Re: Captioning verbiage for the Text Spacing Understanding doc

Hi Laura,

I had to go back and figure out what this was about, but after reading [2]
http://rawgit.com/w3c/wcag21/master/understanding/21/text-spacing.html, it
comes back to me now. Note that my concern/issue is similar to @steverep's
issue (canvas = images of text).

At issue is that not all captions are provided as stand-alone text files,
and as such, not all captions will have the technical ability to be
'reformatted' because they are, for all intents and purposes, "images of
text" burned into the "moving pictures" video file (in a similar way that
text may be part of a JPEG or PNG file).

In effect, there are multiple ways of providing captions. The most common
are:

   1. The caption file is and remains a stand-alone text file (using a
   timed-text markup language, either TTML or WebVTT), and the caption file is
   supplied via the @track element, like this:
         <video poster="myvideo.png" controls>
             <source src="myvideo.mp4" srclang="en" type="video/mp4">
             <track src="myvideo_en.vtt" kind="captions" srclang="en"
   label="English">
         </video>
   (This is often referred to as out-of-band captions)

   2. The caption file (again, similar to above, a time-stamped text file)
   is "bundled" inside of the video wrapper (i.e. .mp4 or .mkv) - these are
   file formats that are wrappers or containers that traditionally include the
   H.264 encoded video stream, the AAC encoded audio stream, and the wrapper
   formats can also include other associated data files such as the WebVTT
   file. Here, the user-agent "looks inside" the video wrapper for text files
   that correspond to caption files, and when present exposes them to the end
   user. See
   https://www.w3.org/TR/2011/WD-html5-20110113/video.html#sourcing-in-band-text-tracks
   for more specifics, but the key thing to remember is that the captions are
   still being supplied via a seperate time-stamped text file, it's just that
   the delivery method (in-band) is different: the caption file is 'embedded'
   into the video wrapper, and not referenced via the stand-along @track
   attrribute.

   3. A third method of providing captions is done at the post-production
   level, where text is 'burned' directly into the moving image files. While
   less common today than in the past, for certain languages and under certain
   conditions this remains a valid and viable option (as it meets the
   functional need of providing captions). These types of captions are also
   traditionally known as "Open Captions" as they are present and
   persistent to all users, and cannot be disabled or hidden, as the text
   rendered on screen is actually part of each individual frame of the video
   asset itself.

It is this third condition that, due to technical limitations, needs to be
exempted from the SC - there is no way for the end user to manipulate the
on-screen 'text', because it isn't text, it's a special type of image. The
core difference is that there is no "text file" being used to provide the
captions, so there is nothing that can be further modified by the
user-agent(s).
(The following video may help better explain the production process:
https://www.youtube.com/watch?v=WzADacd7NMo)

To wrap things up, perhaps the following may meet your need?

Examples of text that are typically not affected by style properties
<https://www.w3.org/TR/WCAG21/#dfn-style-properties> are images of text
<http://rawgit.com/w3c/wcag21/master/guidelines/index.html#dfn-images-of-text>
and
video captions embedded directly into the video frames, and not provided as
an associated caption file, which are not expected to adapt.

​HTH.

JF​






On Thu, Jan 4, 2018 at 6:32 AM, Laura Carlson <laura.lee.carlson@gmail.com>
wrote:

> Hi John,
>
> At the November 13, 2017 AGWG teleconference, you mentioned that the Text
> Spacing Understanding doc should have verbiage to explain what types of
> captions are not expected to adapt to the SC's metrics [1].
>
> I put a place holder in the document [2] so it wouldn't fall through the
> cracks. Would it be possible for you please suggest appropriate verbiage to
> use?
>
> We are very fortunate that you have deep expertise in captioning. Thank
> you very much for your consideration.
>
> Kindest Regards,
>
> Laura
>
> [1] http://rawgit.com/w3c/wcag21/master/guidelines/index.html#text-spacing
> [2] http://rawgit.com/w3c/wcag21/master/understanding/21/text-spacing.html
>



-- 
John Foliot
Principal Accessibility Strategist
Deque Systems Inc.
john.foliot@deque.com

Advancing the mission of digital accessibility and inclusion