No abstract

No abstract

No abstract

As Digital Signal Processors and digital signal processing technology are in increasing process, optimization of previously developed code for Digital Signal Processors has become more and more important. This paper presents an approach for porting Dolby Pro Logic II audio decoder, from the reference platform to the targeted DSP architecture. The overall system is firstly analyzed and then ported to the DSP platform. The system features are validated and verified using software simulations and audio system measurements in real-time conditions. The application of targeted DSP system is in audio processing for digital TV receivers, mobile phones, headsets, etc. The results have shown that execution time is decreased in the final system, and that higher sampling rates are enabled for better performances.

An increased complexity of modern computer based systems is accompanied by advanced software architectures embedded into a system. Common approach for designing consumer electronic products includes usage of legacy code with some modifications in order to run it on the targeted platform. Shorter development time and budget constraints are also influencing the process of system design. Thus, reuse of existing legacy source codes as well as inclusion of available open source contributions is widely used in industrial practice. Since each embedded system is having its specific limitations, reuse of embedded software without adaptation to the new platform is not always possible. This paper describes a case study of software reengineering of Dolby Virtual Speaker from the reference platform to the required DSP based system on a chip, including system verification and system validation in real-time conditions. The critical reengineering decisions are elaborated with given results and lessons learned during project implementation.