Embedding Media Subtitles are a transcription or translation of dialogue, for cases when the sound is available but
may not be understood, especially if it’s in another language.
Captions are a transcription of dialogue but also describe any other sounds, music cues, or other
essential audio information. Unlike subtitles, captions are intended as a replacement for the audio
if the audio isn’t available, especially valuable to deaf viewers.
Descriptions provide text descriptions of a video if the video isn’t available, or for blind users, in
which case the text descriptions might be synthesized as audible speech or printed to a Braille
device.
Chapters are titles meant to act as navigation points, to jump to a particular segment of the media
source, like the scene selection on a DVD.
Metadata is information about the media, only intended for use by scripts or user-agents, not for
display.
The track element has an optional srclang attribute to declare the language of the text track. If the
track’s kind is set to subtitles, a srclang attribute must be present, though it may have an empty value
(srclang="") meaning the track has no language. An optional label attribute offers a title for each text
track, which could appear in a menu so your users can select one track from several options.
This is all brand new in HTML5, and is still very much in a state of flux as the specification marches toward
completion. As of this writing, only a few of the very latest browsers have implemented support for the
track element, but we can hope others add it soon. We’ve tried to boil it down to a brief introduction here,
and although the markup is simple, actually implementing text tracks for rich media is a much more
complex matter. One major tangle is defining and standardizing the format of the timed text tracks
themselves, and that part is still ongoing.Required Attributes
src: The URL of the text track data.Optional Attributes
kind: The kind of text track, specified by one of the keywords: subtitles, captions,
descriptions, chapters, or metadata. The track’s kind defaults to subtitles if the attribute
is missing.
srclang: The language of the text track, specified by a valid language tag (see langtag.net for
more on language tags). This attribute must be present if the track element’s kind is
subtitles, though srclang may have an empty value (in which case the track has no specified
language).
label: A human-readable title for the text track, to assist the user in selecting which track to
render. This is especially helpful when there are multiple tracks.
default: Indicates the default text track for the media element when there is more than one
track. This attribute can only appear in one track element per media element.