Method and circuit for carrying out forward and inverse quantization by varying a reference step size

DETAILED DESCRIPTION What is claimed is: 1.
A method of carrying out a forward quantizing operation by processing a succession of input digital signals into a succession of digital quantized codes, respectively, said digital quantized codes being disposed at successive instants spaced by a predetermined interval of time, respectively, in relation to step sizes which are decided at the respective instants with a current one of step sizes decided for disposition of a current one of said quantized codes at a current one of said instants with reference to a previous step size, a previous quantized code, and a preselected number smaller than unity, said previous step size being one of said step sizes that is decided at a previous instant, said previous instant being one of said instants that is said predetermined interval prior to said current instant, said previous quantized code being one of said quantized codes that is disposed at said previous instant, said forward quantizing operation comprising the steps of: monitoring said quantized code succession to provide a level signal representative of, at said current instant, a level dependent on particular ones of said quantized codes that precede said current quantized code and include said previous quantized code; setting a reference size variable at a predetermined period longer than said predetermined interval and adjusted in accordance with said level; deciding the step size at a subsequent one of said instants that is said predetermined interval after said current instant, with reference to said current quantized code and said current step size and additionally to said reference size; dividing said input signals by the step size to produce a sequence of quotient digital signals representative of quotients which appear as results of the division, respectively; and quantizing said quotient signals into said digital quantized codes, respectively.
2.
A method as claimed in claim 1, wherein said setting step comprises the steps of: preparing a plurality of size levels each of which is different from the others and which are determined in consideration of said reference size; and selecting one of said size reference levels as said reference size in response to said level signal