We think that corrections and clarifications are needed in part 12 as well. We perhaps should clarify that width and height apply only to visual tracks, and volume only to audio tracks. We need to clarify that visual means ‘visually presented’, i.e. not necessarily (but commonly) video.
We think we should define/reserve 0,0 to mean “provided by the external context”, and the values are irrelevant for non-visual tracks. Perhaps tidying up the semantics for audio as well might be advisable.
We are willing to look also at the text for a subt track reference, which means that the subtitles are associated with the identified track (or the group, if it is a member of the group). Otherwise, perhaps the subtitles are associated with the composition of all visual tracks?
This leaves the ugly question of where the text track has a known aspect ratio but unknown absolute size, in particular where alternatives are authored for different aspect ratios.
The alternative is to put something sensible that has the right aspect ratio, since if you are allowed to override, at least this gets us aspect ratio information and also some default if the track is played ‘by itself’.
But it does seem we need to allow 0,0 (‘don’t know’). Perhaps we need to use tkhd flags to say something (e.g. “provide aspect ratio only, not size’).
We could also consider a box in the sample entry, like pasp, to document sizing, aspect ratio, etc.
We will try to make corrigenda to parts 12 and 30 at this meeting.
Dostları ilə paylaş: |
