HLS WebVTT handoffs: timestamp maps and segment boundaries

Identify the local and media timing relationship

Inspect the authorized text resource for a WebVTT signature and timing-map information. LOCAL describes a cue-time anchor and MPEGTS describes its corresponding media timestamp on the defined transport clock. Keep the presentation's discontinuity and timestamp-wrap context; a raw number interpreted as milliseconds can place otherwise valid text at the wrong moment.

For illustration, LOCAL 00:00:00.000 paired with MPEGTS 900000 corresponds to a 10-second media-clock anchor on a 90000-unit-per-second clock. That arithmetic does not establish the zero point of a finished local movie. Determine the actual presentation-to-file mapping through a format-aware packaging or conversion workflow before using the cue times in a new standalone sidecar.

Check cues that span a segment boundary

A cue overlapping more than one media interval can appear in the relevant WebVTT segments according to the HLS requirements. A naive concatenation can therefore create duplicated dialogue or invalid repeated headers. Preserve the full cue interval and original context when reviewing a segmented-to-sidecar handoff.

Do not deduplicate simply by matching words: an instruction can genuinely repeat later in the recording. Compare actual mapped timing, cue payload and source context. Identifiers can be helpful references but are not universally unique proof of one spoken event. Test a crossing cue against the visible scene before and after the media boundary.

Distinguish a valid quiet interval from unavailable text

A valid caption segment covering an interval without cues can be empty of dialogue. Compare its structure and the expected playlist interval rather than declaring every small response corrupt. An HTTP failure, an HTML error page or a missing referenced resource is a different observation from a valid WebVTT segment with no speech to display.

Review the saved output in its actual receiver, including mapped timing, crossing cues and an interval that should be quiet. Keep a source copy and document any conversion losses. SocialVidDownloader does not expose an HLS WebVTT assembler or promise source-caption export. Use an authorized format-aware text workflow; do not scrape private delivery tokens or guess offsets to manufacture a successful caption handoff.

Worked example

Example: a sentence begins before one segment ends and continues into the next. Both relevant text segments can describe it. A standalone caption conversion should present the intended interval once, with correct mapped timing; deleting every repeated phrase could instead remove a later genuine repetition of the instruction.

Related guidance

• WebVTT parsing: keep headers, cue text and metadata distinct
• Captions appear too early or late: offset versus drift

Sources and verification

• HLS WebVTT segment and timestamp requirements

Reviewed 2026-10-12. Product behavior is limited to available source formats and provider access; it is not a promise that every public link works.