| 摘要 |
PURPOSE:To prevent the increase in a table amount by multiplying a compression/expension coefficient by a product of a constant and a value being the result of subtraction of a product of a constant and a present quantized step size from 1. CONSTITUTION:In quantizing an input signal X(n), a coefficient in response to an output In-1 delayed from a delay circuit Z<-1> by 1 sample's share is multiplied and the result is inputted to an (MIn-1) multiplier MUP1. On the other hand, after a step size DELTA(n-1) is delayed by one sample at the delay circuit Z<-1>, the result is inputted to multipliers MUP2, 3. The multiplier MUP3 obtains [alpha. DELTA(n-1)] by multiplying the DELTA(n-1) by a coefficient alpha, the output is subtracted from 1 by an adder AD, to [1-alpha.DELTA(n+1)] the multiplier MUP2 multiplies the result by the DELTA(n-1), this output is inputted to the multiplier MUP1, the multiplication of [MI(n-1).DELTAn-1.(1-alphaDELTAn-1)] is performed and its output is inputted to a quantizer Q as the step size DELTAn. |
