Every story deserves a bridge.
A film carries its culture in every line. BridgeSub sets two languages on one screen, so viewers follow the story in their own words while staying close to the original, and films from elsewhere find new audiences.
Why it matters
Audiences are already reading.
Cinemas are still winning their audience back, and viewers at home watch more films from other languages, often with subtitles on.
16% below
Global box office in 2025, against the 2017–2019 average, though 6% above 2024.
Over a third
of all Netflix viewing in the second half of 2025 was of non-English titles.
80%
of viewers in the UK, US, France, Germany, and Spain use subtitles at least sometimes, though 81% report no hearing difficulty.
A second language widens the room.
A film festival screens films from many languages to one hall. A neighbourhood cinema has regulars who are still learning the local language. A family watches together while reading different languages. Two subtitle languages on one screen serve all of them at once.
One track is all a player has to show.
Subtitles usually arrive as one file per language, and a player shows the one you pick. BridgeSub writes both languages into a single subtitle track, timed together, so a player that shows one track shows both.
What BridgeSub does
Stacking two files is not merging them.
Each scene below plays the same clip twice. Switch between Before and With BridgeSub to compare; the strip under the clip draws every subtitle on the clip's timeline.
Merge · timing
Two files, two clocks.
Subtitle files for the same film are often timed against different releases, so one language starts late and slips further as the film runs. BridgeSub measures the offset and the drift between the two tracks and corrects them before pairing lines.
Both files shown as delivered. The Chinese file starts about a second late and keeps slipping.Merged by BridgeSub. The offset is corrected and each line pairs with its translation.
Before
With BridgeSub
Drag along the timeline to move through the clip.
Merge · meaning
Same sentence, different breaks.
Every language has its own rhythm. Translators break sentences in different places, and one language may say in a few characters what another needs a full line for. Paired by timing alone, half a sentence flashes against the wrong line. BridgeSub groups the cues that carry one sentence in each language and checks each pairing by meaning.
Paired by timing only. Three cues in language 1 against four in language 2; language 2 changes mid-sentence.Grouped by BridgeSub. Each screen carries one complete thought in both languages.
Before
With BridgeSub
Drag along the timeline to move through the clip.
Merge · on screen
Built to be read in the dark.
Each language keeps at most two lines, and a marker sets the second language apart, so every viewer's eyes go back to the same place. An optional line width keeps lines short, and a sound label stays in a cue of its own instead of joining a line of dialogue.
Merged without layout rules. Long lines wrap unpredictably and the sound label rides along with the dialogue.With BridgeSub layout. Two lines per language, a dot before each Chinese line, lines kept short, and the sound label on its own.
Before
With BridgeSub
Drag along the timeline to move through the clip.
Alignment
Subtitles that land on the word.
When a track runs early or late, BridgeSub listens to the film's audio with on-device speech recognition and retimes each line to the speech. For another language, it transfers the timing from a track you have already accepted as aligned. The audio never leaves your Mac.
The English file as delivered, more than a second behind the speaker.Aligned by BridgeSub to the speech in the clip.
Before
With BridgeSub
Drag along the timeline to move through the clip.
Translation
The same name, from the first scene to the last.
Translating a few lines at a time, a model can spell a name one way in one scene and another way later, or translate a job title word by word. BridgeSub first reads the whole subtitle file and builds a glossary of names, terms, and tone that every batch must follow. You choose the provider, Ollama or any OpenAI-compatible service, and confirm before any text is sent.
Translated batch by batch. The name and the mission are spelled one way here and differently elsewhere in the film, and the job title is translated word for word.Translated with the glossary. Names, the mission, and the role follow the glossary in every batch.
Before
With BridgeSub
Drag along the timeline to move through the clip.
On the roadmap
Built next with cinemas and festivals.
BridgeSub is in development, and none of the work below is in the app yet. If you run a cinema, a festival, or a screening series, tell us what your audience needs and we will build with you.
A second language, chosen by the audience
On a cinema's website, viewers ask for a second subtitle language for a screening, and the cinema prepares the bilingual file with BridgeSub.
XML subtitle formats
Reading and writing the XML subtitle files used in cinema and broadcast delivery, such as DCP subtitles and TTML.
SMPTE timecode alignment
Aligning subtitles to SMPTE timecode, so timing matches the editing and mastering timeline frame for frame.
- Gower Street Analytics, global box office estimate for 2025.
- Netflix, What We Watched: the second half of 2025.
- XR Extreme Reach, Global Accessibility Report, survey of 3,000 consumers in five countries, May 2026, as reported by Advanced Television.
- Clips: NASA's Goddard Space Flight Center. NASA does not endorse BridgeSub. The subtitles on this page are prepared to show each feature; they are not unedited app output.
- Netflix is a trademark of Netflix, Inc. Other names are trademarks of their owners.