Linear Prediction and Long Term Predictor Analysis and Synthesis

Main Article Content

Xiaohua Zhou Banu W. Yohanes


Spectral analysis may not provide an accurate description of speech articulation. This article presents an experimental setup of representing speech waveform directly in terms of timevarying parameters. It is related to the transfer function of the vocal tract. Linear Prediction, Long Term Predictor Analysis, and Synthesis filters are designed and implemented, as well as the theory behind introduced. The workflows of the filters are explained by detailed and codes of those filters. Original waveform files are framed with Hamming window and for each frames the filters are applied, and the reconstructed speeches are compared to original waveforms. The results come out that LP and LTP analysis can be used in DSPs due to its periodical characteristic, but some distortion might be coursed, which examined in the experiments.


Download data is not yet available.

Article Details



[1] L. Flanagan, Speech Analysis Synthesis and Perception, Academic, New York, p. 119, 1965.

[2] E. N. Pinson, Pitch-Synchronous Time-Domain Estimation of Formant Frequencies and Bandwidths, J. Acoust. Soc. Amer. 35, pp. 1264-1273, 1963.

[3] L.B. Rabiner and R.W. Schafer, Digital Processing of Speech Signals, Prentice Hall, New Jersey, 1978.

[4] B. S. Atal, Characterization of Speech Signals by Linear Prediction of the Speech Wave, Proc. IEEE Symp. on Feature Extraction and Selection in Pattern Recognition, Argonne, Ill., pp. 202-209, 1970.

[5] B. S. Atal and M. R. Schroeder, Adaptive predictive coding of speech signals, Bell. Sys. Tech. J., vol. 49, pp. 1973--1986, October 1970. doi:10.1109/ICASSP.1980.1170967

[6] C.H. Ritz, Decomposition and Interpolation Techniques for Very Low Bit Rate Wideband Speech Coding, PhD Thesis, 2003.

[7] G. H. Golub and C. F. Van Loan, Matrix computations, Johns Hopkins University Press, Baltimore, 1989