This software project provides Internet-delivery of an archive containing more than twelve hundred graded listening examples for the aural skills curriculum. A Windows PC/Macintosh-compatible interface provides a rich aural/graphic environment with General MIDI instrumental sounds rendered through Yamaha MIDXtra soft-synthesizer technology. The comprehensive repertoire for dictation includes exercises on basic rudiments (intervals, chords and scales), melodies, four-part harmonic settings, and varied textures from musical literature. The listening examples are musically shaped performances rendered in MIDI format along with high-resolution score graphics. Screen images from the software are shown in Figure 1.

The presentation outlined below addresses pedagogical and design issues considered in realizing the project. It documents the evolving distribution formats used during the development of the software, provides sample Director Lingo scripts, and outlines strategies for creating a self-starting CD-ROM that can access the Web for unlimited expandability.

The underlying pedagogy for the application can be outlined as follows:

• Media for study should be as rich as possible.

• Musical examples for study should be performed in a musical way.
• The computer can be a practical tool for delivering vast amounts of rich media for aural dictation.
• The media for study – the actual dictation exercises – should hold up to close scrutiny.

• The student’s task is aural dictation.

• In aural dictation, the task of writing down music (yes, with pencil and paper) provides benefits that may be diminished in some computer-assisted instructional models.
• Formatting electronic media is the task of the instructor / developer (and not the student). Some CAI models preoccupy the student with the formatting of answers for grading, deflecting attention from the primary task of aural dictation.
• Computing tasks should be as transparent as possible. A very simple interface with hot keys for starting and stopping sound help assure that the student can focus on the primary task of writing down the music heard.

• The best authority for choosing assignments and providing feedback is the instructor (not the computer).

• Grading is more accurate and informative when done by the instructor (rather than the computer). This enhances the student’s experience in aural dictation work.
• Random generation of assignments and computer grading based on student input mitigates some valuable opportunities for learning, such as previewing assignments in class and going over assignments in class after they are graded. The object of study – the actual dictation material – is vaporous. With little priority placed on the actual materials for listening, quality of the student’s experience may be diminished.

The Macromedia Director authoring platform combined with the Yamaha MIDXtra soft-synthesizer provided a flexible medium for building Windows / Macintosh-compatible aural dictation software. With instrumental timbres that rival those of many stand-alone MIDI synthesizers, the MIDXtra soft-synthesizer helped insure that students doing lab work at remote locations with varied computer setups had access to similar high-quality musical sound. The MIDXtra soft-synthesizer has a complete set of Lingo commands for controlling MIDI functions.