The research focuses on:
Expressive intentions of music performance, from acoustic and perceptual analysis to the development of techniques for rendering expressiveness in a synthetic performance. So far the research has led to the development of an original modeling approach that allows to morph among performances with different expressive contents, both at the symbolic and signal levels. Sound rendering is obtained by interfacing the expressiveness model with a dedicated postprocessing environment, which allows for the transformation of the event cues. The processing is based on the organized control of basic audio effects. Recent research is focusing on extending these results towards the development of Multisensory Integrated Expressive Environments (mixed reality applications in the performing arts such as interactive dance, music, or video installations): algorithms for the recognition and processing of the expressive content of the user gestures (expressive feature extraction from signal computed using hw/sw motion capture systems).
Example: tracking and control of expressive gesture (video, 9MB).
Digital music libraries. This field places particular emphasis on: a) audio restoration (algorithms for noise reduction), b) archiving/cataloguing (multimedia database), and c) the integration of musical audio content and meta-data descriptions, with a focus on flexible access. Applications range from large distribute libraries to mobile access platforms. In the audio restoration field, different approaches are investigated to address the problems of a) intelligibility recovery, b) vocal pattern recognition, c) adapting musical audio documents to the current aesthetic taste. Sergio Canazza has developed specific techniques based on: 1) frequency-domain methods, such as various forms of noncasual Wiener filtering or spectral subtraction schemes, incorporating knowledge of the human auditory system; 2) time-domain restoration by signal models such as Extended Kalman filtering (EKF); 3) restoration by source models. Recently Sergio Canazza has developed a software (Photos of GHOSTS: Photos of Grooves and HOles, Supporting Tracks Separation) ables to synthesize audio recorded in shellac discs from the still image of the carrier surface.
Rosina Gioiosa Trubia – Sta terra nun fa pi mia (This land is not for me)
Shellac 78 rpm 10” Brunswick 58073B (E 26621/2)
Recorded in New York 23rd, February, 1928
Original (played by turntable)
Original (played by grammophon)
Restored by EKF