DAFX: Digital Audio Effects, Second Edition

13.5 Conclusion

In this chapter we have introduced AM-DAFX; a class of digital audio effects which can do automatic mixing. As well as demonstrating a number of examples, a framework has been presented into which future AM-DAFX can be placed. The system architecture utilises cross-adaptive processing of features extracted from the input signals. Depending on the feature extraction mechanisms, the AM-DAFX have been classed as either dynamic or accumulative. Optimising the accuracy of the feature extraction mechanism can significantly improve the performance of the AM-DAFX.

The intention of AM-DAFX is to aid or replace certain tasks which are normally undertaken by the user. Only time will tell how autonomous a DAFX will become, and to what extent they will be accepted by the user. AM-DAFX is at present a growing field of research and several commercial devices based on such principles have emerged. The use of different configurations and topologies in the implementation remains to be explored. The AM-DAFX proposed here deal mainly with technical mixing constraints, and are meant to be used as a tool which allows the sound engineer to concentrate on more creative aspects of the mix. In the future, more extensive use of perceptual models may not only improve the performance of AM-DAFX, but may also allow more subjective mixing decisions to be explored. Research into these areas is, however, likely to remain controversial.

Video examples of the implementations mentioned in this chapter ...