Healthy Strings: AI-Powered Posture Tracking for Violinists
🎻 Violin Posture, Sound, Performance & Technology
Healthy Strings is an interdisciplinary research project investigating posture, movement, playing technique, sound, and physical strain in violin performance. The aim is to develop an open, transparent, and practical method for recording violinists with standard consumer cameras and analysing their movements without markers attached to the body.
The project brings together music physiology, violin pedagogy, physiotherapy, musical acoustics, music psychology, movement analysis, sensor technology, biorobotics, and data science. Rather than prescribing a single “correct” posture, it investigates how instrument-specific functional asymmetries can be distinguished from postural and movement patterns that may increase physical strain.
Research Aims
The project is gradually developing a standardised protocol for video and audio recordings of violinists. Markerless motion data will be combined with expert assessments from physiotherapy and instrumental pedagogy, as well as with acoustic parameters. In the long term, this work is intended to provide a scientifically grounded basis for research, teaching, prevention, and individual feedback.
Key Research Questions
A central question is which small set of scientifically and practically relevant variables can meaningfully connect posture, movement, playing technique, sound, and physical strain. The project also investigates which asymmetries are functional, which movement patterns might contribute to strain over time, which camera perspectives provide reliable measurements, and how movement strategies relate to intonation, dynamics, and timbre.
Methods and Technical Approach
The methodological development is based on the SInES Tools developed at the University of Vienna. Machine-based pose tracking with BlazePose reconstructs three-dimensional body landmarks and movement models from standard video recordings. The analyses include joint angles, movement trajectories, velocities, accelerations, and selected postural parameters. The results are compared with synchronised audio data, expert assessments and, where available, reference data obtained using Noraxon systems, motion capture, and electromyography (EMG). As the research design is still being developed, the recording protocols, variables, and validation procedures will be further refined through pilot studies.
Current Project Status
Pilot recordings are being used to compare different camera positions and recording configurations. Initial methodological comparisons show good agreement for large-scale movements of the upper body and arms. Small rotations, particularly of the forearm and wrist, remain an important focus for further validation.
The next stage will further standardise the recording protocol, multimodal synchronisation, and assessment criteria. In 2026, an international research retreat in Lucca and Pisa will support the joint development of methods, the collection of standardised recordings, and the preparation of further research projects and funding proposals.
Future Applications
- Accessible movement analysis using standard cameras or smartphones.
- Transparent and comprehensible feedback to support healthy and efficient practice.
- Applications in violin pedagogy, music physiology, prevention, and physiotherapy.
- Multimodal research data on movement, posture, and sound.
- Open and reproducible workflows for research and teaching.
Project Leadership and Scientific Coordination
- Ao. Univ.-Prof. Mag. Dr. Matthias Bertsch, Motion-Emotion Lab, Centre for Music Physiology, mdw – University of Music and Performing Arts Vienna – bertsch@mdw.ac.at
- Univ.-Prof. Dr. Christoph Reuter, SInES Lab, Department of Musicology, University of Vienna
Project Team
- Matthias Bertsch (1, 8) 🇦🇹 – research coordination, music physiology, acoustics, and data collection
- Christoph Reuter (2, 8) 🇦🇹 – musical acoustics, audio analysis, and SInES Tools
- Alexandra Türk-Espitalier (1, 8) 🇦🇹 – violin pedagogy and postural assessment
- Tina Margareta Nilssen (1) 🇦🇹/🇳🇴 – violin performance, music physiology, and expert assessments
- Tobias Großhauser (1) 🇦🇹 – sensor technology and interaction design
- Alessio Nacuzi (3, 4, 8) 🇮🇹 – violin pedagogy, biorobotics, motion capture, and data collection
- Bronwen Ackermann (6, 7) 🇦🇺 – performing arts health and physiotherapy
- Alberto Carratello (4, 8) 🇮🇹 – violin performance, pedagogy, and data collection
- Christoff Zalpour (6, 8) 🇩🇪 – physiotherapy, prevention, and movement analysis
- Elisabeth Grain (6, 8) 🇩🇪 – physiotherapy and movement analysis
- Giorgio Stefano Gnecco (5) 🇮🇹 – machine learning, data science, and research funding
- Alberto Bologni (4) 🇮🇹 – violin pedagogy and participant recruitment
- Veronica Santoro (3) 🇮🇹 – biorobotics and data analysis
- Calogero Maria Oddo (3) 🇮🇹 – biorobotics, neuroprosthetics, and sensor technology
- Nikolaus Ballenberger (6) 🇩🇪 – biomechanics and the analysis of posture and movement in violin and viola performance
- Dirk Möller (6, 8) 🇩🇪 – music physiology, physiotherapy, EMG, and movement analysis
Partner Institutions
- Motion-Emotion Lab, Centre for Music Physiology, mdw – University of Music and Performing Arts Vienna, Vienna 🇦🇹
- SInES – Space for Interdisciplinary Experiments on Sound, Department of Musicology, University of Vienna, Vienna 🇦🇹
- The BioRobotics Institute, Scuola Superiore Sant’Anna, Pisa 🇮🇹
- Conservatorio di Musica Luigi Boccherini, Lucca 🇮🇹
- Scuola IMT Alti Studi Lucca, Lucca 🇮🇹
- INAP/O – Institute of Applied Physiotherapy and Osteopathy, Osnabrück University of Applied Sciences, Osnabrück 🇩🇪
- University of Sydney, Sydney 🇦🇺
- Austrian Society for Music and Medicine (ÖGfMM), Vienna 🇦🇹
Publication
- Bertsch, M., Reuter, C., Türk-Espitalier, A., Nilssen, T. M., Großhauser, T., Nacuzi, A., Carratello, A., Grain, E., Brüggemann, J., & Leitz, T. (2026). Healthy Strings: AI-Powered Posture Tracking for Violinists: Markerless 3D Motion Analysis in Violin Performance – Validation, Multimodal Synchronisation, and Physiologically Informed Posture Assessment. Proceedings of the Symposium Beyond the Score – Health & Well-Being in Music and Performing Arts. https://doi.org/10.13140/RG.2.2.36013.04324
Presentation
- Bertsch, M., Reuter, C., Türk-Espitalier, A., Nilssen, T. M., Großhauser, T., Nacuzi, A., Carratello, A., Grain, E., Brüggemann, J., & Leitz, T. (10 April 2026). Healthy Strings: AI-Powered Posture Tracking for Violinists [Poster presentation]. Symposium Beyond the Score – Health & Well-Being in Music and Performing Arts, Salzburg.
Contact
Ao. Univ.-Prof. Mag. Dr. Matthias Bertsch
Motion-Emotion Lab, Centre for Music Physiology
mdw – University of Music and Performing Arts Vienna
bertsch@mdw.ac.at