This article treats mixing as a mathematical optimization problem and embarks on creating a model that balances the different dynamics of an audio mix. The model seeks to create a trade-off between the artistic objectives, the auditory perceptual constraints, and the engineering methods. The researchers indicate that the equations found in the mathematic models are beneficial as they help the engineers arrive at solutions. The paper centers primarily on the numerical optimization of the mixing process based focusing on the non-linearity of auditory perception (Terrell, Simpson & Sandlers, 2014). The paper first describes the different metrics in the optimization process and focuses on using loudness to describe the selected mix using the loudness of the whole mix and the loudness balance vector. The researchers’ method comprised mainly of two components; the optimization algorithm and the model to use to choose the metrics of the target mix (mix loudness and loudness balance). The authors then used a case study to provide some insights into how human perceive an audio mix and to show how automatic mixing is done.
The researcher found that it would be difficult for a human to adjust simultaneously the multiple faders and maintain information of the numerous possible channels that can be found in a typical audio mixing scenario (Terrell, Simpson & Sandlers, 2014). The researchers also indicate that finding the best-fit parameters based on multiple acoustic conditions and other since it would be difficult to compare the different audio perceptions including the mixer’s own perception. Another problem would be the fact that it would be difficult to sample the various audience scenarios. With a proper understanding of the human mixing process, the study came up with a best-fit model that simulate the ability of a listener to listen to different scenarios ant the same time.
Reference
Terrell, M., Simpson, A. & Sandlers, M. (2014). The Mathematics of Mixing. Journal of the Audio Engineering Society, 62 (1), 4-13.