[Sora配音] The Jukebox karaoke machine project has been restarted, with Sora supporting audio generation, background music, and voiceover.

Sora has gone viral. Although it has not yet been officially released, many industry insiders have already begun discussing how to create videos, and dubbing is naturally a key technology in this area, so it has also become a hot topic. As the topic grows hotter, more moderators have begun to join in. Among them, an administrator nicknamed VB hinted that JukeBOX, although very old, has always existed.



However, recently we have noticed that OpenAI is making new moves with Jukebox and has resumed research and development on it. Most people believe it follows closely after Sora and is presumably intended to integrate with Sora for audio output. This seems to confirm the VB moderator’s verification.

Music example





The moderator selected the first piece of music, and you can see that it has corresponding lyrics. From this, it is not hard to infer that it is the same as the music generation tools released by other major companies: all are text-to-music.




If Juke is meant to work with Sora, then we believe it should be able to provide dubbing for it, including background music, vocals, and so on.

Perhaps you can use text-to-video prompts to simultaneously generate text-to-music.