Fractal Wavelet Compression of Audio Signals
Robert A. Wannamaker, Edward R. Vrscay · 1997
A novel approach to audio signal compression is described which marries wavelet transform coding with fractal-based compression techniques. The signal is divided into frames which are subjected to a Fast Wavelet Transform, with subsequent fractal coding of high frequency wavelet coefficients using a collage approach. Results from initial implementations are examined. Compression ratios above 6:1 should ultimately be attainable with good fidelity signal reconstruction. 1 Introduction Over the past decade, audio compression has become almost ubiquitous due to the enormous appetite of multimedia systems for data. Current state-of-the-art audio compression schemes transform the incoming signal (using a discrete cosine transform or a set of subband filter banks) and quantize each transformed datum according to its psychoacoustic salience as determined by some psychoacoustic model [1]. The signal reconstructed from the coded data will not be identical to the original signal, but should (ide...