Towards a Parallel Computing Framework for Direct Sonification of Multivariate Chronological Data

Abstract

This paper presents a generic and scalable framework for direct sonification of large multivariate data sets with an explicit time dimension. As digitalization and the process of data collection gathers momentum in many fields of human activity, such large data sets with many dimensions of different data types are common. The specificity of our framework is uniformness of the synthesis technique on different temporal scales achieved by using direct sonification of particular data rows in corresponding sound grains. This way, both distinctiveness of individual data rows and patterns on the higher scale should become perceivable in the synthesized audio content. In order to attain scalability, the implementation relies on parallel computing.