| 摘要 |
PURPOSE:To attain the conversion of a voice and a sound signal meaningful to sound quality by revising difference from attributes between nearby frequency components is time signal information. CONSTITUTION:Band split filters 2-4 and MDCT circuits 5a-5d have a function of converting sound time signal information into plural frequency components. The difference of the attribute of at least one local peak of the plural frequency components from that of other frequency components almost within a critical band based on the audible characteristic. As means revising the attribute, a frequency shift peak detection circuit 12, a dissonance frequency detection circuit 11, a masking curve detection circuit 16, a mask circuit 10, a frequency shift peak detection circuit 12, a minimum audible curve generating circuit 17 are employed. Then the harmonic relation between frequency components in the critical band backed up audible is controlled to adjust the sound quality of a voice and a sound signal to be useful for persons. |
