| 摘要 |
PROBLEM TO BE SOLVED: To provide a sound signal generation device capable of performing efficient musical sound waveform generation while reducing useless time slots within each sampling cycle as further as possible when performing musical sound waveform generation of a plurality of sounds in parallel, and also capable of suppressing increase of a case where the musical sound waveform generation is ruptured.SOLUTION: In the sound signal generation device configured to generate musical sounds of a plurality of channels in parallel, a cache memory 401 is provided for each channel and waveform data are read ahead in the cache memory 401. On the basis of the amount of waveform data read ahead in the cache and/or on the basis of a speed of progress of address information, a predetermined number of channels to preferentially read ahead waveform data into the cache memory 401 are determined from among the plurality of channels. Regarding the channel determined to preferentially perform read-ahead, waveform data are read ahead from a waveform memory and the cache memory 401 is filled with the waveform data. |
