Guide
Subtitles done right: SRT, VTT, speaker labels and translated captions
Two problems, not one
Captioning is a transcription-and-timing problem: turn spoken audio into text, broken into cues that appear and disappear at the right moments. Translating subtitles adds a second hard problem on top — each translated cue has to carry the meaning *and* still fit the time the original line occupied.
Doing both well is why generic auto-captions and machine-translated subtitle dumps so often read badly.
Soft vs burned-in
Soft subtitles are a separate file (SRT, VTT) that a player overlays on the video. They can be turned on and off, restyled, and swapped between languages without touching the video. Burned-in subtitles are baked into the pixels permanently.
For almost every professional workflow you want soft subtitles: no video re-encoding, no quality loss, and one video can carry captions in many languages. Traxlate delivers files, not burned-in pixels — you mount them on your player or upload them to YouTube.
File formats
You'll usually want one of:
- SRT — the universal subtitle format, supported nearly everywhere.
- VTT — the web-native format, used by HTML5 video.
- TXT — a clean transcript with no timing, for repurposing as an article or notes.
- Bilingual SRT — source and target language stacked in each cue, useful for review and language learning.
Speaker labels and cue timing
Good subtitles identify who's speaking when it matters, and their cues are timed precisely to the audio. When you translate, those speaker labels and cue timings have to survive — a translated subtitle file where the cues have drifted off the audio is unusable.
Traxlate keeps the original cue timing and speaker structure and reflows the translated text into the same cues, so a translated SRT lines up with the video exactly like the source did.
Translating into many languages
Caption once, then translate into as many target languages as you need. Each translated subtitle file is independent and ready to mount — there's no re-transcription per language, because the timing and structure are shared.
Already have an SRT?
You don't always need to start from the audio. If you already have a subtitle file you can re-time it (fix an offset or frame-rate mismatch), convert between formats, or translate it into more languages — without re-running speech recognition, which is the expensive part.
Free with a dub
If you've already dubbed a video on Traxlate, the matching subtitle files — in every language you dubbed into — are included at no extra cost. The transcription and translation work was already done for the dub.