| 摘要 |
PURPOSE: An audio data compression device is provided to search for masking level necessary to audio data compression by recording weighted value of masking level of each unit in memory and operating logically scale factor of each unit in the recorded weighted value and data transformed on frequency axis. CONSTITUTION: A sub-band filter unit(100) filters audio data input into prescribed band. A frequency axial transform unit(101) transforms band output in the sub-band filter unit(100) into data on frequency axis. A unit dividing unit(102) divides data output in the frequency axial transform unit(101) into prescribed unit. A scale factor detection unit(103) detects maximum scale factor in unit determined in the unit dividing unit(102). A scale factor quantization unit(104) makes scale factor of the unit out of the maximum scale factor detected in the scale factor detection unit(103). A memory unit(105) records value masked by other unit of each unit determined in the unit dividing unit(102). A logic operation unit(106) operates logically scale factor of each unit made in the scale factor quantization unit(104) and weighted value recorded in the memory unit(105), and searches for masking level of each unit. A normalization unit(107) divides data determined in the unit dividing unit(102) by scale factor made in the scale factor quantization unit(104). A comparison unit(108) compares masking level operated in the logic operation unit(106) with scale factor of unit quantized in the scale factor quantization unit(104) and outputs the result value. A bit allotment unit(109) allots bit to data normalized in the normalization unit(107) according to quantization noise permissible level made in the comparison unit(108) and outputs compressed audio data.
|
