W3C home > Mailing lists > Public > public-html-a11y@w3.org > April 2011

Re: text track associations

From: Mark Watson <watsonm@netflix.com>
Date: Fri, 8 Apr 2011 09:41:18 -0700
To: Sean Hayes <Sean.Hayes@microsoft.com>
CC: Silvia Pfeiffer <silviapfeiffer1@gmail.com>, Eric Carlson <eric.carlson@apple.com>, Ian Hickson <ian@hixie.ch>, "public-html-a11y@w3.org" <public-html-a11y@w3.org>
Message-ID: <AE8313F6-D319-4F11-A51E-CE9B199DCEE1@netflix.com>
I think that a starting point for the design should be an assumption that all the tracks of a multitrack presentation have the same start time. If they have different durations (and they always will, at least a little bit) then we pad them to the longest.

This situation is easy to deal with. As soon as you break from it, even a little bit, a lot of new questions arise.

I'm not an expert in director's commentaries, but it seems unlikely that some content would break with the model above in just a small way (i.e. just at the start or just at the end). If you break with the common timeline model it is probably in some major way such as extended audio descriptions which require numerous pauses in the video. I wonder if those cases should be dealt with as an entirely separate piece of content ? Ultimately, someone has to author the new version of the video, place the pauses correctly etc. and it seems better if this is done in the media formats themselves rather than asking the browser and page to handle it. One could imagine solutions where we can create "new tracks from old" by application of edit lists or some such so that providing the extended audio descriptions version does not require duplicate storage of the main audio and video. Maybe such things already exists ? But the new video would appear as a separate track to HTML whose duration did match with the extended audio descriptions track.

...Mark 



On Apr 8, 2011, at 2:16 AM, Sean Hayes wrote:

> In a multitrack API, since caption tracks are essentially the audio for Deaf and HoH use, we will need some way for these to follow the selected audio asset.
> 
> If I have a presentation, with the option of switching to different a different audio resource (e.g director commentary which is longer, starts before and ends after the main movie). The captions for it need to be timed to the commentary audio, and so if it's in a markup element, live in that element. If the user has captions turned on and wants the commentary audio, then the captions for the commentary soundtrack will need to replace the captions for the original audio track, and render into the rectangle allotted for the main video.
> 
> I'm not seeing how I would create this. 
> 
> 
Received on Friday, 8 April 2011 16:41:46 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:05:19 UTC