Video Captioning in HTML5
At first glance, video captioning might not seem like a huge deal. Maybe it might appeal to you
if you need to meet accessibility requirements in your workplace, or you're just doing it to be nice to
people with hearing difficulties.
You're pretty far from the mark though. Yes, accessibility is an important driver for video captioning on the web.
But it's about much more than ‘deaf people’
you should be providing quality captioning for them anyway!)
In addition to important accessibility benefits, captioning your video can allow automatically generated
transcripts, text search-ability within videos (imagine what will be possible once Google supports crawling
WebVTT!), chapter markers and powerful metadata capability, and much more!
You could provide a track of time-specific comments on the video (the same way SoundCloud
allows commenting on specific parts of audio) or even a live twitter feed on your video (similar in style to the
ABC'S Q&A program.)
“Alright!”, you say. “I'm convinced. How does HTML5 allow me to caption my video?”
For a while, there wasn't an easy way to do it. Guys like Bruce Lawson were coming up with
creative ways (if a little hacky)
to implement support, in lieu of a proper standard.
Now, however, there's a really nice way to provide support - the HTML5
<track> element and its
API. There's a slight catch though. No currently shipping browser supports either of these features/standards.
They will soon - Webkit has support in development, and
Firefox can now parse the
<track> element (but can't do anything with it yet.)
But you want to caption your HTML5 video now. And for that, you need Captionator.js.
designed to provide support for the
supports subtitles in SUB, SRT,
Youtube's SBV, WebVTT, and
WebVTT with Google's proposed timestamp syntax. It will soon support
LRC, DFXP/TTML and the
WebVTT v2 proposed features.
Like most polyfills, it's intended to be a drop-in addition to your existing HTML5 video (and/or custom player)
specification. This makes it easy to write your own interface for allowing the user to select assistive or
media enhancement tracks, and when browsers land support for the feature, it'll move out of the way and let the browser
take over (and you won't have to change a line of code!)
Captionator.js will automatically detect and enable
<track> elements specified in markup - it's a totally
painless solution! See the Captionator.js documentation to get started!
Captionator.js Out & About
Captionator was recently recommended and demonstrated by Jer Noble of Apple at WWDC 2011!
Check out Session 502: Advanced HTML5 Media Controllers.
(Unfortunately, you'll need to sign up for a free Apple Developer account - it's
a great video though!.)
Silvia Pfeiffer presented on WebVTT to an audience at GoogleIO
in March this year, and used Captionator.js for the demo. This is a great talk for getting your head around WebVTT
as a format.
CaptionCrunch, the old development release, is now stable!
CaptionPlanet is the next
stage frontier in the Captionator.js roadmap.
At this stage, intended to be the 0.6 release milestone, the goal for CaptionPlanet
is performance, compatibility, better testing, and broader format support.
Some things which slipped from the 0.5 CaptionCrunch release (in the interests of getting
the new Captionator.js WebVTT renderer out the door) will be delivered in CaptionPlanet
- Support for WebVTT inline styles (Captionator currently parses these, but does nothing with them)
- Enabling support for cue styles in page stylesheets
- LRC (and possibly SSA/ASS, depending on whether the current renderer will support the formats) subtitle support
This release is slightly less ambitious than 0.5 CaptionCrunch, but every bit as important.
As Captionator.js is an open source project, you can
see all these new features come to fruition in real time by following progress of the
branch on github! Just be aware that things might not always be pretty, or even work.