“…Despite this popularity, Music Information Retrieval (MIR) research on polyphonic music has been rather constrained to other types of musical material such as pop/rock (Ryynänen and Klapuri, 2008;Bittner et al, 2017), jazz (Abeßer et al, 2017;Abeßer and Müller, 2021), or piano music (Sigtia et al, 2016;Nakamura et al, 2018) in the past decades. However, several works have recently addressed singers' intonation and interaction in vocal ensembles (Devaney et al, 2012;Dai and Dixon, 2019;Cuesta et al, 2018;Weißet al, 2019), the analysis of vocal unisons (Cuesta et al, 2018(Cuesta et al, , 2019Chandna et al, 2020), the estimation of multiple fundamental frequency (F0) values from polyphonic vocal performances (Su et al, 2016;Schramm and Benetos, 2017;McLeod et al, 2017;Cuesta et al, 2020), or the separation of singing voices (Gover and Depalle, 2020;Petermann et al, 2020;Sarkar et al, 2020). Most of the tasks mentioned above commonly rely either on analyzing separate audio tracks (stems) for each singer of the ensemble, or on the ground truth individual F0 contours.…”