MPEG-4, The Coding Of Audio Visual Objects, Part II

The following is a precise definition of the many objects within the MPEG-4 codec. A movie box contains all the metadata necessary for a video presentation and is kept at the very beginning of a video or at the very end. A media box is 32-bit in size and it holds all the information regarding the specific media used for the video. The movie header box contains specific information dealing with the presentation and includes the creation time, modification time, timescale and duration. The track box contains a single track of presentation.



This presentation can contain one or more tracks. Each one is an individual having its own temporal and spatial information. The tracker header box describes exactly the characteristics of one track and is sited exactly at the beginning of each track. It like the movie header box provides the following information: creation time, modification time, duration, layer, alternate group, volume, matrix, and width and height of the track’s presentation size.



A track reference box provides information to another track, which is referenced in the main track box. Also contained in the track box, is the media header box, which contains all the information about the media contained in the track. This information could be creation time, modification time, timescale, duration and language. The media header box or Meta box contains the handler reference box. This describes the structure or format of the ‘meta’ box contents. It furthermore declares the process by which the media data is shown. It shows whether the Meta data is pre-defined, handler type and whether it is reserved or not.



The Media Information Box contains specific header boxes, which describe the different types of media by track. It does this by assigning to them a specific header box. The Video Media Header Box has the information of graphicsmode and opcolor. Graphicsmode specifies the specific composition mode for a particular track. The opcolor is a set of three colors available to that track.



These represent the RGB divisions of the available colors. The Sound Media Header Box can also be present in a Media Information Box giving the balance of the audio if perchance it is stereo. Monaural is 0.0. The Hint Media Header Box gives general information without protocol for hint tracks. It presents information, which gives MaxPDUsize (maximum size of the PDU in a track), avgPDUsize, maxbitrate, avgbitrate, and reserved status.