Subtitle Trimmer

The Subtitle Trimmer cuts an SRT down to a single time window and rebases it so the clip begins at zero. You paste the full track into the SRT Content box, type a Start timecode and an End timecode, and the tool returns a new SRT containing only the cues that fall inside that window, with every timestamp shifted back by the start value and the cues renumbered from 1. The output downloads as a .srt file.

This is the everyday companion to cutting a clip out of a longer programme. Social teams pulling a two-minute highlight from an hour-long panel, e-learning producers slicing a lecture into modules, and journalists preparing an excerpt for broadcast all end up with a trimmed video and a subtitle file that still thinks it belongs to the full-length master. Loading that untrimmed SRT against the clip shows nothing at all for the first few minutes, because every cue is timed far beyond the end of the new video. Doing the arithmetic by hand across a few hundred cues is exactly the sort of task that produces off-by-one-second errors.

The logic is precise about boundaries. Both timecodes are converted to milliseconds, then a cue is kept if its end time is later than your start and its start time is earlier than your end - in other words, any cue that overlaps the window at all is retained, including one that was already on screen when the clip begins and one still on screen when it ends. Each surviving cue then has the start value subtracted from both its start and its end, with a floor at zero so a cue straddling the in-point simply begins at 00:00:00,000 rather than going negative. If either timecode is missing or cannot be parsed, the tool tells you instead of guessing.

Everything is computed in the browser with no upload, so unreleased footage transcripts stay private and even a long master track trims in an instant.

• Precise two-point window: The Start timecode and End timecode fields define the extract exactly, using the same HH:MM:SS,mmm form you already read off your editing timeline.
• Overlap-inclusive selection: A cue that is only partly inside the window is kept rather than discarded, so a line of dialogue already on screen at the in-point is not lost from the clip.
• Automatic rebase to zero: Every retained cue is shifted back by the start value, which is what makes the output line up with a video that was cut at the same point.
• Negative-time protection: Cues straddling the in-point are clamped to 00:00:00,000 instead of producing invalid negative timestamps that break parsers.
• Clean renumbering: The clip is written out with sequential indices from 1, so it is a valid standalone subtitle file rather than a fragment with gaps in its numbering.
• Clear input validation: If a timecode is blank or malformed the tool asks for both timecodes rather than silently returning an empty or wrong result.
• Local and immediate: The trim runs in your browser with no file upload, and the result is downloadable straight away as a .srt.

1. Note the in-point and out-point of your video clip from your editor, written as hours, minutes, seconds and milliseconds.
2. Paste the full-length subtitle track into the SRT Content box, or click Load sample to try the tool on a demo file first.
3. Type your in-point into Start timecode, for example 00:10:00,000, matching the exact frame where your trimmed video begins.
4. Type your out-point into End timecode, for example 00:12:30,000; anything wholly outside this window is dropped.
5. Run the trimmer and confirm that the first cue in the output now starts at or near 00:00:00,000 and that the last cue ends before your clip duration.
6. Download the result as a .srt file and load it alongside the trimmed video to check the first and last lines land on the right words.

Example 1 - a 30-second highlight from a long interview. The clip starts at ten minutes into the master, so every kept cue is shifted back by exactly ten minutes.

Start timecode: 00:10:00,000
End timecode:   00:10:30,000

Before
145
00:10:02,400 --> 00:10:05,100
The turning point was the second quarter.

After
1
00:00:02,400 --> 00:00:05,100
The turning point was the second quarter.

Example 2 - a line already on screen at the in-point. The cue ran from 00:09:58,000 to 00:10:01,500, so it overlaps the window and is kept, then clamped at the front:

After
1
00:00:00,000 --> 00:00:01,500
...which nobody expected.

Without the clamp this would have become a negative start time and been rejected by most parsers.

• Clips ship with captions on the first try: The trimmed subtitle file drops straight onto the trimmed video with no manual offset step and no guesswork.
• Eliminates hand arithmetic: Subtracting an offset from several hundred timestamps by hand is slow and error-prone, and a single mistyped digit desyncs the rest of the file.
• Keeps accessibility intact on excerpts: Social and promo clips stay captioned, which matters for platform requirements and for viewers watching with sound off.
• Nothing important gets cut off: Because partly overlapping cues are retained, the first and last lines of dialogue in your clip are still there.
• Produces a valid standalone file: Renumbering and zero-clamping mean the output passes validation instead of tripping parsers with odd indices or negative times.
• Unreleased material stays private: The trim happens in your browser, so rough cuts and embargoed transcripts are never uploaded anywhere.