Personalized Multi-Track Leveling Algorithm

This paper analyzes the importance of music mixing and presents a personalized volume control algorithm. We discuss the constraints of existing multi-track leveling methods and highlight the necessity for a tailored approach that incorporates individual preferences and creative perspectives. The primary objective of this study is to enhance the creative process of mix engineers, musicians and producers, through the application of a customized multi-track scaling algorithm. This method aims to autonomously address the initial phase of the mixing process, namely the loudness control of each track. It achieves this by effectively determining the optimal balance between the various elements of the mix, taking into account the workflow and decision-making processes typically employed by professional mix engineers. The proposed method presents a formulation of the volume leveling as an optimization problem, which uses a customized genetic algorithm that aims to determine the optimal coefficients for the multi-tracks that compose a song.