A system, and an apparatus of speaker, and a method for forming a virtual audio is provided. An audio source is processed by a second audio source processing program according to parameters which are entered by the user, and filtered to be a high frequency audio signal. The audio source with orientation is outputted by the orientation speaker.