With both 3.5" and 5.25" optical disks the head does not actually touch the surface of the disk — whereas hard disks rely on directly creating and sensing changes in polarity of the magnetic material on the disk, optical works on the principle of sensing the differences in the reflection of a laser beam.

The laser is also used to write the information onto the disk in the first place. The disk is made up of various layers which will generally consist of a protective layer, a reflective layer, a recording layer and a substrate. On writing the information, the power of the laser beam is increased so that the recording layer is significantly heated at the spot at which the beam is aimed. With WORM disks, this causes the spot to be permanently changed in some way and, depending on the method used, will result in a 'pit' or hole (as shown in Figure 1), a bubble or a change in colour — all of which will cause a beam to be reflected differently from the areas of the disk which have not been heated. On reading the information back, a lower-powered beam is aimed at the disk and the reflection of the beam is translated into 1s and 0s, depending on how it is reflected.

With MO disks, the principle is similar — however, instead of using a recording layer which is permanently changed, it consists of material whose magnetic polarisation can be changed when heated to a sufficiently high degree. To start with, the polarisation of the recording layer is in a uniform direction as shown in Figure 2a. In order to change the polarisation of a spot on the recording layer, the laser beam is switched on and heats up the spot. Working alongside the beam is a biasing magnet, and the recording layer under the heated spot takes on the polarisation of the magnet as shown in Figure 2b. As the spot cools, the recording layer maintains this polarisation — when the disk is read, the change in polarisation will cause a change in the way the beam is reflected.

This mechanism can be likened to having tiny blocks of ice, each in their own container and each containing their own arrow. If left alone, the position of the arrow is frozen and cannot be changed. However, if required, a block can be melted, the direction of its arrow switched and the block refrozen.

After the recording process is completed, the information is verified or checked during what is called a 'verify pass'. This adds to the time required for the recording process. Furthermore, unlike hard disk where existing information can simply be overwritten, in order to re-record over the optical, there must first be an erase pass which realigns the polarity of all the spots on the recording layer so that they are once more all in the same direction. This, again, adds to the time required for the recording process.