Rumo à sala de estudos aumentada: experiências com suporte computacional para o desenvolvimento técnico e perceptivo na performance musical

Abstract

Digital technology is increasingly present in most music-related activities and its pervasive use in music production and consumption became the absolute standard in the last couple decades. Musical instrument teaching and learning has also been subject to the these changes, with the development of a great number of hardware and software tools, video-games and online communities devoted to it. In the context of western classical music education there is still some resistance to these novel approaches and, as such, interesting opportunities are still to be explored. There is a growing number of studies addressing the issue but few of them explore real-world teaching scenarios. The capacity to be expressive through music is strongly dependent on performers abilities to control subtle modulations of acoustical parameters through their instruments in a conscious and contextualized manner, being the techniques that allow for such control one of the most crucial components of a musicians expertise. Identifying the correlations between slight modulations of sound attributes and the bodily actions that control them is a key technical aspect of musical instrument playing. In the traditional master/apprentice model of learning, this is usually accomplished by means of instruction given through verbal feedback and exemplication, also resorting to metaphors and analogies. Focusing on the development of technical skills for musical expression, we conducted a year-long study with a class of undergraduate clarinet students and professionals at a music school of western classical tradition of a Brazilian university. Two undergraduate subjects were oered, comprising weekly practice sessions with an experienced clarinet teacher and periodical classes with other masters of the instrument. Digital signal processing techniques for Music Information Retrieval were used, allowing for the extraction of several acoustic descriptors from the audio recorded. Through this data we could assess information related to note articulation, timbre, dynamics, timing, and tuning, used to describe several expressiveness-related aspects of a musical performance, and explore how the acquisition of the fore-mentioned skills in a teaching scenario relate to the physical properties of musical sound. The approach applied here is also open to the possibility of using external recordings from other performers as reference by comparing extracted data from it and those of the students. In this thesis we present a series of practical studies related to the role of computational tools and augmented feedback on the development of musical performance. We will be focusing on the development of technical expertise for musical expressiveness from an acoustical perspective. Our experience demonstrated that these techniques can be used to rene the relationship and dialog teachers have with their students while also incorporating novel references to their practice.