The audio-visual initial debug (AVID) method is proposed as a first-line software development tool for locating errors in new source code. The concept involves using a computer speech synthesizer connected to the terminal to provide an audio play-back of the source code while the programmer visually follows his original hand-written text. Avid as currently envisioned is specifically designed to analyze and speak Fortran source code but a different version can be built for any computer language. Certain standard and user-defined key words are pronounced while all other symbols are spelled character by character. Speech is accomplished via inexpensive, phoneme-based, add-on equipment with essentially unlimited vocabulary. Speaking rate, volume, frequency, and intonation are user-controlled. In addition the avid software allows the user to position control at any line in his text and move in any direction. In conjunction with the avid software development, an in-depth survey of existing speech synthesis technology and equipment will be completed; a demonstration of AVID's capabilities will be made available; and recommendations for Phase II development and enhancements will be compiled.
