The present invention suppresses audio skipping. A frame number acquisition unit 102 acquires a current frame number S11. An audio data acquisition unit 104 acquires an audio data S12 of a current frame. A signal processing unit 110 holds the audio data S12 of the current frame in the sub-buffer 114 when the current frame number S11 is discontinuous with a last normal frame number. The signal processing unit 110 stores the audio data stored in the sub-buffer 114 into the main buffer 112 and sets it as a reproduction object, if the current frame number S11 matches an expected value S14 before audio data of N frames is stored in the sub-buffer 114, where N is a predetermined value.