DubWeaver: Dub Films Offline

Dub any film into any language, in your own voice, on your own device.

DubWeaver is a dubbing studio that runs entirely on your phone or tablet. The AI that removes the original voices runs HERE, not on a server. Nothing is metered by the minute, and it all works with no signal at all.

TAKE THE ORIGINAL VOICES OUT
The separation model runs on the device and lifts the original dialogue out of the mix, leaving the music and effects behind — room for your dub without losing the film's own soundtrack. If a film already ships a music-and-effects track, DubWeaver finds it and uses that instead.

FIND EVERY LINE, THEN WALK THEM
DubWeaver reads a track once and marks where each line starts and ends — the step Pro Tools calls Strip Silence. The breath before a line is kept and the tail left to decay, and two arrows walk you from mark to mark: skip, record, skip.

RECORD YOUR DUB AGAINST THE PICTURE
Arm a take and perform while the film plays. Count-in, monitoring through your headphones, and the take lands aligned to the frame. As many takes as you like, on as many voices as you like.

MEASURE YOUR OWN DEVICE
Cheap hardware lies about its own audio delay, and dubbing apps inherit the lie. The Calibrador plays a click train through your headphones, hears it come back through the microphone, and places every take by what it measured. Fifteen seconds, once.

SHAPE THE VOICE WHILE YOU HEAR IT
High-pass, compressor, de-esser, three-band EQ and room reverb — set them and hear them, rather than finding out at export.

SUBTITLES WRITTEN FROM WHAT IS SAID
Point DubWeaver at a track and the words come back with the times they are spoken at, in any of its nine languages — on the device, so nothing is uploaded and nothing is charged by the minute. Import an .srt as well, or write captions yourself.

CAPTIONS THAT MOVE
Seven styles: typing letter by letter with a cursor, a word at a time, karaoke, a word lighting up as it is spoken, a pop, a fade. One press styles every caption on a track, and what the player shows is what the exported file carries. Thirteen bundled faces travel with the project. Burn them in, export an .srt, or both.

1,500 SOUND EFFECTS
Whooshes, risers, applause, door slams, rain and more, in 52 categories. Search in your own language, tap to hear, star your favourites and add one at the playhead. All are public domain (CC0), free to use in the videos you publish.

FIFTEEN LOOKS FOR THE PICTURE
Picked by looking rather than from a menu: each is a small square of your own frame already wearing it. Nine are free. Six more come with the subscription, and four of those are drawn into the picture rather than graded onto it: a 1970s CRT television with its scanlines, film grain that moves the way stock does, a light leak, and worn VHS tape. Each look sits on its own track, so one film can carry several.

MIX IT PROPERLY
Level, pan and fades per clip. Automatic ducking so the music steps out from under the dialogue. Loudness normalised to streaming, podcast or broadcast targets. Deliver in stereo or discrete 5.1, at the resolution of your source, rendered on the device.

NOTHING LEAVES YOUR DEVICE
Video, audio and projects are stored on this device and nowhere else. No account to create, no cloud to sign in to. The app downloads its four AI models once each and each sound effect the first time you use it, every file against a known checksum, and makes no other request of its own. The only other connection is Google Play confirming your subscription — neither carries your film.

NINE LANGUAGES
English, Spanish, Portuguese, French, Italian, German, Japanese, Simplified Chinese and Russian. The app follows your device and lets you change it whenever you want.

SUBSCRIPTION
One month free, then one subscription: no watermarks, no project limit, no per-minute charges. Without it the basic tier keeps working — five projects, delivery up to 720p, and every project you already made stays open.

Recent changes:
The AI tools in plain sight: an AI tab with a card for each (remove the voices, write subtitles, detect lines, the voice studio) and an AI button in the top bar. Transitions on any cut, with the sound crossfading. Clip speed from 0.25× to 4× with the pitch kept, and freeze frame. Copy, paste and duplicate a clip; reset and keyframe on every inspector row. Every tool has its own icon. AI models download only when a tool first needs them, and a download picks up where it left off.