Subtitle Time Shifter for SRT and VTT
Subtitle time shifter moves every supported SRT or VTT cue earlier or later by the offset you choose.
UTF-8 source only. Maximum 200,000 characters and 2,000 cues. Files are read in this page; no video upload or transcription. The page loads analytics scripts, so avoid confidential captions.
Paste text or choose file. File extension is a hint, not proof of format.
Positive moves captions later; negative moves them earlier. 1500 = 1.5 seconds. Negative underflow rejects entire file.
Supported subset: numbered sequential UTF-8 SRT with comma milliseconds, or WebVTT with dot milliseconds. VTT NOTE, STYLE, REGION, cue settings and header text remain unchanged for same-format shifting; rendering semantics are not checked. Inline timed payloads stop shifting; markup, settings, metadata and nonserial identifiers stop cross-format conversion rather than disappear. Output uses LF and omits BOM. Overlap is advisory; end not after start blocks task use. See W3C WebVTT draft and Library of Congress SRT description. SRT has no universal formal grammar.
Why shift subtitle timing here?
One offset for every cue
Move a track by one constant number of milliseconds when its delay stays the same throughout.
Keep the file format
Review SRT or WebVTT timing changes, then export in the original format.
Stop before negative time
If any cue moves before zero, the whole shift stops. No zero-duration cue or partial export is offered.
What does a subtitle time shift change?
Every cue has a start and end. A positive offset delays both; a negative offset moves both earlier. For example, add 1500 milliseconds to move an SRT start of 00:00:01,000 to 00:00:02,500.
If speech begins at 12.5 seconds while its caption appears at 11 seconds, try +1500 ms. This estimate comes from a line you identify, not audio analysis. Check first and last cues against media: one offset cannot repair drift that grows across a recording.
What stays with the cue?
- Text and identityThe original cue text, line breaks and identifiers remain beside their shifted times.
- Time notationSRT uses comma milliseconds, while WebVTT uses dot milliseconds. Shifted times keep source notation.
- Rich WebVTTNOTE, STYLE, REGION and settings stay in place, but timed tags inside cue text stop export because they also need shifting.
Common Use Cases
Constant late captions
Move cues earlier if every caption appears after the corresponding spoken line, provided none starts below zero.
Constant early captions
Delay cues if they appear before speech. A +1500 ms shift moves both boundaries of each cue.
Corrected track
Compare the first and last timestamps before saving a new same-format track for an editor or player.
How to shift SRT or VTT timing
Add timed captions
Paste SRT or VTT text, or open a UTF-8 subtitle file. Confirm the detected format.
Choose signed milliseconds
Enter a whole number: positive delays, negative shows captions sooner, and zero changes no times.
Review cue times
Run the shift; fix malformed ranges or reduce a negative offset if any cue would start before zero.
Save the changed track
Copy or download only the current result with the original subtitle extension.
Review before saving
Current result only
Editing source, format or offset clears the prior output so it cannot be exported by mistake.
Named underflow
A rejected shift names the affected cue instead of clipping it to zero and changing its duration.
No false drift fix
Every cue moves by the same amount. Gradually changing delay needs another correction method.
Before you export
- •Check direction: Positive means later; negative means earlier.
- •Pick a spoken line: Use a line with recognizable speech to estimate the offset.
- •Inspect both ends: Check cue times near the beginning and end against the recording.
Shift a supported subtitle track
Paste or open SRT or VTT above, then choose a whole-millisecond offset.
Review the changed times and save the result, or correct any source problem before trying again.
Related tools
Frequently Asked Questions
How do I shift SRT subtitles?
Paste or open an SRT, enter a signed millisecond offset, run Shift, then download .srt. A positive offset delays cues.
Does it also shift VTT?
Yes. Supported WebVTT cues keep dot milliseconds and export as .vtt. Inline timed cue text stops export instead of being left behind.
What happens before zero?
The whole shift is rejected and no subtitle file is exported. Choose a smaller negative offset.
Will this correct gradual drift?
No. One constant offset moves all cues equally; it cannot repair timing that drifts across a recording.