SRT (SubRip) is a plain-text subtitle format: numbered cues, a start --> end timestamp, then one or more lines of text. YouTube, many editors, and players accept SRT for captions. This toolkit shifts timing, merges files, and cleans SRT text entirely in your browser.

How do I shift subtitle timing if captions are early or late?

Use Shift times. Positive seconds move every cue later (fix early subs). Negative seconds move cues earlier (fix late subs). Enter 1.5, -2, or minute:second values, then download the new .srt. No video is uploaded — only the subtitle text is adjusted.

How do I merge two SRT files?

Open Merge files, add both (or more) SRT files, set order with the arrows, and optionally add a time offset per file if one track starts later. The tool sorts cues by start time, renumbers them, and lets you download a single merged SRT.

What does Clean SRT do?

Clean renumbers cues, strips common HTML tags (including ), trims whitespace, can collapse blank lines, drop empty cues, optionally strip speaker labels, and shorten overlapping end times so cues do not cover each other. Turn options on or off before download.

Does this convert SRT to VTT or burn subtitles into video?

No. This tool edits SRT text only: shift, merge, clean, copy, download. It does not convert to WebVTT/ASS, burn captions into a video file, or auto-transcribe audio. Use it when you already have SRT and need timing or cleanup fixes.

Will my subtitle files be uploaded to your servers?

No. Parsing and processing run in your browser. Files are not uploaded for conversion. That keeps private captions on your device while you work.

Can I use this for YouTube captions?

Yes. Export or download SRT from YouTube Studio (or your editor), fix drift with Shift, combine language or part files with Merge, tidy formatting with Clean, then re-upload the SRT in Studio. Always preview a few cues against the video after large shifts.

Why are my subtitles overlapping after merge?

Overlaps happen when two sources cover the same time range or offsets are wrong. Use Clean with “Fix overlapping times,” or set per-file offsets in Merge so tracks sit sequentially. Spot-check the cue preview before publishing.

What timestamp format does SRT use?

SRT uses hours:minutes:seconds,milliseconds with a comma, e.g. 00:01:05,200 --> 00:01:08,400. This tool accepts comma or dot milliseconds on import and writes standard comma-based SRT on export.

Is the SRT toolkit free?

Yes. Free, no signup, unlimited local use. Product claims match reality: browser-side text processing, not a cloud caption service or speech-to-text engine.

How do I know how many seconds to shift my SRT?

Pause on a clear spoken word, note the video clock, then compare to that word’s cue start in the SRT. Shift ≈ speech time − caption time. If captions are early, the result is positive (move later). If late, use a negative shift. Check open, middle, and end — if drift changes across the video, a single global shift may not be enough and you may need a new export from your editor.

Can I merge English and Spanish SRT into one file for YouTube?

Usually no — YouTube expects separate caption tracks per language. Merging two languages into one SRT stacks lines on the same timeline and confuses players. Keep one SRT per language and upload each track in Studio. Use Merge for sequential parts of the same language (part 1 + part 2), not for bilingual mashups.

Does shifting SRT fix wrong words in auto-captions?

No. Shift only moves timestamps. Wrong words, missing names, and bad punctuation need manual edit or a better transcript. Use this tool after the text is acceptable and only timing or file structure is wrong.

What is the difference between SRT and VTT?

SRT (SubRip) uses numbered cues and comma milliseconds (00:00:01,000). WebVTT starts with a WEBVTT header and often uses dot milliseconds plus optional styling/notes. This toolkit is built for SRT. If you only have VTT, convert or re-export as SRT from your editor or caption tool, then shift/merge/clean here.

Free SRT Toolkit | Shift, Merge & Clean Subtitles

Complete Guide

The Complete Guide to SRT Subtitles: Shift, Merge & Clean Captions

Your video is ready. Captions almost are — except every line is a second early after the re-edit, or you have two half-files from a long interview, or the export is full of HTML tags YouTube does not need. Hand-editing hundreds of cues is a waste of a creator’s day.

SRT (SubRip) is still the workhorse subtitle format for YouTube uploads, many editors, and simple players. An SRT toolkit that shifts timing, merges files, and cleans formatting lets you fix real caption problems without installing heavy desktop suites or uploading private dialogue to a random cloud converter.

This free tool runs entirely in your browser: drop or paste SRT, choose Shift, Merge, or Clean, preview, then copy or download. It does not transcribe speech, burn captions into video, or convert to every exotic format — and that honesty is intentional. Search phrases this guide answers in plain language include how to shift SRT timing, how to merge SRT files, how to clean subtitle files for YouTube, and how to fix captions that are early or late after a re-edit.

Below is a complete guide to SRT structure, measuring drift, when to shift vs merge vs clean, worked examples, troubleshooting, privacy limits, and a safe YouTube Studio workflow — written for real packaging work, not inflated claims.

For on-video chapters (description timestamps), use the YouTube Chapters Builder instead — different job, same packaging mindset.

In this guide

›What an SRT toolkit is ›How to use it ›Why captions timing matters ›SRT format basics ›Shift times ›How to measure drift ›Merge files ›Clean SRT ›YouTube workflow ›Worked examples ›Common mistakes ›Troubleshooting ›Privacy & limits ›Quick reference

FoundationsWhat is an SRT toolkit?

An SRT toolkit is a set of utilities for SubRip subtitle files: parse the cues, change their timing, combine tracks, normalize text, and write a valid .srt again. Creators use it when captions exist but need mechanical fixes — not when they need a first transcript from audio.

This page’s tool focuses on three jobs. Shift: add or subtract the same amount of time from every cue. Merge: stack multiple SRT files (with optional offsets) into one numbered track. Clean: renumber, strip common HTML, trim whitespace, drop empty cues, optionally remove speaker labels, and repair simple overlaps.

Who it is for

YouTubers fixing Studio or editor captions that drifted after a re-cut.
Podcasters & interviewers combining part-1 / part-2 subtitle exports.
Course creators cleaning markup-heavy exports before LMS or YouTube upload.
Editors who want a fast browser pass without opening a full NLE.

What this tool is not

It is not automatic speech recognition, not forced alignment of audio, not burn-in / hardcode subtitles into MP4, and not a full subtitle studio for styling ASS/SSA. Bring an SRT; leave with a better SRT.

WorkflowHow to use the SRT toolkit (step by step)

The workspace sits above this guide (tool-first layout, same idea as our PDF tools). Pick a job, load files, set options, preview, download.

Choose Shift, Merge, or Clean

Shift for global drift. Merge for multiple files. Clean for numbering, tags, empties, and overlaps. You can clean after a merge by downloading, then switching to Clean with the new file.

Load SRT

Drop a .srt, browse, load the sample, or paste text (single-file modes). Merge accepts several files — order them and set per-file offsets if needed.

Set the job options

Shift: enter seconds (1.5, -2) or minute-style values. Merge: offsets in seconds per file. Clean: toggle renumber, strip HTML, speaker labels, whitespace, overlaps, drop empty.

Preview

Check cue count, time span, full text, and the first cues. Parse errors and warnings appear on the loaded file cards.

Copy or download .srt

Download for Studio or your editor. Spot-check three moments in the video (open, mid, end) after any large shift.

StrategyWhy subtitle timing and cleanup matter

Captions are accessibility first: deaf and hard-of-hearing viewers, noisy environments, language learners, and anyone who watches muted on mobile. Timing that is consistently early or late makes captions unusable even if the words are correct. Overlapping cues and HTML junk can confuse players or look broken on upload.

On YouTube, captions also support discovery and comprehension. They are not a substitute for a strong title and thumbnail, but broken captions still hurt trust. A quick shift/clean pass is cheaper than redoing a full retime by hand.

Weak approach	Better approach
Ignore 1s drift on a 20-minute video	Shift the whole SRT by −1s and re-check
Manually renumber 400 cues	Clean → Renumber
Upload HTML-heavy export as-is	Clean → Strip HTML
Leave two half-files for one video	Merge with offset = length of part A

BasicsSRT format in plain language

Each cue is a block: optional index number, a line with start --> end, then one or more text lines, then a blank line. Timestamps use hours:minutes:seconds,milliseconds with a comma in classic SRT (dots are often accepted on import).

1
00:00:01,000 --> 00:00:03,500
Welcome to the video.

2
00:00:03,800 --> 00:00:07,200
Today we fix drifting captions.

Rules of thumb

End should be after start on every cue.
Prefer short lines; two lines max is a common readability target.
Keep encoding UTF-8 for non-English text.
Blank line between cues — this toolkit re-serializes cleanly on export.

Job 1Shift times: fix early or late captions

When every line is early, add a positive shift (captions move later). When every line is late, use a negative shift. Measure drift by pausing on a clear word and comparing spoken audio to the on-screen caption time.

Symptom	Typical shift
Captions appear before speech	Positive (e.g. +0.5s to +2s)
Captions lag behind speech	Negative (e.g. −0.5s to −2s)
Intro cut removed from video	Negative by cut length
Silence/slate added at start	Positive by added length

Clamping at zero

If a negative shift would push start times below 0:00, this tool clamps at zero and shortens the duration as needed so the file stays valid. Extreme negative shifts on cues near the open can squash early lines — re-check the first few cues.

Shift will not fix random per-line errors or wrong words. Those need an editor or a new transcript. Shift is for global, consistent drift.

Input formats this tool accepts for shift

Seconds as a number: 1.5 , -2 , 0.25
Minute:second style: 1:30 for ninety seconds (and optional fractional forms where supported)
SRT-like clock values when you paste a full timestamp style string the parser understands

Start with a half-second or one-second nudge, download, and check a known line of dialogue. Large multi-second shifts are fine when you cut a long cold open, but always verify the open, a mid-roll section, and the outro — cuts in the middle of a video can make a single global shift imperfect if only part of the timeline moved.

MethodHow to measure subtitle drift against the video

Guessing the shift amount wastes time. A simple measurement routine gives you a number you can trust before you touch the whole file.

Pick a clear word early in the video

Choose a word that is easy to hear and easy to see on the caption line — a proper name, a number, or the first content word after the intro music.

Note the video clock when the word is spoken

Pause the player on the exact frame (or as close as your player allows). Write down the timeline time — for example 0:12.400.

Note when that word appears in the SRT

Open the SRT cue that contains the word. Use the cue’s start time (or the moment the word is meant to show). Example: cue starts at 0:11.200 but speech is at 0:12.400 → captions are about 1.2 seconds early.

Compute shift = speech_time − caption_time

If speech is later than the caption start, you need a positive shift. If speech is earlier, you need a negative shift. In the example: 12.4 − 11.2 = +1.2 seconds.

Apply, then verify a second and third cue

Check one cue near the middle and one near the end. If mid/end drift differs a lot from the open, a single global shift may not be enough — the edit may have changed only part of the timeline, and you may need a new export from the editor.

Constant drift vs stretch

Global shift fixes constant offset (everything is early by the same amount). It does not fix “stretch” problems where captions get more wrong over time because the video frame rate or speed changed. Stretch issues need a retime in an editor or a new caption export.

Job 2Merge files: one track from many

Merge is for sequential parts (interview hour 1 + hour 2), alternate takes, or multi-export pipelines. Load multiple SRTs, set order, and apply an offset to later files so their timeline continues after earlier content.

Offset examples

Part B starts after 1 hour of Part A: set Part B offset to 3600 seconds (or 1:00:00 style if you enter supported formats).
Same timeline, two languages: merging into one SRT usually confuses players — keep languages as separate uploads instead of one mixed file.

After merge, cues are sorted by start time and renumbered. If two sources overlap in time, use Clean → fix overlaps, or fix offsets so tracks are sequential.

Job 3Clean SRT: renumber, strip, fix overlaps

Clean options are toggles. Defaults favor common YouTube-ready cleanup: renumber, strip HTML, trim whitespace, collapse extra blank lines, fix overlaps, drop empty cues. Speaker-label stripping is off by default (useful for interviews when you want “ALEX:” removed).

Option	What it does
Renumber	Rewrites cue indices 1…n in order
Strip HTML	Removes tags; turns <br> into line breaks
Strip speaker labels	Drops [Name] or NAME: prefixes on lines
Trim whitespace	Collapses extra spaces, trims lines
Collapse blank lines	Removes empty lines inside cue text
Fix overlaps	Shortens previous end if it runs into next start
Drop empty	Removes cues with no text after cleanup

Overlaps

Overlap fix is conservative: it shortens the earlier cue’s end rather than rewriting speech. If two full sentences truly share a time range, you may still need manual edit — but simple export glitches clear up quickly.

YouTubeYouTube Studio caption workflow

Typical path: download existing captions or export SRT from your editor → fix in this toolkit → upload SRT in YouTube Studio (Subtitles). Prefer “upload file” over pasting broken fragments when possible.

After re-trimming a video, re-measure drift before uploading old captions.
After shift, play three checkpoints: intro, midpoint, end card.
Keep a master SRT next to the master video for future cuts.
Pair captions with description and chapters for full packaging.

Auto-captions are a starting point. Published quality still depends on your review. This tool speeds mechanical fixes; it does not replace reading for accuracy.

ExamplesWorked examples you can copy as a habit

Example A — Intro cut by 2 seconds

You removed a 2-second slate from the start of the master. Captions still include the old open, so every line is late by about two seconds. Shift by −2 , download, and check the first spoken line. If the slate had no dialogue, you may also delete the first empty or music-only cues with Clean → drop empty after a manual pass.

Example B — Two-part interview

Part A is 58 minutes; Part B’s SRT restarts at 0:00. In Merge, load Part A then Part B. Set Part B offset to 3480 seconds (58 × 60). Preview: Part B’s first cue should land near 58:00. Download one file for the combined upload.

Example C — HTML from an editor

Export shows lines like <font>Hello</font> or <b>Really?</b>. Use Clean with strip HTML on. Renumber and fix overlaps if the export also stacked ends. Re-upload to Studio and confirm styling is plain text.

Example D — Auto-captions slightly early

YouTube auto-captions sit ~0.6s early on a talking-head video. Measure with the drift method above, shift +0.6 , spot-check three points, then replace the track in Studio. Still edit wrong words — shift does not fix vocabulary.

Save masters

Keep original.srt and fixed.srt next to the video master. If you re-cut again next month, you can re-shift from the original instead of stacking shifts on an already shifted file (which invites rounding confusion).

AvoidCommon SRT mistakes

Shifting the wrong direction — early captions need +time; late captions need −time. Test with 0.5s first.
Merging without offsets — both files start at 0:00 and stack on top of each other.
Expecting transcription — empty audio will not produce SRT here.
Uploading uncleaned HTML — strip tags when Studio or players show raw markup.
Skipping the spot-check — always verify a few cues against picture after large shifts.
Using one merged file for two languages — upload separate tracks instead.
Stacking shifts blindly — if you shift +1 then later +0.5 from the already shifted file, document the total (+1.5 from original) so you can reverse it.
Fixing stretch with shift only — if drift grows over time, re-export captions from the final timeline instead.

Fix itTroubleshooting: when the output looks wrong

No cues found

The file may not be SRT (try opening in a text editor). WebVTT starts with WEBVTT and uses different timing. Convert or re-export as SRT from your tool, then load again. Also check for empty files or binary content.

Parse warnings / skipped blocks

Some exports insert notes or junk blocks without timestamps. The parser skips blocks it cannot read and continues. Review warnings on the file card; critical dialogue should still appear in the cue list.

Overlaps after merge

Offsets are wrong or sources share the same timeline. Adjust offsets so parts are sequential, or run Clean → fix overlaps knowing that shortens ends rather than rewriting words.

YouTube rejects the file

Re-download from this tool (clean serialization), ensure UTF-8, avoid empty cues, and confirm extension .srt. If Studio still fails, open the file and confirm classic arrow timestamps with commas for milliseconds.

Mid-video still wrong after a perfect open

You likely have non-uniform timeline changes (inserts/deletes mid-edit). Global shift cannot fix that. Go back to the editor, export captions from the final sequence, or split the SRT and shift sections separately (advanced; merge afterward with care).

TrustPrivacy, limits, and honest product claims

All parse, shift, merge, and clean work runs in your browser session. We do not require an account to use the toolkit, and the product is not designed as a cloud caption host. That matches FreeCreatorTools’ privacy-first stance: local processing for files you may not want on a third-party server before publish.

Hard limits of this tool

SRT text only — not automatic speech-to-text.
Not burn-in / hardcode into MP4, MOV, or WebM.
Not a full ASS/SSA styling editor or karaoke designer.
Not a guarantee of YouTube policy compliance for every caption edge case — always preview in Studio.
Very large files may stress mobile browsers; desktop is more comfortable for multi-hour cue lists.

If you need chapters on the progress bar (description timestamps), that is a different feature — use the Chapters Builder. If you need audio loudness or GIF conversion, those are separate roadmap tools. This page stays focused so claims stay accurate.

Cheat sheetQuick reference

Item	Guideline
Format	SRT (SubRip) plain text
Timestamp	HH:MM:SS,mmm --> HH:MM:SS,mmm
Shift +	Captions later (fix early)
Shift −	Captions earlier (fix late)
Merge offset	Seconds to add to that file’s cues
Clean defaults	Renumber, strip HTML, overlaps, drop empty
Privacy	Browser-only — no upload to our servers
Not included	STT, burn-in, VTT/ASS export

Ship checklist

Valid SRT parse · correct job · preview looks sane · three video checkpoints · download · upload in Studio · final watch-through of captions.

Wrap-upFix the file, protect the watch

Good captions are timing plus text. When the words are fine but the clock is wrong — or the file is messy — a focused SRT toolkit saves hours. Shift for drift, merge for multi-part exports, clean for structure, then re-upload with confidence.

Scroll back to the tool at the top of this page, load your .srt, and ship captions that stay readable for every viewer who needs them.