Make and review SRT subtitles from a short recording
Turn a short soundtrack into editable captions, then check wording and timing before publishing.
Prepare a short soundtrack
The local Subtitle Generator uses Whisper Tiny. Recordings must be no longer than two minutes and under 40 MB. Start with a clear WAV or MP3 file. If a video does not decode in your browser, export its audio track from your video editor and upload that instead.
The workspace converts audio to mono at 16 kHz for inference. Background music, overlapping speakers, and long stretches of silence can reduce accuracy. This tool produces an SRT text file; it does not burn captions into a video or export an edited MP4.
Create and review the captions
- Open the tool, upload the soundtrack, and select the spoken language where available.
- Choose Create subtitles. First use downloads approximately 42 MB of model files plus runtime files.
- Listen to the recording while reviewing the caption text and timings. Correct names, punctuation, and missing words.
- Choose Download SRT, import the file into your video editor, and check synchronization against the original video.
Understand an SRT block
This illustrative block shows the format, not a measured transcription:
1 00:00:00,500 --> 00:00:02,800 Welcome to the workshop.
The first line is the caption number. The next line gives start and end times in hours, minutes, seconds, and milliseconds. The text follows, with a blank line between blocks. A caption should end after it starts and should remain visible long enough to read.
When you trim or move a clip after making subtitles, its timing can shift. Import the SRT into the same sequence position as the audio you transcribed. Split a long caption into shorter phrases in your editor when it covers too much of the screen.
Common failure cases
If no timed segments are returned, the workspace cannot produce a useful SRT from plain text alone. Try a shorter, clearer speech segment. If the words are wrong, check language selection and remove loud background music. For a longer video, split it into short clips and adjust each caption file’s offset when assembling the final timeline.