Get a properly timed caption file out of a recording, ready to upload or drop onto a timeline.
SayToType times every line against the audio, lets you correct the ones that came out wrong, and exports a standard .srt or .vtt file — with or without speaker names and timestamps.
Free to start, on Windows and macOS.
Four steps, and only one of them needs your attention.
MP4, MOV, WEBM and more, alongside plain audio files. A video is read for its sound, so a screen capture works exactly like camera footage. Long recordings are split at natural pauses and processed part by part, with progress you can watch and cancel.
Play the recording back with the current line highlighted as it goes. Where a caption starts a beat early or the recognition mangled a name, fix that line by hand before you export.
Export with or without speaker names, with or without timestamps. A two-person interview can ship with “Anna:” in front of every line, or as clean captions with no labels at all.
Translate the finished transcript on a tab beside the original. Timestamps and speaker names carry over, so a second subtitle track is a couple of clicks rather than a second pass over the video.
Two formats, one rule of thumb.
You are uploading to a video platform or handing the file to an editor. It is the most widely accepted caption format there is, and the safe default when nobody has told you otherwise.
The captions are going onto a web page — VTT is the format an HTML5 video element expects for its text track.
Both are plain text. You can open either in any editor, read the timings, and see exactly what will appear on screen — which also means you are never locked out of your own captions.
A file that behaves like a caption file should.
Plain, well-formed caption files. Nothing proprietary, nothing that needs SayToType installed to open on the other end.
Play the video inside the app with the captions running on it, full screen if you want, with the transcript and the translation beside it. What you see there is what the exported file will do.
Captions are only as good as their timing. Every line is editable against playback, with the spoken line highlighted as the audio runs.
Keep the labels for an interview or a panel, drop them for a single-voice narration. The same transcript exports either way.
Translate the transcript and the timings come with it, so the second-language track lines up with the first without redoing the work.
Recognition covers 47 languages, worked out from the recording itself. Subtitling something you did not record yourself is the normal case, not the exception.
Wanted the words rather than the captions? Get a plain-text transcript with speaker labels.
Four jobs that all end in the same file.
A timed caption file that drops onto the timeline, instead of typing captions against a scrub bar for an afternoon.
Most people watch with the sound off at least some of the time. Captions are the difference between being watched and being scrolled past.
One recording, two subtitle tracks: the original and a translated one that keeps the same timings.
Training clips and recorded walkthroughs become searchable and accessible without a separate captioning step.
Both are plain-text caption files that pair lines of dialogue with timings. SRT is the older and more widely accepted of the two and is the safe default for video platforms and editing software. VTT is the format used for the text track of an HTML5 video on a web page. SayToType exports either.
Yes. Play the recording back with the spoken line highlighted and correct any line by hand. What you export is what you last corrected.
Yes, and you can leave them out. Export runs with or without speaker names and with or without timestamps, from the same transcript.
Yes. Translate the finished transcript on a tab beside the original and export that. Timestamps and speaker names carry over, so the translated track keeps the timing of the original.
MP4, MOV, WEBM and more, plus audio files such as MP3, WAV, M4A, AAC, FLAC, OGG and OPUS. A video is read for its sound, so a screen recording works the same as camera footage.
Yes, for file transcription. The recognition runs on a service — SayToType Cloud by default, or OpenAI, Mistral AI or Deepgram if you supply your own key — so the file is transcribed online.
SayToType runs on Windows and macOS and exports SRT and VTT from any recording you give it.
Free to start — every account gets 50 minutes of cloud transcription a month at no cost, and no card is needed to sign up.
