CD-ROM, 2000, US$ 27.00; Norwegian Network for Technology, Acoustics and Music (NoTAM), P.O. Box 1137, Blindern, N-0317, Oslo, Norway; World Wide Web www.notam.uio.no/DSP/; available from CDeMUSIC, 116 North Lake Avenue, Albany, New York 12206, USA; telephone (888) 749-9998; (518) 434-4110; fax (518) 434-0308; electronic mail emf@emf.org; World Wide Web www.cdemusic.org/.

Reviewed by Bill Sack
Buffalo, New York, USA

Very little is currently available in the field of sound creation software for young people, and the few examples that come to mind are little more than toys. Enter DSP, a feature-filled and educational computer music program for children.

The first version of DSP was developed by Jøran Rudi at the Norwegian Network for Technology, Acoustics and Music (NoTAM) in 1997 to teach computer music techniques and repertory to children in Norway. The product under review is the second version, which has been translated into other languages, including English. DSP is a standalone program consisting of software tools for the creation and manipulation of sounds, plus extensive hypertext documentation about the program as well as short tutorials on the physics of sound and the basics of digital audio.

DSP can install and run on any IBM-compatible PC with a 486 or better processor running Windows 95, 98, or 2000. Hardware requirements are a CD-ROM drive, a graphics card capable of displaying thousands of colors, and a 16-bit sound card. The program makes modest system demands and can run on the sort of aging hardware that might be encountered in publicly funded schools. It installs from its CD through the familiar Windows installation dialogue. A copy of Microsoft's Internet Explorer 5 is included; the program needs it to display documentation. The installation process will upgrade older versions or can be skipped if an equivalent or newer version is already present. The DSP executable file is copied onto the machine's hard drive, while the documentation and demonstration files remain on the CD. However, and contrary to the program booklet, the CD must be in the drive in order for the program to run.

DSP's graphics have a stylized, "aged" look reminiscent of Myst, a popular computer game from several years ago. The program's widgets and buttons are witty and clever: instead of a static trash can icon, there is an animated fish with exaggeratedly large teeth that "eats" unwanted soundclips which are dragged to it. This part of the program seems to have been designed to appeal to younger computer users whose previous experience with computers may have been through gaming.

The main interface is the Mix window, a five track timeline on which monaural sounds can be easily arranged by dragging them around with a mouse (see Figure 1 at right). Volume and pan settings for each track can be set using graphic breakpoint envelopes, and the result can then be played back through the audio output or mixed down to a stereo 16-bit WAV file.

Functions for recording, editing, and synthesis are located in submenus of the "Sound" menu. The integrated sound editor can be used to record new files through any sound card input. It has buttons for adjusting the volume of a sound, fading in and out, copying, cutting, and pasting. Unfortunately, only one editor window can be open at a time, making it impossible to cut and paste between two different sounds. Available synthesis functions include FM, Karplus-Strong plucked string, two-formant filtered buzz, eight-partial additive, and white noise, all with parameters controlled by graphic breakpoint envelopes.

Some of the most interesting features of DSP are available through the "Distort" menu. Tools here include Granulate (granular processing), Time Stretch, Scratch (forward and reverse variable speed playback), Spectral Shift, Spectral Sieve, and Algorithms. Spectral Sieve and Spectral Shift are FFT-resynthesis tools; the former filters a sound by selectively removing spectral components and is useful for certain types of noise reduction, while the latter produces drastically altered sounds by shifting analyzed partials up or down. The Algorithms submenu provides functions for a simple algorithmic composition program (see Figure 2). One of four algorithms can be chosen to control the frequencies of the resulting melody: Normal distribution (random, but mostly in the center of the range), Brownian motion, Fractal (1/f noise), and Chaos (the Verhulst Equation).
A more standard array of sound processors can be found within the "Effects" menu: a multi-function filter, chorus, ring modulation, harmonizer, reverberation, and delay. The parameters for each of these, as well as for the "Distortions," are controlled by the same type of time-value breakpoint envelopes used in the Mix window. There is no way to enter precise numeric values to any of these parameters—everything is accomplished graphically. I do not see this as a drawback to the program, but rather as a feature: an intuitive, "sound-oriented" approach to composition is encouraged, and the resulting music will sound very different from that made with sequencers, samplers, or other software tools. I only wish that there was a way to delete a breakpoint. An inadvertent click of the mouse can put an unneeded breakpoint into a graph, and once created it cannot be removed.

Projects in DSP are saved as a collection of sound files and a single text file which contains data about their time positions, pan, and mix values. Sound files which have been created but not yet needed in the Mix window can be stored with the project for later use. All effects and distortions are similarly non-destructive, which is to say that the resulting sound is automatically saved with the project as a new file with a unique name. This is not the case with the editor, however. It is possible to duplicate a sound and then perform some editing operation on the copy, but there is no "Save As..." option. Any action taken on a sound in the editor is irrevocable once the "OK" button is pushed.

The sound tools are fairly powerful and fun to use, but the tutorial program and help files are the most impressive parts of the program. The centerpiece of the documentation is a two-and-a-half minute composition by Jøran Rudi titled And the Birds...?. An accompanying text tells the story of how the piece came to be written, and describes, section by section, the techniques used in the creation and modification of the sounds. The project file for And the Birds...?, as well as all of its soundfiles, are included with the documentation. A good way to become acquainted with the program is to open this project, disassemble it, change parts of it, and put it back together in a new way.

In addition to this demonstration, there is a wealth of context-sensitive help. Clicking on the "question mark" button in any of the windows opens a browser that describes in clear, non-technical terms what can be done in that window. Additionally, there are links in the text that open examples of the function in use; the parameters in these can be changed and tried out by the user. The documentation even includes examples from the computer music repertory; for example, the help file for "Brownian motion" includes a 30-sec excerpt from Iannis Xenakis's La Légende d'Eer!

I found the tutorial aspect of this program to be very well thought out, and it would make a wonderful addition to the curriculum of a junior-high or high-school general music class (is there such a thing in the United States anymore?). DSP would even be useful in entry-level college computer music courses as a supplement to the standard texts and listening. Aside from its use in schools, it could provide a life-changing experience to a young person interested in the composition of music with sounds, but without previous access to the necessary tools, language, or ideas. What prior generations of young proto-composers accomplished with old tape recorders, current and future generations could do with programs like DSP.