In this work, we present the initial design of a modular multi-sensor system for singing voice analysis, and describe its first assessment experiment on the ‘vocal breathiness’ qualitative characteristic. A system case study with two professional singers was conducted, utilizing signals from four sensors.