Method of coding an audio signal by using an orthogonal transformation

摘要

A method of coding an audio signal by using an adaptive orthgonal transformation comprises a step (ST1, ST2) of calculating the sum total of a power of an input audio signal in a time duration of an orthogonal transform length or over the length; a step (ST3) of adaptively controlling the minimum audible level as a threshold level corresponding to the power level on a frequency axis after the orthogonal transformation or on a time axis before them orthogonal transformation; a step (ST4) of excepting samples having a power level under the threshold level from orthogonally transformed samples; and a step (ST5) of quantizing remaining samples. By the method, the threshold level of the minimum audible level is adaptively controlled corresponding to an observed level of an input audio signal. The method has an effect that even if the input audio signal has various power levels, it is possible to usually set the adaptively audible value (the threshold level), thereby effectively reducing a coding amount.