Audio processing is an interesting and challenging area due to the strict requirements of highlevel
human ear perception. Although audio processing is a part of digital signal processing, there has not yet been any realtime
parallel implementation of the essential signal processing tool. We have developed, implemented and analysed a Sliding Discrete Fourier Transform algorithm using a particular vector parallel processing. This investigation is
focused on speeding up realtime
parallel implementation of the SDFT algorithm. It has developed and implemented some realtime
parallel algorithms with different methods of data copy in order to find the fastest algorithm. In order to achieve a whole signal processing
tool for professional use in the audio processing field a realtime
parallel algorithm of Inverse Discrete Fourier Transform has been developed. In order to speed up this algorithm
some versions were developed with different methods of data copy. All algorithms were tested and analysed. Realtime
parallel algorithms were achieved with promising results.
The signal processing tool of the SDFT and the IDFT algorithms was attained.
|Date of Award||1 Jan 2011|
|Supervisor||John Fitch (Supervisor) & Russell Bradford (Supervisor)|
- sliding discrete Fourier transform parallel algorithm
- clearspeed accelerator