Parallel digital signal processing for efficient piano synthesis

While computational acoustics techniques for musical instruments emulation reached a remarkable maturity due to continuous development in the last three decades, implementation into embedded digital instruments lags behind, with only a few notable commercial products to solely employ physics-based algorithms for acoustic instruments tone synthesis. In this paper a parallel DSP architecture for the efficient implementation of the acoustic piano on embedded processors is reported. The resulting model is able to provide faithful reproduction of the acoustic piano physical behaviour and can also be used as an engine for novel instruments that need to provide advanced multimodal output (haptics, spatial audio) with a low-cost embedded platform.