Abstract :
[en] Previous research on vocal pitch accuracy revealed insights into the fundamentals of singing. However, most of the research on singing focused on the analysis of single voices, whereas few aYempts have been made to tackle the challenge of analyzing mulTtrack recordings of singing ensembles. In addiTon, singers have to adjust their way of singing with respect to a given venue’s acousTcal environment (e.g., small room vs. a comparaTvely large space like a church). If it is common that musical performances are greatly influenced by room acousTcs, studies on the effects of room acousTcal features during ensemble singing are rare. In order to invesTgate singing performances across various condiTons, we manipulated the singing condiTon (unison, canon, solo) as well as the acousTcal feedback by applying diverging virtual rooms. Three duets with female singers (N = 6) were asked to sing three different melodies using headset microphones to record each singer separately. Recordings took place in the communicaTon acousTc simulator (CAS) at the House of Hearing (Oldenburg, Germany) to be able to provide different simulated acousTcal spaces (i.e., cathedral, classroom, and dry condiTon) to the singers. ObjecTve measures were performed on each recording and confirmed that the singers sang the melodies with high precision (small pitch interval deviaTons) hardly affected by singing condiTons or by the type of acousTcal feedback. However, the singers tended to driH (larger deviaTons of the tonal center) when singing in canon compared to solo and unison singing. Overall, the analysis of the pitch accuracy showed a general effect of condiTon (i.e., unison, canon, solo), but no general effect of acousTcal feedback and no interacTon between the two variables under study.
