A worked example
From caption text to a usable subtitle file.
Search a short transcript, inspect its timing, and download the same content in three formats. Everything in this example works without an account.
This is an original written demonstration created for Media Caption. The cue timings are illustrative. It demonstrates text cleanup and exports; it is not an audio recording or an AI transcription accuracy test.
1. Separate the words from caption formatting
Caption tracks can carry italic tags, bold tags and encoded characters. Media Caption removes presentation tags and decodes characters for reading, searching and exporting. It preserves the spoken wording and cue times.
Source caption text
<i>Start with a short recording.</i> Check names & numbers against the audio.
Clean reading text
Start with a short recording. Check names & numbers against the audio.
Removing markup does not correct a misheard name, translate a sentence or recover missing speech. Those need a source-audio review or a separate transcription step.
2. Search the words and inspect an export
Search for “timing” to highlight the last cue. Choose SRT to see numbered cues with comma-separated milliseconds, or VTT for a WEBVTT header and dot-separated milliseconds. TXT is useful when you only need the words; its timestamps follow the checkbox.
Four cues · 15.5 seconds of illustrative timing
- 0.0sStart with a short recording.
- 3.2sCheck names & numbers against the audio.
- 8.0sExport subtitles for your video.
- 11.5sPlay it back and check the timing.
1 00:00:00,000 --> 00:00:03,200 Start with a short recording. 2 00:00:03,200 --> 00:00:08,000 Check names & numbers against the audio. 3 00:00:08,000 --> 00:00:11,500 Export subtitles for your video. 4 00:00:11,500 --> 00:00:15,500 Play it back and check the timing.
3. Check the result against your recording
A correctly formatted file can still contain inaccurate captions. In your own project, listen to the start, middle and end. Check names and numbers, then confirm that cues appear when the words are spoken. If you trimmed or rearranged the video, the original timestamps need adjustment.
Try a video you work with
Paste a supported public YouTube link, then sign in or create a free account to unlock a caption track for 1 credit. Reopening and downloads are included. When captions are missing, AI transcription uses account credits. Review generated text before publishing or quoting it, and use content you have permission to reuse.