27 June

If you watch a lot of films and TV shows that weren’t created in your native language, you may have experienced the frustration of reading sub-titles. Or if it’s dubbed, the usually horrible mismatch between the dubbing actor’s voice and the original, making it sound fake. AI is already revolutionising this problem. AI can firstly use the same voice as the original actor, and then also capture the actual physical experience of the actor in the scene, so a dub doesn’t sound like another actor reading lines. Dubbing sounds fake because the dubbing actor is in a sound booth somewhere, not in the physical reality of a scene, which shapes how speech is produced. AI can create the voice as if it‘s being produced within that physical scene. And it goes even further. It can actually manipulate the movement of the mouth in the scene so that it looks like it’s speaking your language. For those worried about loss of actor’s rights, many actors embrace the idea because they’re paid extra for the additional translation. And it’s their voice people in other countries hear, not some dubbing actor’s. This allows the full artistic effect of a production to not be lost in translation by people having to read or listen to fake-sounding dubbed voices. It creates a universal experience that crosses cultures.