kadirb4rut/video-dubbing-translator — reverse-engineered prompt

Build me a local video dubbing app that runs on my computer and opens in the browser. I want to upload a video, choose the original language or auto detect it, pick a target language, then click one button to create a dubbed version.

The app should pull out the speech, keep or rebuild the background audio, transcribe what people say, translate it, generate a new voice track, and put everything back into a final video I can play in the page. It should show clear progress logs while it works and let me open the output folder when it’s done.

Please include setup scripts for Windows, Mac, and Linux, plus a model download or check button so a normal user doesn’t have to run lots of commands. If the computer has an NVIDIA CUDA setup, add an optional lip sync checkbox using LatentSync, but keep the normal dubbing workflow working without it.

Use Python for the local app and look up current docs online if needed.