W3C home > Mailing lists > Public > public-webplatform@w3.org > December 2013

Re: VTT authoring guide on Web Platform Docs?

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
Date: Fri, 13 Dec 2013 12:51:08 +1100
Message-ID: <CAHp8n2nP8Xr0tRF8E50iOrsg5-KqWZTMy9cjp5HREpU3+RL_ZQ@mail.gmail.com>
To: Carlos Araya <carlos.araya@gmail.com>
Cc: David Singer <singer@apple.com>, "public-webplatform@w3.org" <public-webplatform@w3.org>, "public-texttracks@w3.org" <public-texttracks@w3.org>, Doug Schepers <schepers@w3.org>
Nice start! Some random feedback.

WebVTT stands for "Web Video Text Tracks" and not "Web Video Time
Tracks", although I have to wonder if that might be better naming. ;-)

Google Chrome only has the region feature behind a flag - everything
else is released. And no: nothing is really different since the blink
fork, apart from bug fixes.

I'm not sure I would include ogg theora still in your examples - WebM
and MP4 is likely sufficient.

Not sure what you want to put into the "Accessibility and ARIA
considerations" section - almost all <track> features cover
accessibility fully and don't require ARIA. Maybe if you want to talk
about audio descriptions, you could say that the plan is to have the
text be voiced by the browser for blind users. However, this is not
implemented yet and is basically waiting for speech synthesis to be a
native browser feature. It can currently be achieved only with browser
extensions. I'd remove that section TBH.

Where did you read "Optional Cue Settings separated from the time by
no more than 2 spaces"? In the spec it says one or more SPACE or TAB
characters are used for separation.

What does this mean: "As far as HTML5 video is concerned the only
difference between captions and subtitles are the way the <code> tags
are structured:" ? I don't follow...

You have <p> tags in the description example - they are not conformant
(and some of them are not closed properly). Just remove them. Every
line is a "paragraph" anyway.

As for metadata tracks: they exist to transport any information that
is supposed to be time-aligned with the video or audio's timeline.
This could be base64 encoded images, JSON, or anything else that a JS
developer would want to interpret at a certain time instance on the
video/audio timeline.

In your conversion from SRT to VTT, you write "Replace them with Cue -
prefixes." I don't understand what that means. Do you mean to give
cues a name (id)? If so, that's not necessary - the numbers of SRT
work perfectly fine, too.

Timestamp tags are not just useful for Karaoke, but also for paint-on
captions. You might want to add that use case.

Some markup has gone wrong around the bold and underline tags.

Adding @default to the text track not only makes that track the
default, but also turns it on by default so it is displayed. Might
need some clarification.

The last example of <track> use has a @kind=captions collection twice
- I think the latter one was supposed to be a @kind=descriptions

That was my quick first reading. HTH.


On Fri, Dec 13, 2013 at 11:17 AM, Carlos Araya <carlos.araya@gmail.com> wrote:
> I took a first stab at writing the VTT tutorial. It's currently hosted in
> Github to make it easier for people to contribute examples and edits to the
> page.
> https://github.com/caraya/vtt-demos/blob/master/video_captioning.md
> Please take a look at it and let me know what you think needs to be changed
> and what other things I should add.
> Carlos
> On Thu, Dec 5, 2013 at 6:00 PM, Silvia Pfeiffer <silviapfeiffer1@gmail.com>
> wrote:
>> Great to hear your interest!
>> I would indeed encourage to write about implemented reality, even
>> though that will probably continue to improve over the next few
>> months.
>> You could explain the features as specified and then explain what
>> works in the browsers now.
>> This could, in fact, contribute to the testing efforts of the <track>
>> element in HTML, see http://www.w3.org/html/wg/wiki/Testing .
>> Cheers,
>> Silvia.
>> On Fri, Dec 6, 2013 at 12:44 PM, Carlos Araya <carlos.araya@gmail.com>
>> wrote:
>> > I'd be interested in looking at working on something like this. However
>> > the first question that comes to mind is whether we write to the way the
>> > specification is written or to the way it is supported. As far as I know
>> > there is no complete implementation of the VTT spec available,is there?
>> >
>> >
>> > Carlos
>> >
>> > Sent from my iPad
>> >
>> >> On Dec 5, 2013, at 12:43 PM, David Singer <singer@apple.com> wrote:
>> >>
>> >> Hi
>> >>
>> >> it would be really cool to have an authoring guide for VTT (unlike the
>> >> specification, which is essentially an implementation guide) on
>> >> WebPlatformDocs.  There are some possible starting points, below, though of
>> >> course to use any of this material wed need to work with the authors and
>> >> get permission.
>> >>
>> >> Is there someone who would like to pull this together and spearhead
>> >> getting a good authoring guide on this important site?
>> >>
>> >>
>> >>
>> >>>
>> >>> http://dev.opera.com/articles/view/an-introduction-to-webvtt-and-track/
>> >>> http://developer.mozilla.org/en-US/docs/HTML/WebVTT
>> >>>
>> >>> http://blogs.msdn.com/b/ie/archive/2011/10/12/html5-video-captioning.aspx
>> >>> http://www.delphiki.com/webvtt/
>> >>>
>> >>> http://demosthenes.info/blog/580/Make-Online-Video-Accessible-And-Searchable-With-WebVTT
>> >>>
>> >>> http://www.accessiq.org/news/features/2013/03/webvtt-and-captioning-on-the-web
>> >>
>> >> David Singer
>> >> Multimedia and Software Standards, Apple Inc.
>> >>
>> >>
Received on Friday, 13 December 2013 01:51:56 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:20:56 UTC