A system and method are provided for enhancing an audio recording which comprises a recording of a sound signal obtained from the play-out of an audio signal via a speaker. The audio signal, and thereby the sound signal, may represent certain audio content, e.g., a radio station or TV audio. To perform the enhancing, the recording of the sound signal is suppressed using the audio signal, thereby obtaining an intermediate audio recording. An original version of the audio content is then added to the intermediate audio recording to obtain an enhanced audio recording. This original version is generally of higher quality as it generally does not represent a background audio component but rather was purposefully recorded or generated.