Feature Request?: Re-time subtitles with speech-to-text

[wave-dash]

I have a subtitle file where every line is mistimed by as much as +/- 1 second. Is there a way I can re-time each line, on a line-by-line basis, using Whisper? I don't see a way to use Whisper's time codes with existing subtitles.

I really don't want to just use Whisper's output. Setting aside the issue of plain old accuracy, my current subtitles have stuff that wouldn't get transcribed by speech-to-text that I want to keep (speaker tags, sound effects, occasional furigana).