RE: Support for advanced caption features (inc rollup)

we are regularly presenting pre-recorded television content on the internet - content that was originally captioned live with rollup captions.

this is a real use case.

-glenn

Glenn Goldstein | Vice President, Media Technology Strategy | VIACOM
glenng@mtvi.com | 212-846-3210 | 1515 Broadway, New York NY 10036


________________________________
From: Anna Cavender [mailto:annacc@google.com]
Sent: Tuesday, December 04, 2012 6:19 PM
To: Ian Hickson
Cc: Silvia Pfeiffer; public-texttracks@w3.org; Loretta Guarino Reid
Subject: Re: Support for advanced caption features (inc rollup)



On Tue, Dec 4, 2012 at 2:42 PM, Ian Hickson <ian@hixie.ch<mailto:ian@hixie.ch>> wrote:
On Tue, 4 Dec 2012, Anna Cavender wrote:
> >
> > The requirements are similar (though not even remotely identical, for
> > example there's no way to drag and drop cues on a TV), but the
> > constraints are vastly different (for example, live TV historically
> > couldn't have a two-second delay loop and users care more about
> > syncing captions to the picture than readable captions, but on the
> > Internet a two-second delay is a non-issue even for live streams and
> > so we can get readable captions and still get the sync right).
>
> A two-second delay will certainly be an issue for real-time interactive
> communication.  Real-time captioning in these scenarios should be
> displayed as soon as it is available and any delay will negatively
> affect access to the conversation.  The idea of rendering regions seems
> crucial here.

Given that neither <video> nor WebVTT support real-time captioning, this
is kind of moot in a discussion of WebVTT and <video>.

If we wanted to support real-time captions of low-latency video we'd have
to do far more than just put roll-up captions in WebVTT.

(Note that instant messaging isn't "roll-up", by the way. It's line-at-a-
time, and as a near non-stop user of this medium, I'm pretty confident in
saying that that is fine.)

To clarify, I was proposing live captioning for video conferencing, not instant messaging and not necessarily roll-up, as a use case for the Region.  There exist proposals to use WebVTT in DASH [1] and HLS [2], so this isn't too far fetched.  A Region would enable appending of cues as a captioner types.

Indeed, much more would be required to support real-time captions in WebVTT, but it is a potential use case for the Region.

[1] http://concolato.wp.mines-telecom.fr/category/general/mpeg/dash/
[2] http://tools.ietf.org/html/draft-pantos-http-live-streaming-09


--
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'