Keep the text you approved
A script, a reviewed verbatim transcript or a proofread caption file goes in as is. Sonix adds the timing and never rewrites a word.
Reviewed by David Dat Nguyen, CXO Last updated
Existing transcripts
Your words.
Our timing.
Paste a transcript or an SRT with the media, and Sonix aligns every word to the recording, so a document becomes a synced transcript with timecodes, subtitles and all the exports.
Scripts, verbatim transcripts from another service, old subtitle files: aligned instead of transcribed, with the text exactly as you wrote it.
30 free minutes on every new account, no credit card.See pricing
Pasted text or SRT
Your wording, untouched
Every export afterwards
Your text, now synced
Three reasons to align instead of transcribing again.
Get everything a transcript gets
Once aligned, the text behaves like any Sonix transcript: click a word to play it, export DOCX with timecodes, SRT and VTT subtitles, CSV, or editor timelines.
Rescue old subtitle files
Paste an SRT beside its video and the cues become a word-timed transcript you can re-split, re-time and export in a new format.
When the words already exist
How three kinds of teams align.
Transcription agencies
Timecodes and subtitles added to work the team already finished.
- Align a proofread transcript to deliver a timed DOCX and an SRT from one upload
- Turn a client's legacy transcripts into synced files without re-transcribing
- Split speakers from the text's own labels when the pasted transcript carries them
Podcasters
The show notes and the episode page, from a script you wrote.
- Align the episode script to the final audio and export chapter timecodes
- Publish a synced transcript on the episode page from the approved copy
- Produce a VTT for the player from the same alignment
Accessibility teams
Compliant captions from text that has already been reviewed.
- Align a corrected transcript to the lecture video and export captions
- Re-time an old SRT that drifted after the video was re-edited
- Set lines, characters per line and cue length for the platform
How alignment works
The settings and limits, as they are in the product today.
- An option on the upload formAlign
- Turn on "Align a transcript" in your upload settings, pick that card when you upload the media, then paste the text or the SRT. The button reads Align instead of Transcribe.
- Text or SRTTXT or SRT
- Plain text works line by line. An SRT keeps its cues as the starting paragraphs; the form warns you if a pasted file looks like SRT so you can pick the right format.
- Speakers come from the textFrom the text
- Alignment does not detect voices. Keep speaker labels in the transcript, or through the API set split_speakers so lines starting with a name become labeled paragraphs.
- Billed like a transcriptionSame hours
- An alignment uses the same hours as transcribing the file, on every plan including the trial minutes.
- Realign after editsTools menu
- Edit the aligned text as much as you like, then run Realign timecodes from Tools to retime every word against the audio.
- The APItranscript_text
- Send transcript_text or a transcript_url with the media and the optional split_speakers flag, and the file is aligned instead of transcribed.
Questions teams ask about
existing transcripts
How would a transcription agency use existing transcript alignment?
Upload the client's audio with the proofread transcript pasted in, and the finished text comes back timed to the word. Export a DOCX with paragraph timecodes and an SRT for the same recording, then deliver both. Legacy transcripts the client already paid for can be aligned the same way instead of being transcribed again.
Does it change my wording?
No. Alignment places your words on the recording and leaves the text as you provided it. If the audio and the text differ in places, the nearby words are placed as closely as the audio allows, and you can correct them in the editor.
Is aligning cheaper than transcribing?
It is billed the same way: an alignment uses the hours for the length of the file, like a transcription. What you save is the editing, since the text is already right.
Will it detect speakers?
Not from the audio. Speaker labels come from your text: keep them in the pasted transcript, or through the API set split_speakers, and each labeled line becomes a labeled paragraph. For detection from the voices, transcribe the file instead.
Can I align a caption file and export it in another format?
Yes. Paste the SRT in with its video, and once it is aligned you can export VTT, TTML, a burn-in MP4 or a DOCX from it, with the subtitle line and timing options applied.
Pages around this one
The feature this page belongs to, and its neighbors.
Everything else Sonix does
with the same recording
Get started
Align a transcript you already have
Upload a recording with its script and click through it by the words. 30 free minutes, no credit card.
4.98
10xfaster than real time
30free minutes to start
