A Fractal Self-Similarity Model for the Spectral Representation of Audio Signals

Ferreira AJS, Sen D, Sinha D. A Fractal Self-Similarity Model for the Spectral Representation of Audio Signals. In: AES Convention 118. Audio Engineering Society; 2005. Paper 6467. Available from: https://aes.org/publications/elibrary-page/?id=13183

@inproceedings{Ferreira2005_13183,
  author = {Ferreira, Anibal J. S. and Sen, Deep and Sinha, Deepen},
  title = {{A Fractal Self-Similarity Model for the Spectral Representation of Audio Signals}},
  booktitle = {AES Convention 118},
  note = {Paper 6467},
  year = {2005},
  month = may,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=13183}
}

TY  - CPAPER
TI  - A Fractal Self-Similarity Model for the Spectral Representation of Audio Signals
AU  - Ferreira, Anibal J. S.
AU  - Sen, Deep
AU  - Sinha, Deepen
T2  - AES Convention 118
M1  - Paper 6467
PY  - 2005
DA  - 2005/05/06
UR  - https://aes.org/publications/elibrary-page/?id=13183
PB  - Audio Engineering Society
LA  - en
AB  - In the application of conventional audio compression algorithms to low bit rate audio coding one is faced with the unsatisfactory tradeoff between coarser quantization and audio bandwidth reduction. BandwidthExtension has therefore emerged as an important tool for the satisfactory performance of low bit rate audio codecs. In this paper we describe one of a newer class of Frequency Extension techniques which are applied directly to the high frequency resolution representation of the signal (e.g., MDCT). This particular technique is based on a Fractal Self-Similarity Model (FSSM) for the short-term frequency representation of the signal and takes advantage of the high frequency resolution of the MDCT, namely in terms of parameter estimation.. The FSSM model, which may include multiple dilation and translation terms, has been found to be effective for a wide variety of speech and music signals and provides a compact description for long term correlation that may exist in frequency domain.. The Structure of the FSSM model is presented, issues related to parameter estimation, and its application to audio coding for bit rates of 8-48 kbps are discussed.
ER  -