number of practitioners wrote, largely for the student, chapters on their own area of expertise. One of us (RCB) was the instigator and editor of this book, and so again it was tempting to assume that we had slain the dragon of ignorance, and there is now no barrier to the brave new world. Added to the book most students have access to the InterNet, with its additional material, software updates, and opportunities for revisions of the text. One is still tempted to ask, What can go wrong?

We have experience of teaching beginning students. They are not empty vessels into which we pore knowledge. They come with an experience. Many of the students who arrive in synthesis classes have already been exposed to the sounds of the DJs, Collage Artists, Turntableists, Remix Artists, Hip Hop and Ambient Scene, and come with expectations. They have had synthesisers for years, and are used to plugging in the MIDI keyboard and the sequencer, and immediately getting sounds with which to make music. The concern here is that the initial enthusiasm may die before the skills and excitement in crafting detailed musical sounds and moments are fully developed.

We are expecting the students to learn how to create sounds from absolute basics, and it is a hard task, and inevitable will be slow for most students.

In this paper we propose three ways in which we can seek to alleviate some of these problems; use of physical modelling, use of Worldwide Web interfaces, and issues of performance.

Hardware Synthesizers and Software Synthesis

Among the things that make hardware synthesisers so easy to use are the way in which they have packages synthesis techniques which were explored slightly earlier in software synthesis. These are made available at the push of a button, as presets. For example the FM synthesis technique²

became the basis of much popular music through Yamaha synthesisers.

Using this as our clue, we have created a structured program of presets which are designed to engage the student’s imagination and to develop enthusiasm. This has three levels.

Level 1: The creation of a number of unit generators based on physical models

Level 2: The creation of instruments built from simpler opcodes for the same model.

Level 3: Access to C code to implement the models.

Level 1 – Physical Model Opcodes

There are a large number of ways in which sounds can be created, but the family of physical models have the advantage that they are quickly related to our experience, and as such the controls over the sounds are ,natural‘. It is for this reason that we