Website video caption publishing workflow
Captions need to match the same version as sound and picture. Speech, relevant sounds and speaker identity should remain understandable without hearing the audio. Automatic text needs review.
- Starting condition
Captions need to match the same version as sound and picture. Speech, relevant sounds and speaker identity should remain understandable without hearing the audio. Automatic text needs review.
- Implementation
Connect video file, language track, caption file and version ID in publishing. Include language selection, timing review and correction ownership. Assess transcripts and audio description separately against the content’s needs.
- Verification
The edited passage and ending follow speech in every published language track.
- Decision
Bind approval to both video and track versions. Review timing after edits and verify the file referenced by the HTML.
The edited passage and ending follow speech in every published language track.
Which check determines the next step?
Bind approval to both video and track versions. Review timing after edits and verify the file referenced by the HTML.
Scope the development
Captions need to match the same version as sound and picture. Speech, relevant sounds and speaker identity should remain understandable without hearing the audio. Automatic text needs review.
ImbaHub can help with a bounded implementation on your website. We start from the real environment, affected page types and an agreed visitor task.
What the implementation needs
Connect video file, language track, caption file and version ID in publishing. Include language selection, timing review and correction ownership. Assess transcripts and audio description separately against the content’s needs.
Delivery covers the agreed components, configuration, test cases and documentation. Start with a representative page. Define languages, devices and external dependencies before extending the work to the rest of the website.
| Check | Expected outcome |
|---|---|
| Starting condition | Captions need to match the same version as sound and picture. Speech, relevant sounds and speaker identity should remain understandable without hearing the audio. Automatic text needs review. |
| Implementation | Connect video file, language track, caption file and version ID in publishing. Include language selection, timing review and correction ownership. Assess transcripts and audio description separately against the content’s needs. |
| Verification | The edited passage and ending follow speech in every published language track. |
| Decision | Bind approval to both video and track versions. Review timing after edits and verify the file referenced by the HTML. |
A fictional website example
Fictional example: An instruction film gains a safety step halfway through. Old captions remain attached, so every later line appears too early.
Checks before acceptance
The edited passage and ending follow speech in every published language track.
Compare the opening, an edited passage and the ending against the audio. Mute playback and check speaker changes and meaningful sounds. Test every track in the delivered player rather than only reading the file.
Delivery and ongoing ownership
Assign responsibility for changes after handover and document how to restore the previous working version. Scope and pricing follow review of the current solution and access; fixed packages or outcome levels are not assumed.
Sources and further reading
Questions, answered
What should we send to scope the work?
Connect video file, language track, caption file and version ID in publishing. Include language selection, timing review and correction ownership. Assess transcripts and audio description separately against the content’s needs.
Which check determines the next step?
Bind approval to both video and track versions. Review timing after edits and verify the file referenced by the HTML.