Dolby AC-4 Audio

Dolby AC-4 is a lossy audio system supporting channel-based, immersive, and personalized presentations.

One AC-4 programme can offer several listening choices without storing a separate complete soundtrack for every choice. The decoder selects a presentation: the relevant audio components and the information needed to combine and render them.

This is a central reason AC-4 is more than a newly named AC-3 bitstream.

Assemble a hypothetical programme

Suppose a programme supplies music/effects, English dialogue, French dialogue, and associated description narration. It can define presentations using different combinations:

Presentation Illustrative selected components
English main Music/effects + English dialogue
French main Music/effects + French dialogue
English described Music/effects + English dialogue + description

The table illustrates a permitted component-oriented approach, not a rule that all AC-4 programmes use this exact structure. A programme can also use a complete main mix or other defined component types.

Sharing music/effects avoids duplicating that content in every language combination. The presentation metadata tells the decoder which components belong together and how the available controls should affect them.

Selecting and mixing the required substreams

  1. Read the available presentation information.
  2. Select an appropriate language/accessibility presentation.
  3. Find and decode its required substreams.
  4. Apply the supplied mixing, gain, loudness, and rendering information.
  5. Produce the supported channel or spatial output.

If description narration is an associated component, the decoder may also need instructions for balancing it with the main material. A standalone narration component is not necessarily the whole film soundtrack.

This also explains why one container audio-stream entry can expose several meaningful choices. A conventional track list and an AC-4 presentation list describe different layers.

Channel and object components support different render jobs

Channel-based material has defined output roles. Object-based material supplies sound with spatial information for rendering. AC-4's standards cover both traditional channel coding and immersive/personalized uses.

Dialogue enhancement depends on what the delivery provides. Separate dialogue or suitable enhancement information can enable a controlled change; a decoder cannot assume an arbitrary mixed recording contains an independently adjustable voice merely because the codec is AC-4.

Preserve a selected result or the selectable structure

AC-4 is specified in the ATSC 3.0 AC-4 audio system and other delivery contexts. An Emby recording can therefore contain picture that is readily decoded while the soundtrack needs AC-4-specific presentation handling.

Converting a selected presentation to ordinary stereo or surround creates a useful listening result. It generally flattens the selected components into that mix, so the output no longer contains all the original language/accessibility choices as separately selectable AC-4 presentations.

Preserving the encoded structure and creating a compatible listening copy are different goals. The distinction follows directly from which components and presentation instructions survived the operation.

See Also

References