"Digital Sound Recording and Reproduction"

This event was originally planned as a visit to 'Planet Roland' in Leeds for demonstrations and a talk on digital sound recording and reproduction. However, unforeseen circumstances meant that the visit was not feasible and Mark was not able to participate in the event. Glyn very kindly offered an alternative talk on the same subject and Leeds Metropolitan University were equally generous in allowing us to use one of their lecture halls as the venue.

Glyn began his talk with an overview of the nature of sound, reminding us that sound is vibrations (waveforms) in the air. The human ear detects these vibrations and converts them into signals which are processed by the brain. These vibrations can be very complex. It should also be remembered that sound is analogue by its nature. The seven senses of hearing were outlined: frequency; loudness; dynamics; transients; harmony; timbre; and direction. The faithful capture and reproduction of all these is a challenge for the recording process. There is also a certain amount of psychology involved as the same sound waves will be 'heard' by different people in different ways.

We then had a brief history of recording; from human memory (the earliest, and at one time the only, way of recording sound) through the phonograph and electromechanical/electromagnetic recording through to CD, DVD and solid-state devices familiar to us today. A discussion of microphone techniques reminded us that the method of sound recording used is just one factor; the type of microphones and their placement during a recording session is of paramount importance.

The workings of a simple Analogue to Digital Converter (ADC) were then explained. Basically, this converts an analogue signal into a binary representation of the waveform at a point in time. The number of bits used in the binary representation, together with the frequency with which the conversion is done (the sampling frequency) determine how accurately the waveform is represented by the succession of binary numbers. For example, standard CD audio uses a resolution of 16 bits and a sampling frequency of 44.1kHz.

A number of different digital audio formats have appeared in recent years to address the fact the standard digital recording requires a lot of storage space; one minute of CD audio requires about 10MB. Smaller files can be obtained by compressing, which uses algorithms to remove 'redundant' information from the sound file; this assumes that some of the information in the file encodes sound that is inaudible. Whether the resulting compressed file is still a faithful representation of the recorded sound continues to be debated among hi-fi enthusiasts, music lovers and technologists! To some extent this is subjective (the 'psychological' aspect of sound recording mentioned earlier) and different levels of compression may be appropriate for different applications. The difference between CD audio and the same track as an MP3 may be easily discernible if listening in a quiet environment on high-quality playback equipment, but much less (if at all) distinguishable when listening in the car or outdoors on a portable device.

We finished with a discussion of some of the equipment used by recording engineers; one illustration showed that a typical portable sound recording rig could be seen as a 'PC in disguise' with a specialised control panel and some additional specialist hardware. Basic digital audio mastering (as with basic digital video editing) could indeed be done on a typical PC. More specialised equipment allows more control over mastering of the digital signal, such as equalisation, dynamics (the dynamic range), and 'cut and paste' editing. As with the earlier comments about microphones, the mastering that is carried out on the recorded sound has a significant effect on the final result irrespective of the recording format used.