How audio and video are merged?

What you call as the Audio-Video merge is generally termed as Audio-to-video synchronization, lip sync or audio-video sync. It refers to a relative timing of an audio or sound and its related video or image in parts during its creation, its post-production mix, during its transmission, its reception and while it is play-back processing. It is generally referred as a time when both the sound and the video are having a time-related cause and effect relationship.

The AV-sync can create an issue in broadcasting a program on television, while video conferencing, or even in a film.
The Digital or analog audio video streaming or those video files usually contain a type of explicit AV-sync timing, either in the shape of an interleaved video and audio data or it can be done as an explicit relative time-stamping of data. The processing of a particular data has to follow the relative data timing by either taking action like stretching between or creating an interpolation of the data it has received. If this processing of data does not follow the time, then the AV-sync error takes place. It will then increase in whenever data it loses because of the transmission errors or due to the missed timed processing.