Compact Disc - The Inside Story

Part 1 - General Principles

[Contents] [Next Part]

The Compact Disc revealed itself to the world in mid 1982, heralding a new age in sound reproduction for the consumer. A marriage of precision manufacturing, optics, lasers, servo control, digital and analog technologies, it represented one of the leading edges of human scientific achievement. The system was developed jointly by Philips and Sony, to provide a permanent storage medium of high capacity and quality in a convenient package.

This series of articles will cover most of the topics involved in CD; the disc medium and manufacturing, laser optics, digital processing, analog conversion, filtering, specifications, sub-codes and servo control. As this is such a hi-tech device, I cannot help using scientific terminology at times. I hope you will understand (emotionally and technically). Lets make a start.

The original audio signal is converted from a continuous electrical wave into a digital form. The digital signal is then stored on a spiral track which winds its way from inside to out, around the rotating disc. The only recorded information on the disc is the presence or absence of tiny "bumps" along the track. The player follows this track using a laser beam, which has to merely discriminate the two states of the signal. Very complex electronics keep the laser following the track and converts the signal back into analog form. Sounds simple? No; each link in the chain is very refined in its innovation, and of course, each has its own characteristic problems.

An electrical wave consists of a voltage level that varies over time. If this voltage level is measured at fixed time intervals and recorded, it should be possible to reproduce the original wave at a later date by reversing the process. The accuracy of the reproduction is determined by the number of voltage levels measurable and how often it is done. CD has 65,536 discrete levels on a linear scale, and samples 44,100 times per second. The binary number system (base 2) is used to represent the levels, 16 bits or places correspond to the 65,536 (decimal) levels. By theory, a sine wave of 22,050 Hertz can be reproduced in this manner. Obviously, the lower frequency and higher level signals are reproduced more faithfully.

A little calculation shows CD stereo requires 1,411,200 bits per second to be recorded. The audio bits are group together into frames in a rather complicated manner; with extra bits added for error correction and control information. The frames are then further encoded using EFM (Eight to Fourteen Modulation) and ultimately appear on the disc as bump transistions which occur between 694 and 2,545 nanoseconds (thousandth of a millionth) apart, in 231 ns steps. This encoding will be discussed in detail in a later part.

In the next part I will begin getting down to the nitty-gritty, with details on the physical disc.

[Contents] [Next Part]

Originally published in MAC Audio News No. 156, November 1987, pp 14-15.

Copyright © 1987 Glenn Baddeley. cd1.html was last updated 2 December 1996.