Computers in music are now everywhere and they have opened up new possibilities for amateur and professional musicians alike. Computers have profoundly changed production methods in the music and entertainment industries, permeating much of the output of radio, TV, video, film, CD and even live performance. Self-contained music computers like the Yamaha QY70 costing just Pounds 300-400 can be programmed in a bedroom to do almost everything needed to develop ideas in popular music, produce a top ten single, or to accompany a live performance. ÁñÁ«ÊÓƵ computers with additional hardware and software can manipulate full orchestral scores or explore the outer reaches of electroacoustic music. Computers also now have considerable importance in music education, and computer generated music has important roles in new areas such as websites, multimedia products, computer games and virtual environments. The democratisation of computer music is well under way - and we are probably just at the beginning of a long series of new developments.

Of course, there are dangers as well as opportunities. New tools alter the kinds of music that get composed and produced, but a particular danger can come when the gap between a musician's intuition and the knowledge needed to realise that intuition gets too wide: there may be temptations to take whatever the machine can provide most easily. Curt Roads' Computer Music Tutorial is intended to combat exactly this problem, among others. The former editor of the Computer Music Journal, with several colleagues, has distilled years of scholarly research and practical experience into a meticulously written teaching resource. The book is aimed principally at musicians, especially composers and performers. It does not aim to turn musicians into engineers, nor to act as a detailed "how to" guide for current systems. Instead it goes back to fundamentals and takes the musician through the basics, focusing comprehensively on digital audio, sound synthesis, signal processing, sound analysis and music programming.

A strong section on the musician's interface deals with musical input devices (with copious examples), and also covers performance software, music editors, music programming languages and tools and strategies for musical knowledge representation. Psychoacoustics is covered, though not the psychology of music as such. Sensibly, system interconnection has its own section, and there is a long appendix on Fourier analysis. This book will provide computer musicians with a sound foundation for dealing with not just the current generation of tools, but probably a few future generations as well.

This book also fulfils a key role in educating the next generation of engineers, tool designers, and researchers. Where designers fail to think about fundamental principles, poor design features in existing tools are sometimes imitated and perpetuated. A case in point is the present frequency range of CDs, less than that of the best analogue records and tape recorders, leading to a loss of information, "air", above 22KHz, which surprisingly many people can hear. For budding serious computer music composers, performers and development engineers, this book will be the essential teaching text.

ÁñÁ«ÊÓƵ

There are things that this book does not attempt to do, and it certainly does not cover all of the fundamental areas of concern routinely covered by the Computer Music Journal. This is not a criticism, simply an acknowledgement that no single book can cover all of the fundamentals. The expert, up-to-date coverage of the original staples of computer music - sound generation and processing - is unsurpassed. For example, physical modelling gets its own chapter as just one of six chapters in a comprehensive section on synthesis principles. On the other hand, music cognition gets little coverage in its own right. As computer music has matured, there has been an increased emphasis by researchers on understanding how music is perceived and processed by the human mind. This has played a vital role in influencing sound design, musician interface design, editors, composition systems, and systems that can listen to and interact with human musicians. This emphasis is reflected in the sections dealing with sound analysis, the musician interface and psychoacoustics, but wider implications of music cognition for computer music are not considered in depth. Interdisciplinary researchers principally interested in specialised areas of computer music will need to read more widely. This applies in the case of specialised concerns such as: applications of the cognitive psychology of music, advanced musician machine interface design, hyperinstrument design, artificial intelligence and music, advanced music representation, computers and musicology, and computers and music in education. In any book this size there is room for one or two niggles. A heroic attempt at briefly presenting relevant computer science concepts for musicians sometimes stresses older views - for example the algorithmic view of programming as problem solving is stressed rather than the declarative view of programming as system description and construction - although object technology and logic programming are both covered, along with Lisp. Some topics could be made more accessible: for example, opportunities are missed in explaining in simple ways techniques such as oversampling and pitch shifting. The explanation of pitch shifting rapidly descends to implementation-dependent technicalities like address pointers that are ultimately irrelevant.

The book aims to be international in outlook and this is mostly successful. Work from Europe and around the world (as opposed to from America) is well represented, if not always prominently. But such cavils are few, and they are very minor.

ÁñÁ«ÊÓƵ

This will undoubtedly and deservedly become a standard text in studios, conservatoires, music departments and computer music labs all over the world.

Simon Holland is a lecturer in computer science at the Open University, with research interests in knowledge media and music.