| 摘要 |
PURPOSE:To always perform an appropriate control even in case of changing a scale factor, and to reduce quantize distortion, by calculating a correction function which prevents the quantize distortion generated by the introduction of a leak constant based on a quantization signal, and impressing it on an adaptive control part. CONSTITUTION:An adaptive correction part 8 calculates a correction function G expressed in the next expression by the value Y of a scale factor 6. G=a.Y<(1>-<beta)>(0<<a<1). The correction function G is a factor to correct an error between Y and Y corresponding to the value Y of the scale factor 6. After calculating the correction function G, the adaptive correction part 6 impresses the correction function G on an adaptive control part 7. The adaptive control part 7 calculates a function YK+1=GKM((IK){YK}<beta> multiplied by the correction function G, and outputs the function Y to a normalization part 2 as the value of the scale factor 6. Therefore, it is possible to reduce the quantize distortion, such as overload distortion generated in an adaptive quantizer at the time of increasing the Y, or adversely, a particleshaped noise at the time of decreasing the Y, etc.
|
