发明名称 DECODER FOR ADAPTIVE PREDICTING CODE SIGNAL
摘要 PURPOSE:To obtain a device without distortion or offset by expanding a quantization residual signal in response to a maximum value of a reproducing signal in one frame so as to apply arithmetic operation even when a processor of a fixed decimal point operation is used and processed. CONSTITUTION:A residual decoder 2 of a decoder section 4 of the prestage decodes a residual code 102 and outputs a quantization residual signal 105 via a demultiplexer 1. A reproducing signal 106 is calculated by adding a predicting signal 107 and the signal 105. The reproducing signal 106 by one frame outputted from the decoder section 4 is stored tentatively in a buffer section 5 and its maximum value is detected by a maximum value detection section 6. The detection section 6 sends a scaling quantity 109 to an expander 7 or a reducing device 8 of the post-stage decoder 9 and the arithmetic operation of the quantization residual signal 110, the reproduced signal 111 and the prediction signal 112 expanded respectively is applied by the expanded scale. Thus, the arithmetic operation accuracy is improved and distortion or offset is not caused.
申请公布号 JPS63261925(A) 申请公布日期 1988.10.28
申请号 JP19870096431 申请日期 1987.04.20
申请人 MITSUBISHI ELECTRIC CORP 发明人 MURATA KENJI
分类号 H03M7/38 主分类号 H03M7/38
代理机构 代理人
主权项
地址