On the use of pitch predictors with high temporal resolution

The use of pitch predictors in linear predictive coding systems is an efficient way to represent periodicity in the speech signal. Typically, the predictor is described by one parameter representing the delay in samples and 1 to 3 predictor coefficients can provide interpolation for periodicities that are not a multiple of the sampling interval and allow for a frequency-dependent gain. The authors describe a first-order pitch predictor whose delay is specified as an integer number of samples plus a fraction of a sample at the current sampling frequency. This realization has a better performance than conventional multiple coefficient predictors and leads to more efficient coding of the predictor parameters