Technology For Real-Time Visual Feedback In Singing Lessons

Abstract

Computer technology has advanced to the point where audio analysis techniques, previously the hallowed preserve of speech science and engineering research laboratories, are now available to anyone who makes use of a modern PC machine in the home or office. Much of this analysis can be carried out in real-time. Thus it is currently perfectly feasible to make available real-time visual feedback in singing lessons, and a number of systems are now available. There is a considerable body of (well-established) literature describing quantifiable changes that occur when the singing voice is trained in a context within which visual feedback can be offered. This paper discusses voice production and those elements that can be quantified using today's multimedia PC machines, the use of real-time visual feedback in singing lessons and the importance of some understanding of the principles behind typical analysis algorithms employed in such systems. It is shown that real-time displays are useful in singing lessons. Displays should be clear and unambiguous and users must be aware of what is expected of them. User-controlled parameters should be altered with caution with an informed expectation of what the output might look like and the impact of any errors that are likely to occur.