Automatic Subtitles and Translation in Fermata Auto
For years, the story of subtitles in a media player was simple: you either had a subtitle file or you did not. Fermata 2.0 changed that. Release 2.0.0 (24 January 2026) introduced automatic, real-time subtitle generation, and release 2.0.1 (6 June 2026) added an Opus-MT based subtitle translator. Together they let you watch a video or listen to an audio stream in one language and read it in another.
What the feature does
In the developer's words, automatic subtitle generation allows watching video and listening to audio streams with subtitles. The speech in the stream is recognised and turned into text while the media plays, so you do not need a pre-made subtitle file. The translator then converts that text into the language you prefer.
Why it is useful
- Foreign-language videos with no subtitle file.
- Live streams and IPTV channels where subtitles do not exist.
- Learning a language by reading along.
- Accessibility for people who are deaf or hard of hearing.
- Noisy places where you cannot turn the volume up.
What is Opus-MT?
Opus-MT is a family of open-source neural machine-translation models trained on the OPUS corpus of parallel texts. Using open models means translation can run in a self-contained way rather than depending on a paid cloud service. It is one reason the 2.0 APK is larger than earlier versions.
How to use it
- Update to Fermata 2.0.1 or newer from the download page.
- Open a video or audio stream.
- Open the subtitle options in the player and enable the automatic subtitle feature.
- Choose a translation language if you want translated text.
Menu names may change between releases, so look for options mentioning subtitles, generation or translation. If you cannot see subtitle controls at all, remember that video subtitle and audio-track selection needs the VLC engine; see our engine guide.
What to expect from accuracy
Automatic speech recognition is impressive but not perfect. Expect good results with clear speech, a single speaker and common languages. Expect mistakes with:
- background music and noise,
- overlapping speakers,
- strong accents, slang and names,
- technical vocabulary.
Translation adds another layer, since a mistaken transcript becomes a mistaken translation. Use it as a helpful guide, not as a legal or medical record.
Performance, battery and storage
- Processing power: real-time recognition is demanding. Newer phones cope better. If playback stutters, close other apps.
- Battery: expect more usage than normal playback; keep the phone charged when using it for long sessions.
- Storage: models take space. The size of the 2.0.x APKs (roughly 115 to 153 MB for arm64) reflects that.
- Privacy: because the models are open and can run on the device, your audio does not have to be sent to a third-party service for the basic feature.
Using it in the car
Subtitles are especially convenient when watching video while the car is parked and the cabin is noisy. As always, never watch while driving. See our video guide for the setup.
Tips for better results
- Use a recent, capable phone.
- Prefer clear audio sources.
- Switch to VLC if subtitle controls are missing.
- Keep Fermata updated; the feature is new and improving.
Frequently asked questions
Can Fermata generate subtitles automatically?
Yes. Since version 2.0.0 Fermata can generate subtitles in real time, and 2.0.1 added an Opus-MT based translator.
Do I need internet for automatic subtitles?
The feature is designed around open models that can run on the device, but you still need internet to stream online video. Check the release notes for the version you use.