The optical disk has some useful characteristics that have allowed it to survive alongside magnetic media. John Watkinson takes a look.
The magnetic hard drive has demonstrated extremely high capacity, transfer rate and access speed, but the optical disk carries on because it is not directly comparable and has advantages in some applications.
Optical disks did not come from the computer industry. The first optical disks were developed by Philips and were known as Laservision because they recorded analog video. Being analog they were of no interest to the IT community.
However, developments in digital audio led to the Compact Disc, which had a recording density that made the computer industry take notice. Today when practically every industry is moving towards IT solutions, it is hard to think that the CDROM was designed using audio industry technology. The audio industry did it again with the DAT format, having a storage density that made the computer tapes of the day look silly.
But that was long ago, before everyone had a computer. Now that IT has become universal, there won't be any more dedicated audio, video or data media.
The head to disk spacing in a modern magnetic drive is microscopic and everything has to be sealed to prevent contamination. A removable medium is out of the question. The last removable magnetic disk was the floppy disk, which has been eclipsed by the memory stick.
The optical disk operates by focusing on the medium from a distance. Although that distance is not large in absolute terms, it is huge compared to the clearance in a magnetic drive and that means the drive can survive serious contamination without damage. Optical disks became the dominant removable media.
Some types of optical disk are read only, but they work on phase contrast, which operates entirely on their mechanical shape. Such disks can be cheaply and simply replicated by pressing, just as the vinyl audio disk was. Optical disks are useful where the same data need to be distributed to a large audience and are used for pre-recorded music, movies and software. Such applications do not need rapid random access. Optical disks can provide random access, but for reasons that will become clear, they are not as fast as hard drives. Applications such as audio and video playback and software loading tend to be linear.
Magnetic disks and vinyl disks both tend to run at constant rpm. This is known as constant angular velocity or CAV and it means that the speed of the track varies with radius. The bandwidth or information capacity also varies with radius, but the hard drive lives with that and reduces the storage density on outer radii by keeping the clock speeds constant. The constant rpm allowed the heads easily to seek from one radius to another because the bit rate is independent of radius. To minimize the range of track speeds, the magnetic disk records tracks over a band some way from the spindle.
Optical disks took another route. Unlike the hard drive, optical disks use a single spiral track per surface, so there is no concept of a cylinder in an optical drive. To optimize capacity in linear applications, it is the track speed that is held constant. This is known as constant linear velocity or CLV. CLV requires the spindle speed to change according to radius and it means that data can be recorded over a much greater range of radii than is possible in the hard drive, by going in close to the spindle.
The greater range of radii and the increase in density from the use of CLV increases the capacity of optical disks to compensate for their lower superficial density compared to magnetics. If a CLV optical disk needs to seek to a different radius, it will also have to change the spindle speed. Another issue is that it is not possible accurately to seek to a spiral track, so a seek would consist of a coarse track crossing followed by reading the track to find out where the pickup has arrived, followed by a correction seek.
The track contains regularly spaced headers so that the pickup always knows where it is and can compare that with where it should be. This takes longer than the deterministic seeks of the cylinder-based hard drive.
The optical disk resists contamination in a way that is not available to the magnetic drive. The lens has a large numerical aperture (NA) in order to resolve smaller bits, and this means that the light focusses down onto a spot on the data surface over a large angle. Fig.1 shows that the data surface is not on the front of the disk, but is on the back, so light from the pickup has to travel through the thickness of the disk. This means that where the light enters the disk it is a long way out of focus and enters though a large area, diminishing the impact of dust and debris.
Fig.1 - The light from the pickup arrives at the surface of a CD over an angle of about 27 degrees. Refraction at the surface changes that to 17 degrees inside the disk, which makes the entry circle about 0.7mm in diameter. This is very much greater than the size of a bit on the disk and dust particles on the disk don't obstruct the light.
The resistance to contamination is completed by the use of a powerful error correction system. All of the other components were in place and the Compact Disc became viable when real-time Reed-Solomon error correction became possible at consumer prices.
As there is no physical contact between the pickup and the disk, the optical drive needs an active focusing system and an active track-following system. The use of CLV requires active control of the spindle. Disks will never be flat and the hole will never be precisely in the middle, so the focus and tracking servos are designed to deal with runout. Optical disks usually start playing from the center, rather than from the outside like a vinyl disk. There will be less runout near the spindle so the pickup can acquire focus more easily.
The operation of the three servos in the optical drive puts constraints on the channel coding that are not seen in magnetic drives. In attempting to turn the spindle at the right rpm, which has no constant correct value, the CLV servo may analyze the run lengths in the data signal. As the channel code is run length limited, there are constraints on the maximum and minimum times between transitions. If the drive sees a minimum run length that is too short, the track is going too fast, whereas if it sees a maximum run length that is too long, the track is going too slowly. Without run length limits, that would not be possible.
Focus and tracking is done using the same pickup that reads the data. This means that the channel code must be DC-free so that the data modulation does not interfere with the servos. With a DC-free code the servos simply contain a low-pass filter that rejects the data signal and leaves only the positioning information.
Once the track is approximately at the right speed, the data separator will achieve phase lock with the clock content of the channel code and the readout bit rate will be accurately known.
There are two ways of controlling the data rate. Using a time base corrector, data can be read out of memory at a constant rate, for example the sampling rate of an audio disk. If the disk track is running too slowly, the memory will tend to empty, whereas if the track is going too fast, the memory will fill up. It is thus possible to adapt the spindle speed to keep the average contents of the TBC constant and this can be done by comparing the read and write addresses.
In video disks such as DVD and BluRay, the pictures use a variable compression factor. Instead of the quality going down when a difficult scene is compressed, the bit rate goes up instead. This means that the optical drive has to support a variable bit rate. That is difficult for the data separator because it is harder to lock to the clock content of a variable rate signal. Instead the drive uses skipping. The linear speed corresponds to the highest data rate needed, and whenever the actual date rate is lower, the timebase corrector is prevented from overflowing by skipping the drive back like a stuck vinyl disk so it plays the same piece of track over again until more space has opened up in the TBC.
Using headers, every data block can uniquely be identified, it is possible to create a contiguous file in memory even if it was not played all at once. A small variation on the technology allows the drive to continue working in conditions of shock or vibration, such as in hand-held or automotive applications.
Drives of this kind have a larger than usual time base corrector and if the pickup is knocked off track, data continues to be read from the TBC while the pickup gets itself back to the right track again. Using the headers, the data from before the skip can seamlessly be joined to the data after the skip.
You might also like...
The explosion in digital technology that led to Compact Discs, DVD, personal computers, digital cameras, the Internet and digital television broadcasting relies heavily on a small number of enabling technologies, one of which is the use of Reed-Solomon error correcting…
The first burst error correcting code was the Fire Code, which was once widely used on hard disk drives. Here we look at how it works and how it was used.
The CRC (cyclic redundancy check) was primarily an error detector, but it did allow some early error correction systems to be implemented. There are many different CRCs but they all work in much the same way, which is that the…
The mathematics of finite fields and sequences seems to be a long way from everyday life, but it happens in the background every time we use a computer and without it, an explanation of modern error correction cannot be given.
Here we look at one of the first practical error-correcting codes to find wide usage. Richard Hamming worked with early computers and became frustrated when errors made them crash. The rest is history.