Video (or audio, too) format that stores multiple languages in the same file, but only the parts that are different?

Question

DVDs or similar often offer different dubs of the same movie, TV show, etc. on one disk. Storing a whole different video track (and audio track, too) for a movie seems a bit wasteful, given that the parts where nobody actually speaks (and, for that matter, when no on-screen text needs to appear in another language) don't need to be duplicated.

So, instead, why not provide the whole movie in some "canonical" language (maybe English), and then if other languages are available only store the parts of the movie that actually need to be translated, seamlessly loading them as needed? Is there a video or audio format that does something like this (not necessarily on DVD)?

Answer 1

You don't need separate video tracks. All the differences are in the audio, which is a minuscule part of the overall data on a BVD. It's a simple matter for a player to select one of multiple audio tracks.

Answer 2

Very interesting thought. In some films on screen text (e.g. signage) is changed. One could imagine a sort of patch being played instead of the "original" image. I imagine one could code something like that together in html5 for digital delivery. I don't think, however, that it's possible on the existing disc platforms. In any case the use cases are few.

Answer 3

On the post-production level obviously you have this. It happens that you edit separate language versions at the same time, using a multitrack audio file that contains a basic mix and then the parts that are different in each language. On the distribution level, as somebody mentioned (for home use), there is no real need for that, the audio part of the movie is so small in comparison to the video part, I guess the logic that would be necessary to achieve the "overlaying" of audio would be more complex than the advantage (by saving a little space). In the cinema world dolby atmos is gaining some ground and it seems it is slowly entering the home cinema world as well. With dolby atmos, this idea is entirely possible and probably exists already.

As far as I understand dolby atmos, it contains the raw tracks with mixing information (volume, panning) as meta-data which can be applied to different setup, room sizes or language versions (as in your question)