Although physics-based sound synthesis is not widely used, we have more and more accurate knowledge of the behaviour of the instruments. Traditional techniques generate the resulting sound by the modification of prerecorded samples, therefore, they hide the sound production mechanism, the inside parameters have no meaning to a musician. The physics-based method focuses on the internal structure, this way, the parameters of the model will have more direct interpretation in the real world, for example the length of the string or the weight of the hammer.
The goal of my thesis is to analyze the possibilities of implementing a relatively new method, the resonator approach in an FPGA circuit. These systems are mostly realized in DSP (digital signal processor) architectures on the market, but the pipeline and parallel operation give reason for existence of FPGA’s in this area, newer devices already have DSP blocks.
In the first part of the thesis I present the structure of the piano and the possibilities of sound synthesis including the physics-based waveguide and resonator model. The second part is about the reference implementation of the chosen resonator model in MATLAB environment, and the designed hardware for the FPGA.
Finally, I delineate my achievements and the available further developments.