Subtitles
Create, review and publish multilingual subtitles
Generate an editable transcript from hosted video, or upload your own WebVTT or SRT file for each language. DynTube converts SRT to WebVTT for delivery, keeps every language as a separate track, and serves the published result in the same player.
One track per language, in the player you already use
The problem
Why most video libraries have no captions
Captioning an hour of video by hand can take several hours, so growing libraries quickly accumulate a backlog. Clearing it makes video usable for deaf and hard-of-hearing viewers, useful with sound off, and searchable by what was actually said.
The consequence compounds. Every month of new video widens the gap, so the backlog grows faster than any plan to clear it, and eventually "we should caption everything" becomes a thing people stop saying.
Machine transcription supplies the timed first draft while a reviewer supplies the judgement. That changes a back catalogue from a blank-page transcription task into a focused review of names, specialist terms, numbers and timing.
The mechanism
Choose the subtitle workflow that fits your content
Described against the current transcript contracts: locale listing, and transcription and translation request workflows.
- Upload prepared subtitle filesWebVTT or SRT · per language
- If captions are already prepared, upload a WebVTT or SRT file and assign its language, label, direction and default status. SRT timestamps are converted to WebVTT and the published track is delivered through the player.
- Transcription requestPer video
- You request a transcript for a video that is already hosted. The audio is processed by automatic speech recognition and returned as a timed transcript attached to that video.
- Translation requestPer locale
- An existing transcript can be translated into further languages. Each output language is a separate pass over the same source transcript rather than a re-listen to the audio.
- Locales listed per videoTrack management
- The transcript locales that exist for a video can be listed, so you can see what a title already has and what is missing before requesting more.
- Editable before it is publishedReview
- The generated transcript is text you can correct. This is the step that turns a draft into a caption track — names, jargon, product terms and numbers are exactly where recognition is weakest.
- Served by the same playerPlayback
- Published tracks are delivered by the player and embed you already use. There is nothing separate to host and no second integration.
The workflow
Generate, review, publish
Use the generated transcript as a draft, correct the source language, then translate and spot-check playback.
- 01
Request a transcript
- 02
Read it and fix it
- 03
Translate from the corrected transcript
- 04
Publish and spot-check on playback
Why bother
Make hosted video easier to watch and discover
Captions widen access, support silent viewing and turn spoken material into reviewable text.
Viewers who need them
Deaf and hard-of-hearing viewers need an accurate text alternative to the spoken content, reviewed with their use of the track in mind.
Viewers watching without sound
Open-plan offices, public transport, shared rooms, muted autoplay. A large share of viewing happens where audio is not an option.
Second-language audiences
Written words are easier to follow than spoken ones in a language you are still learning — often more useful than a full translation.
Text you can search
A transcript makes the content of a video findable by what was said, both for your own team and on the page around the player.
Questions
Questions about automatic subtitles
Plan the review, locale and accessibility responsibilities before processing a library.
How accurate is it?
Recognition quality depends on audio conditions, overlapping speakers, accents, pace and specialist vocabulary. Run a representative video, review every line, and use the correction time to estimate the rest of your library.
How many languages are supported?
Tell us the source and target locales you need and get current availability confirmed in writing before planning the batch. Each translated language is processed and reviewed as its own track.
Does this make my videos ADA or accessibility compliant?
Captions are one part of accessible video. Your content, obligations and jurisdiction determine the full requirement, so have the generated transcript reviewed and have your accessibility owner approve the published experience.
Will the timing line up?
Usually closely, sometimes not — particularly around pauses, music and overlapping speech. Check it on anything instructional, and correct it where a caption appearing early spoils a demonstration or a reveal.
Can I use captions I already have?
Yes. Upload a WebVTT or SRT file for each language and set its label, text direction and default status. DynTube converts SRT timestamps to WebVTT for delivery, so existing professional captions can be used without automatic generation.
What does it cost?
Subtitling is metered on the minutes of audio processed, per language, as an add-on to a hosting plan. The subtitle pricing page explains the model and routes to a request for current terms.
Run it on one of your own videos
Pick a title with your usual audio conditions and your usual vocabulary, generate a transcript, and read it. Ten minutes tells you how much correction your library will need.