1.6 kbit/s LP vocoder using time envelope

LPC interpolation by approximation of the sample autocorrelation function

“…Energy-weighting gives a better model for the high energy part of transitions. It has been reported in [12] that energy-weighted interpolation improves the performance of a coder at rapid onsets. The larger number of outliers in low-energy parts does, however, influence the objective results in terms of SD.…”

“…One way to incorporate the energies of analysis frames is to multiply the coefficients of an LPC representation with the energy of the corresponding analysis frame and to interpolate this energy weighted representation. For example, a vocoding algorithm is proposed in [12] that puts great effort into accurately coding the time envelope of the speech signal, because the energy contains important perceptual information. In that algorithm the frame energy is used in the LPC interpolation, in connection with LSP's.…”

“…Traditional speech enhancement methods are based on optimizing mathematical criteria, which in general are not always well correlated with speech perception. Several recent methods have also considered auditory processing information [8], [9], and constrained iterative methods using various levels of speech class knowledge [10]- [12].…”