Subtitle translation can mean two different jobs. One tool listens to speech and creates captions. Another reads captions that are already visible and translates them. TranslateWindow does the second job.
Choose the subtitle area once, place the result where it does not cover the video, and continue watching. The live region refreshes when the visible caption changes.
Pin the subtitle line, not the whole video
Draw a narrow region around the place where captions appear. A smaller capture reduces titles, controls, chat messages, and other text that could compete with the subtitle line.
Put the translation window above, below, or beside the player. TranslateWindow leaves the original caption visible so you can compare names, numbers, and unfamiliar terms.
Where visible subtitle translation helps
| Content | What must be visible | What to capture |
|---|---|---|
| Online course | Course captions or burned-in subtitles | The subtitle line |
| Video | Player captions or subtitles inside the picture | The stable caption area |
| Meeting | On-screen captions supplied by the meeting app | The caption panel |
| Presentation | Words on the shared slide or demo | The changing text region |
It translates captions; it does not create them
TranslateWindow does not listen to system audio, microphone input, or a speaker's voice. If there is no visible text, there is nothing for its screen OCR to translate.
Turn on captions in the video or meeting service first. If that service cannot produce captions, use a speech transcription or audio translation tool before TranslateWindow.
Improve recognition before changing models
- Increase the player size when subtitle text is very small.
- Choose a caption style with clear contrast when the player offers one.
- Exclude playback controls, chat panels, and speaker labels from the capture where possible.
- Pause and use a screenshot when a dense caption disappears too quickly.
Keep recurring terms consistent
A technical course or product meeting may repeat names that generic translation handles differently from one line to the next. A translation scene can carry preferred terms and instructions into the result.
For ordinary viewing, the default scene is usually enough. Add rules only when a recurring term is actually getting in the way.
Choose local or cloud translation
With a downloaded local model, recognized caption text can stay on the Mac. If you choose a cloud service or your own API key, the required text is sent to that provider.
TranslateWindow does not include the screenshot, OCR text, source text, or translation in website or product analytics.
Questions about on-screen subtitle translation
Does it work with burned-in subtitles? Yes, if the text is visible and clear enough for OCR.
Can it translate a meeting without captions? No. The meeting app must show captions or other visible text first.
Does it replace the original subtitles? No. The translation stays in a separate compact window beside the source.
Try it on one visible subtitle line
Turn on captions, pin the line, and place the result beside the video. TranslateWindow follows the text without listening to the audio.
Sources and verification
Product facts were checked against these first-party sources. Links open the current official documentation.