TLocal

TLocal is a web application that automates the scoping and translation of on-screen text in video for international distribution. It is built for post-production teams shipping localized cuts of films, episodics, and trailers, where the work of cataloging every title card, lower third, sign, and graphic is currently a manual frame-by-frame slog before a single string gets translated.

 

Traditional localization scoping forces a producer to scrub a cut end to end, screenshotting and timecoding every text element by hand, then wrangling that list through translation, then handing the translated strings back to a graphics artist with rough sketches of where each one belongs. A 30-minute episode can take a full day per locale just to assemble the brief.

 

TLocal replaces that with a pipeline. Upload a cut and shot detection plus text detection identify every on-screen text element and group them by shot. A frame is extracted for each, and Claude transcribes the actual text in context. Reviewers get a timeline and a ticket table where they can edit, approve, or skip each item, then translation runs against the approved set. For each translated item, Gemini regenerates the frame with the new text burned in as a reference.

 

The final deliverable exports as xlsx, csv, or json, scoped to a single locale or all of them. A same-day brief instead of a same-week one — every text element identified, translated, and visually previewed before sending the shot back through the pipeline.