Detecting expressive identity of orchestra conductors
The aim of this research is to help musicologists in the expressivity analysis of orchestral performances. Our proposal is based on the explotation of the musical worm proposed by Widmer adapted to the orchestral music specificities. Assuming this worm is one of the most succesful representations of musical expressivity in terms of computational musicology, we detect the representative differences from different recordings of the same piece in order to identify common and salience tendencies. We apply this methodology to both historic and modern recordings, so, the processes here presented need to be robust in front of different recording techniques and qualitites.
Playing the unplayable
The goal of this research is to create models describing physical and acoustic behavior of playable instruments from museums and conservatoires to predict the sound that could be emitted by unplayable instruments. Our plan is to include machine learning and music information retrieval methodoligiess to the knowledge accumulated by luthiers and curators.
This research is performed in collaboration with Toros Ufuk Senan (SMC student @ UPF), Marco-Antonio Pérez (reserarcher at the Department of Strength of Materians al Structural Engineering @ UPC), Paul Poletti (Luthier and teacher @ ESMUC) and Oriol Rossinyol (Museu de la Música de Barcelona).
This project is partly funded by the WoodMusick COST action.
The safe concert
The aim of this research is to analyze the acoustic level and spectral distribution from sounds generated by the different sections in the orchestra and propose a physical redistribution of the musicians in the stage to prevent future hearing loss problems. This redistribution need to be transparent by the audience and has to be sensible to different ensembles and historical periods.
Gesture analysis in violin performances.
The aim of this research is to help determining whether jazz-violin students perform better (rhitmically speaking) than classic-violin students, using traditional MIR techniques. The pedagogic view of this research is performed in parallel by a PhD researcher. Our particular goal is to find relationships in rhythm recorded in different sessions by different students. Data recorded comes from the audio (both microphone and pickup) and a set of position sensors attached to the bow and the violin. Some state of the art descriptors are computed from this data, and machine learning algorithms are applied to find the rhythmic relationship between students.
More information here.
Room acoustics in churches for oral tradition polyphonic chants.
The aim of this work is to verify wether, for the oral transmitted traditional music research, ethnomusicologic view is not sufficient and it requires some empirical knowledge of the environment. In other words, we propose a multidisciplinar study to establish a relationship between the sound ideal and the analyzed sounds. This research focuses on room acoustics of different churches and oratories to determine how they can influence to the oral tradition polyphonic chants performed there.
More information here.