Real-Time Audio Signal Processing on RISC Workstations via Short-Time Fourier Transforms

Peevers A. Real-Time Audio Signal Processing on RISC Workstations via Short-Time Fourier Transforms. In: AES Convention 97. Audio Engineering Society; 1994. Paper 3927. Available from: https://aes.org/publications/elibrary-page/?id=6305

@inproceedings{Peevers1994_6305,
  author = {Peevers, Alan},
  title = {{Real-Time Audio Signal Processing on RISC Workstations via Short-Time Fourier Transforms}},
  booktitle = {AES Convention 97},
  note = {Paper 3927},
  year = {1994},
  month = nov,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=6305}
}

TY  - CPAPER
TI  - Real-Time Audio Signal Processing on RISC Workstations via Short-Time Fourier Transforms
AU  - Peevers, Alan
T2  - AES Convention 97
M1  - Paper 3927
PY  - 1994
DA  - 1994/11/06
UR  - https://aes.org/publications/elibrary-page/?id=6305
PB  - Audio Engineering Society
LA  - en
AB  - The most recent generation of RISC-based workstations has made it possible to perform powerful signal-processing functions in real time. These functions can be implemented in a high-level language, such as C, while retaining sufficient performance for real time. A useful example that exploits these powerful capabilities has been written for the Silicon Graphics Indigo family of workstations. The specific application implements a real-time signal analysis/synthesis tool based on the overlap add short-time Fourier transform. It generates high-resolution 3-D images of the power spectrum of the audio input in real time. This application utilizes the 100-MHz RISC to compute forward and inverse fast Fourier transforms (FFTs), along with the 3-D surface polygons, while keeping up with real time. There is an intuitive user interface which allows users to examine various spectral features in great detail. In addition, it is possible to draw trajectories for filter center frequencies on top of the spectrogram and listen immediately to the resulting time-varying filters. A demonstration of this application will be shown.
ER  -