Description
1.2.1 Audio-only and Video-only (Prerecorded), 1.2.2 Captions (Prerecorded), 1.2.3 Audio Description or Media Alternative (Prerecorded), and the definition for "synchronized media" all include the clause that A/V used as a media alternative for text be "clearly labeled as such". Ironically, what actually constitutes a "clear label" is left up to the reader to infer - none of the understanding documents make any mention of this or provide examples.
There is one failure technique, Failure of Success Criterion 1.2.2 and 1.2.8 due to not labeling a synchronized media alternative to text as an alternative, linked from 1.2.2 Captions (Prerecorded)'s understanding, that initially seems to provide some information, but it's just a tautology - the failure is not having a clear label, and the test process is "check that it's clearly labeled". Not...very informative or useful.
While I doubt this can be normatively nailed down fully, it would be nice to at least provide some indication of what "clearly labeled" actually means, in practice.