| 摘要 |
PURPOSE:To improve the sound quality of a coded voice signal by reading voice signal coding data from a storage section, detecting the max. differential value, compressing the differential value to a prescribed number of bits, and adding and subtracting the prescribed value. CONSTITUTION:The max. differential value in the memory section 1 is read into a computing circuit 2 and the first data of the voice signal included in the read block is read into the circuit 2. The difference between the previous decoded data and the first data is calculated by the circuit 2 and the result of the calculation is inputted to a comparator 3. The differential value of the result of the calculation and the max. differential value supplied from the memory section 1 are compared and when the differential value of the result of the calculation is larger than the max. differential value, this value is sent to the memory section 1 and is substd. with the max. differential value thereof and is delivered to a shift quantity calculating circuit 4. The circuit 4 calculates a scale value, transmits the value to the memory section 1 and writes the value into a table. The circuit section 4 reads the data from the memory section 1, calculates the difference from the previous decoded value, inputs the value to a shifter 5, shifts the same to a right side and compresses the differential value. The 1 of the bit discarded at the time of shifting is thus added and subtracted. |
