Scene 01Take 1●
Try a short example with captions and a transcript
Play the 30-second demonstration below with captions enabled. It uses a fictional wedding-website welcome and synthetic narration. Eight short phrase cues accompany the three ten-second scenes. The script and simple visual placeholders show the structure without exposing a real family’s photographs or voices.
Download the WebVTT captions, SubRip captions or descriptive transcript. The transcript identifies the narrator, scene headings and planned visual placeholders as well as the spoken words.
The captions preserve every spoken word and use one or two lines per cue. Their timing follows the saved narration and its pauses, with short reading holds. They were checked in the browser at phone width, including scene changes and visible player controls. This is phrase-based captioning, not word-by-word highlighting or a guarantee that every player displays captions identically.
Scene 02Take 1●
Use the right text alternative for the information
| Resource | What it carries | Example use |
|---|---|---|
| Captions | Speech, speaker identification when needed, and meaningful non-speech sound, synchronized with the video | A birthday message that should remain understandable with sound off |
| Transcript | Readable text of the audio, with important visual information added for a descriptive transcript | Names, dates and a scene’s essential action available without operating the player |
| Description of visual information | The visual facts that a listener needs and cannot infer from the soundtrack | Explaining who appears in an otherwise silent historical photograph |
W3C’s captions guidance and transcript guidance explain these different jobs. Captions do not automatically describe a photograph, and a transcript does not replace synchronized captions for someone watching a speaking guest. These references were checked on 24 October 2026.
Scene 03Take 1●
Prepare accurate words before polishing the appearance
- Transcribe the final edit rather than an earlier recording. Correct names, dates and place names with the contributor.
- Identify a speaker when the image or context does not make that clear. Include meaningful sounds such as an off-screen laugh or applause when they affect understanding.
- Time each cue to the relevant speech or sound. Split long passages at sensible phrase boundaries and leave enough reading time.
- Watch the entire exported version with the caption track enabled. Check cuts, overlapping voices, the first cue and the final line.
Automatic transcription is a starting draft. It can mistake an unfamiliar name or turn a quiet sentence into unrelated words. Ask before guessing a sensitive or unclear phrase. For a memorial, let the family approve names, relationships and the wording of personal details before sharing the transcript.
The downloadable example divides longer sentences into readable phrases instead of keeping a full paragraph on screen for ten seconds. Its longest cue lines have at most 37 characters, and its reading rate stays below 17 characters per second. Those are review choices for this sample; judge your own names, language, audience and actual player.
Scene 04Take 1●
Keep text readable over the actual picture
Use a solid or sufficiently opaque caption background when footage changes from dark to bright. Leave space around the text and review it over both faces and detailed photographs. Avoid putting important names where captions or player controls will cover them.
On a phone, test portrait and landscape playback with controls visible, then hide the controls and check again. A line that fits comfortably on a desktop can wrap into several lines in a narrow player. Increase the space available, shorten a cue without losing meaning, or divide it into phrases before simply shrinking the font.
Selectable caption tracks let viewers turn captions on and may let them choose display preferences. Burned-in captions are part of the image and cannot be switched off. They can be useful where a destination discards caption files, but they need their own readability review and do not provide a selectable text transcript.
Scene 05Take 1●
Add captions using a player or an external publishing tool
This example uses a separate WebVTT file attached to the blog’s HTML video player. Its MP4 does not contain burned-in captions. Downloading only the MP4 therefore does not give a recipient the separate caption track; include the text files or use a player that supports them.
For a video you are authorized to publish on YouTube, its subtitle and caption workflow supports uploading a timed file in YouTube Studio. Its supported-format reference covers basic UTF-8 SRT files. Review the imported timing and text before publishing. This is a documented external option, not a claim that the sample was uploaded to YouTube.
For another editor, confirm that it imports your caption format and whether the final export preserves a selectable track or burns text into the image. Test the delivered file in its intended player. This tutorial’s caption files and blog player are separate from AlbumMap’s map-video editing controls.
End card
That’s a wrap
- Try a short example with captions and a transcript
- Use the right text alternative for the information
- Prepare accurate words before polishing the appearance
- Keep text readable over the actual picture
- Add captions using a player or an external publishing tool
- Share the version that your audience can actually use
Helpful answers
Frequently asked questions
Are captions and a transcript the same file?
Captions include timing for display during playback. A descriptive transcript can be read independently and includes important visual context as well as the audio information.
Does the sample MP4 include captions when downloaded alone?
No. The blog player loads a separate WebVTT track. Include the caption file in a compatible player, or make and check a separate export with burned-in captions if the destination needs that format.
Can I trust automatic captions without reviewing them?
Review them against the final edit, especially names, dates, quiet speech and overlapping voices. Seek approval before guessing private or sensitive wording.
