One embodiment of the present invention sets forth a technique for starting playback of an audio portion of digital media content before a video portion of the digital media content is received. An advantage of the disclosed technique is that the perceived delay from the time when a user selects the digital media content to when playback of the digital media content begins is reduced. The audio may be faded-in and one or more still frames of video data may be displayed while the video stream is retrieved. Additionally, a license for DRM (Digital Rights Management) encryption of the title content may be retrieved before starting playback of the video stream, so that the digital media content that is retrieved is also protected.