How to change a name in both the speech and the on-screen text
Tell Vizard Agent the name has to change in the spoken audio and in the subtitles, and that everything else stays. It clones each speaker from their own clean sample, fits each new word into the interval the old one occupied at the same level, and rebuilds the subtitle at the same pixel position.
What do you type into Vizard Agent?
Name both channels explicitly, and give Vizard Agent whatever timestamps you already know. A brief that just says "change this name" reliably gets one channel fixed and the other left behind to contradict it, because nothing in the request said there were two. Approximate timestamps are enough, and list any inflected forms, since a name in a possessive or a question takes a different shape.
Change this name to that one in both the spoken audio and the on-screen subtitles, around 0:00–0:03 and again at 0:10. Keep the voices, timing and everything else as they are.
What does Vizard Agent actually do?
It builds one list and works both channels from it. A word-level transcript gives the precise start and end of every instance, and the frames at those moments are pulled at full resolution so the subtitle's look and position can be measured rather than guessed. In the session behind this article it:
- Extracted a word-timed transcript, listed the exact start and end of every word to be replaced, and pulled full-resolution frames at each moment to compare the subtitles.
- Prepared a clean voice sample for each of the two speakers, cloned them separately, and generated each replacement in the right voice, including an inflected form and a question-intonation version of one name.
- Measured each new word's duration and level, trimmed the silence at its edges, and matched it to the original name's level.
- Measured the subtitles' real pixel positions and ran a small visual trial to confirm the new text blended with the background before committing.
- Ran a second round: the first clones came from short samples and read stiffly, so longer, cleaner samples were used and the words regenerated, re-timed and re-levelled.
When does this not work well?
When the replacement is a noticeably different length. A longer name still has to fit the gap the old one left, so past a small difference either the line runs over the next words or the read is compressed, and Vizard Agent will say which it is trading away. A name said with strong emotion is hard to match, and a name burnt into a moving graphic becomes a tracked replacement.
Common questions
Why does the old name flash back? The subtitle's boundary frames. Vizard Agent adjusts each cue's in and out times.
What if two people say the name? Vizard Agent clones each speaker separately so each one says it themselves.
How many passes should I expect? Two. Vizard Agent uses the first for timing and the second to make it sound natural.