Teaching Audio Processing Software Development in the Web Browser

摘要

Web-based learning progresses from lectures and videos to hands-on software development problems that are to be solved interactively in the browser. This work looks into the technical underpinnings required and available to support teaching the mathematics of audio processing in this fashion. All intensive computations are to happen on the client to minimize the amount of data transfer and the computational load on the server. This setup requires editing source code and executing it in the browser, dealing with the audio computation in the browser, playing back and visualizing computed waveforms. Validating the user's solution cannot be done by a sample-by-sample comparison with a "correct" result but requires a tolerant comparison based on psychoacoustic features.