Track roles

A track’s role is its author-declared purpose. It applies to audio and text tracks only, is never probed from the media, and is stored in the descriptor’s role field. This page is the exact reference for the role vocabulary and how each role is rendered into DASH and HLS.

To set a role, see Label tracks with roles. Roles are validated at index time: a role on a video track, a role that doesn’t match the track’s media type, or an unknown value aborts the run. A role never affects a track’s representation id.

Audio roles

Value	Meaning
`main`	The primary audio.
`alternate`	An alternate version of the main audio.
`commentary`	Commentary (e.g. director’s commentary).
`dub`	A dubbed rendition in another language.
`description`	Audio description for viewers who are blind or have low vision.
`enhanced-audio-intelligibility`	Dialogue enhanced for intelligibility.

Text roles

Value	Meaning
`subtitle`	Translation subtitles (dialogue only). The default when no role is set.
`caption`	SDH / closed captions (dialogue plus non-dialogue sound description).
`forced-subtitle`	Forced narrative subtitles (foreign dialogue or on-screen text), shown even when subtitles are otherwise off.

DASH mapping

Roles are emitted as descriptors on the track’s AdaptationSet; tracks are grouped into adaptation sets by (sample entry, timescale, language, role).

`Role`

Scheme urn:mpeg:dash:role:2011; value is the role string verbatim.

Track	`Role` emitted
Audio with a role	`Role@value` = the role.
Audio with no role	none.
Text	`Role@value` = the role, defaulting to `subtitle` when unset.

`Accessibility`

Scheme urn:tva:metadata:cs:AudioPurposeCS:2007, emitted only for two audio roles:

Audio role	`Accessibility@value`
`description`	`1`
`enhanced-audio-intelligibility`	`8`

No other role emits an Accessibility descriptor.

HLS mapping

Roles set the selection attributes on the EXT-X-MEDIA rendition entries in the multivariant playlist.

Audio renditions

Audio tracks are grouped by sample-entry code. Within a group:

Attribute	Rule
`DEFAULT`	`YES` on the first `main`-role rendition; if no member is `main`, on the first rendition. `NO` otherwise.
`AUTOSELECT`	`NO` for the opt-in roles (`commentary`, `description`, `enhanced-audio-intelligibility`) unless the rendition is the group default; `YES` otherwise.
`CHARACTERISTICS`	`public.accessibility.describes-video` for `description`; `public.accessibility.enhances-speech-intelligibility` for `enhanced-audio-intelligibility`; absent otherwise.
`NAME`	The rendition’s language, qualified by its role when one is set — `nld`, `eng (commentary)`. A counter disambiguates collisions (`eng (2)`).

DEFAULT=YES implies AUTOSELECT=YES, so the group default is always auto-selected even when its role is opt-in.

Subtitle renditions

HLS subtitle renditions are not emitted yet — text tracks currently appear in DASH manifests only (CMAF wvtt sources). The mapping below is the design that on-the-fly text serving is being built toward.

All text tracks share one TYPE=SUBTITLES group.

Attribute	Rule
`DEFAULT`	Always `NO` — the viewer enables subtitles explicitly.
`AUTOSELECT`	`YES` only for `forced-subtitle`; `NO` for `subtitle` and `caption`.
`FORCED`	`YES` only for `forced-subtitle`.
`CHARACTERISTICS`	`public.accessibility.transcribes-spoken-dialog,public.accessibility.describes-music-and-sound` for `caption`; absent otherwise.

dyndo