| 摘要 |
PURPOSE: To simplify the calculation of intensity estimating value and to reduce required operation time by forming the absolute value of an output signal from an inverse quantizer, attenuating higher frequency signal components and accu mulating and limiting the deviation of the absolute value. CONSTITUTION: A sound signal is digitized and supplied to an encoding device 1, a predicted input signal is subtracted from the digitized sound signal, its remaining residual signal is divided by the estimated intensity to normalize the residual signal, and the normalized signal thus produced is supplied to a quantizer/encoder 6. An output signal from the quantizer/encoder 6 is transmitted to a decoding device 2 through a channel 7, supplied to a reverse quantizer/ decoder 5, a signal obtained by multiplying an output signal from the reverse quantizer/decoder 5 by the estimated intensity is added to a predicted signal to form a control signal for a predictor 3. The absolute value to the output signal from the reverse quantizer/decoder 5 is formed and a higher frequency signal component is attenuated to accumulate and limit the deviation of the absolute value of the output signal from the encoding device 1. |
