Abstract: The present invention relates to a time-separated speech coder, that codes the transitional signal of voiced/unvoiced sound through harmonic speech coding.

Abstract: The present invention relates to a harmonic-noise speech coder and coding algorithm of the mixed signal of voiced/unvoiced sound using harmonic model.

Abstract: An inventive method for manufacturing an array of M.times.N thin film actuatable mirrors, the method being capable of preserving the structural integrity thereof, comprises the steps of: preparing an active matrix; depositing a passivation layer and an etchant stopping layer, successively, on top of the active matrix; forming a thin film sacrificial layer; depositing a first mask layer made of a silicon oxide; forming a second mask layer; forming an array of M.times.N pair of cavities, one of the cavities in each pair exposing a portion of one of the connecting terminals; removing the first and the second mask layer; forming an array of M.times.N actuating structures; removing the thin film sacrificial layer to thereby form the array of M.times.N thin film actuatable mirrors. During the formation of the cavities, the cavities is formed into a trapezoidal shape as a result of the difference in the etching rate between the first mask layer and the thin film sacrificial layer.