Subtitle Timing Shifter in the browser is not a generate
/free-tools/subtitle-timing-shifter never spends a credit. Do not pay a model to do an in-browser file job.
/free-tools/subtitle-timing-shifter never spends a credit. Do not pay a model to do an in-browser file job. A uniform offset is arithmetic: every cue moves by the same number of milliseconds. That is the File API and a timestamp rewrite. It is not speech-to-text, not a caption style, and not a reason to open a generator.
Versely has no free tier for generation. Every model costs credits; the app quotes the figure before the charge. Utilities that never touch a model sidestep both the pricing page and the margin on anonymous traffic. They are free because they are free to run.
Shifting timestamps is not a generate
Enter an offset in seconds — negative to pull captions earlier, positive to push them later — and every cue moves by exactly that much. This is the fix when a whole track is uniformly early or late, which is what happens after an intro is trimmed.
The file never leaves the device. No upload, no account, no watermark, and the page keeps working with the network off. That is not a claim about encryption. There is no transmission. Cues that would land before zero are clamped to 00:00:00 rather than dropped. If several cues clamp, the offset is larger than the gap at the start of the file. Shrink the offset. Do not ask a model to “fix the start.”
Uniform offset versus progressive drift
A constant offset fixes a track that is uniformly early or late. Progressive drift means the caption track and the video are running at different frame rates — a 25 fps track against 23.976 fps footage, typically — so the error grows with time. That needs the timings rescaled, not shifted. The shifter will not fix it, and a generate will not diagnose it.
Parse rules are mechanical. WebVTT requires the literal WEBVTT first line. SubRip separates seconds from milliseconds with a comma, WebVTT with a dot. A file with the wrong separator parses as zero valid cues. None of that is a prompt problem. Shift after the trim, or every timestamp is attached to the wrong frame.
Credits exist for models, not for SRT arithmetic
What it costs is the formula book for jobs that actually meter: per second, per 1,000 characters, per megapixel, per tier, per export, captions as a render times a 1× or 2× preset. Credits are the common unit those meters resolve into. A timestamp rewrite has no meter. Putting it through a video model is how a bootstrap spends a clip to do a subtraction.
Do not regenerate the video so the speech “lands on” an old SRT. Do not re-transcribe because the track is 1.2 seconds late — transcription is a caption render (BASIC presets at 1×, DYNAMIC at 2×). An offset is free. Style is a preset multiplier. Timing is the cue list. Quote credits for generation. Use the browser for the file.
Use the captioner when you still need speech-to-text
If there is no subtitle file yet, you are not in this tool. Add captions to a video transcribes the spoken audio and burns styled, timed lines — 21 BASIC presets, 9 DYNAMIC, 165 language codes. That listens. The shifter does not listen. It moves what you already have.
If the words are yours and they change on a clock you can name, that is timed overlays, not an SRT offset. If you need one persistent headline, that is a text overlay. If the whole track is uniformly late after a trim, that is this page.
Three objects. One of them costs nothing and should be done on the machine you already have.
FAQ
Does the shifter spend Versely credits?
No. It never calls a model, an API, or the network. Generation on Versely always costs credits; this is not generation.
Will it fix captions that drift further out as the video goes on?
No. A constant offset cannot fix a frame-rate mismatch. Progressive drift needs the timings rescaled. Shift only when every cue is early or late by the same amount.
Do my subtitle files get uploaded?
No. The file is read into memory in your browser, rewritten there, and offered as a download. Disconnect the network and it still works.
When should I pay for captions instead?
When you do not have a cue list yet, or when the words on screen should be the speech in the file. That is the transcribe-and-burn path. When you already have an SRT that is uniformly late, shift it here and keep the credits.