Fix "Improved screen reader accessibility of captions" #1340

[nfreear]

Hi @rylan and @johndyer,

I had my concerns about Pull #1340, but didn't have time to deal with the it until now.

As the code stands for the majority of videos, Pull #1340 does not improve accessibility for screen reader or other users, it degrades it

Most video (the default case), contains spoken commentary, dialog, a talking head or similar. In these cases subtitles or captions, aimed at non-native speakers (subtitles) or those who are hard of hearing (captions) do not contain any extra information over what is audible. So, for a screen reader user, turning them into a live-region merely obscures and interferes with the existing audio track.

There are at least three special cases:

1. The "subtitles" track is actually a text-audio description track, aimed at those who are blind. It contains text that is timed to be spoken by a screen reader when no one is speaking in the audio track of the video;
2. The listener is struggling to hear the speech in the audio track of the video, for example, because of a strong accent, or poor sound levels. The listener mutes the audio, and uses TTS to read the subtitles aloud;
3. A special case, where the video does not contain any spoken words, and a "subtitles" track is used to convey information. This is the case for "echo-hereweare.mp4" and "mediaelement.srt".

My current fix is to add a flag, to explicitly turn the captions into a WAI-ARIA live region.

As it stands, this flag needs to be set when the video player is created, so it doesn't deal with all the special cases above (it deals with 3, partially with 1). It does however, fix the issue I'm raising.

Best wishes,

Nick

[nfreear]

Hi @rylan,

I've realised that what I put above may be a bit brusque, and doesn't give you credit for what is a useful and bold idea! (Its something that I've thought about for a while, and discussed with colleagues here at the Open University.)

We perhaps need to put our heads together, and work out how best to implement your idea, so that an end-user can dynamically switch on live-region pseudo-captions.

Yours,

Nick