The short answer
Use timed subtitles for reading alongside video, transcripts for standalone reading and reference, and a separately scoped voice-over script and production process for translated speech.
Define the output
W3C WAI — Captions/Subtitles explains synchronized captions and translated subtitles, including necessary audio information. Terminology varies by region, so describe the actual requirement.
W3C WAI — Transcripts explains text alternatives and relevant visual information. Decide whether the audience needs text inside the video or a separate document.
Provide the right inputs
- The correct video version, duration and spoken language.
- Existing transcript or source subtitles.
- Speaker names and approved terminology.
- Publishing platform and target viewers.
- Required subtitle, script or finished-video outputs.
Flag work-in-progress edits. New cuts may require timing updates even where the translated wording stays the same.

Separate files or embedded text?
A separate subtitle file can support language selection on compatible platforms. Embedded text cannot easily be changed or switched off after export.
Confirm the platform’s supported import format and test a sample. Specify whether the scope includes subtitle placement and video export or only SRT/VTT delivery.
Write for reading time
Viewers need to read while following the picture. Concise wording and sensible line breaks help, but requirements should be agreed for the platform rather than assumed universally.
For training content, check action names, numbers and alignment with on-screen steps. Avoid obscuring important demonstrations.
Review against the video
Watch the actual sequence to check timing, speaker changes, names, cuts and mobile legibility. A text-only review cannot reveal every playback issue.
For voice-over, agree pronunciation, voice, timing and recording/editing responsibilities separately. Script translation does not automatically include production.
Frequently asked questions
Can I use automatic subtitles as the source?
Check them against the audio, terminology and timing before translation or publication.
Does a per-minute quote include video export?
Ask explicitly. Transcription, translation, timing, embedding and export are distinct tasks.
References
Project checklists and examples are written by LocalizeZone. Linked references support the specific statements marked in the article.
Discuss your project
Share a sample video, duration and publishing platform. We can discuss suitable deliverables and the required production steps.

