It is intended to remove a risk of illegal high quality restoration while enabling trial viewing of contents such as music, to make it possible to reproduce contents with high quality by acquiring a relatively small amount of data, and to make it possible to implement hardware at a lower cost. To generate trial listening data containing dummy data, a pseudo random number R is generated first. A remainder r of R divided by 128 is calculated. The remainder r is coded and incorporated into high sound quality restoration data. A bandwidth limitation region of trial listening data is replaced by dummy data starting from its rth bit. As a result, the head position of the dummy data in the trial listening data varies frame by frame, whereby the safety of the trial listening data can be increased.