Funding period: 01.08. 2004 - 31.07. 2007

The recent rapid expansion of digital communications and emerging technologies in wired data networks, especially in the Internet, as well as in wireless communications (e.g. Wireless LAN and multimedia messaging) require novel functionalities in digital audio signal processing. These functionalities include new approaches to coding and synthesis of audio signals. Sound source models (SSMs) have a potential to achieve a parametric representation of audio and music. They simulate the physical behavior of various sound production mechanisms. Conceptually, SSMs provide means to generalize the technology presently used in the GSM standard for coding and transmission of speech to a broader range of sound sources.

This task requires an investigation of the acoutic properties of sound sources, development of new sound source models, and tools for analyzing the audio signal, selecting the appropriate SSM, and extracting the model parameters. These are the main objectives of the MAPS project in the framework of modeling of sound sources.

Human perception nowadays sets increasingly important criteria for the quality of synthetic audio. The main objectives in the MAPS project in the framework of perception of sound sources are to find perceptually justified parameter spaces and quantizations for sound source models, and to evaluate the perceptual quality of synthetic sounds based on models of auditory perception.

TKK Laboratory of Acoustics and Audio Signal Processing

Prof. Matti Karjalainen
Prof. Vesa Välimäki

Dr. Cumhur Erkut
Dr. Hanna Järveläinen

• Järveläinen, H., Karjalainen, M., Importance of inharmonicity in the acoustic guitar, submitted to Computer Music Conference, ICMC05, 2005.
• Erkut, C., Modular interactions and hybrid models: A conceptual map for model-based sound synthesis, submitted to EUSIPCO, Antalya, Turkey, 2005.
• Välimäki, V., Pakarinen, J., Erkut, C., and Karjalainen, M., ''Discrete-time Modelling of Musical Instruments,'' Reports on Progress in Physics, vol. 69, no. 1, pp. 1-78, January 2006. Invited paper.