Aaltodoc - homepage
Communities & Collections
Browse Aaltodoc publication archive
EN | FI |
Log In
  1. Home
  2. Browse by Author

Browsing by Author "Tolonen, Tero"

Filter results by typing the first few letters
Now showing 1 - 4 of 4
  • Results Per Page
  • Sort Options
  • No Thumbnail Available
    Development of a mobile interactive musical service
    (2002) Airas, Matti
    Helsinki University of Technology | Master's thesis
  • No Thumbnail Available
    Model-Based Analysis and Resynthesis of Acoustic Guitar Tones
    (1998) Tolonen, Tero
    Helsinki University of Technology | Master's thesis
  • No Thumbnail Available
    Nonlinear Modeling Techniques for Plucked String Instruments
    (1999) Tolonen, Tero
    Helsinki University of Technology | Licentiate thesis
  • No Thumbnail Available
    Object-based sound source modeling
    (2000-10-13) Tolonen, Tero
    Doctoral dissertation (article-based)
    This thesis proposes techniques for object-based audio and music. The work can be divided into two parts corresponding to model-based synthesis of musical instruments and computational analysis of audio and music. The contributions of this work are in signal and auditory analysis, sound object modeling, and sound signal synthesis. The first half of the thesis considers linear and particularly nonlinear discrete-time modeling of plucked string instruments. The nonlinear tension modulation phenomenon that is present in every vibrating string has been integrated into the model. Its simulation has been developed into a computationally feasible structure and it has been demonstrated that the models produce the desired effects on the synthetic tones in the desired manner. The proposed models can reproduce accurately such phenomena as variation of the fundamental frequency and coupling of the tension modulation via the bridge to the body and to the vibrating strings. The second half of this thesis mainly concentrates on computational audio and music analysis. Methods are presented for separation of harmonic sound sources using sinusoidal modeling. A computationally efficient and auditorily motivated multi-pitch analysis model has been developed and integrated with the sinusoidal separation techniques. Nonlinear least squares (NLS) methods are proposed for signal analysis. The first NLS technique performs high-resolution sinusoidal analysis which uses a priori information of the signal components obtained from the multi-pitch analysis. Another NLS-based method is used for refining a fundamental frequency estimate. It is also applied to detection of vibrato of prominent tones. The third NLS technique is used for detection of inharmonicity of a piano tone in a three-voice chord. The analysis methods are applied to obtaining a model-based representation of two-voice guitar music. Listening experiments were conducted measuring the perception of the variation of the decay pattern in model-based plucked string synthesis. The experiment results provide perceptual thresholds for changes in overall decay and frequency-dependent decay. The results are applied to parameterization of the synthesis model as well as designing a perceptually motivated parameter estimation method.
Help | Open Access publishing | Instructions to convert a file to PDF/A | Errata instructions | Send Feedback
Aalto UniversityPrivacy notice | Cookie settings | Accessibility Statement | Aalto University Learning Centre