Skip to content
This repository

HTML5 polyfill for closed captioning with the <track> element, and implements the WHATWG Timed Text Track specification.

branch: master
README.markdown

Captionator

Simple closed-captioning polyfill for HTML5. Just 8KB when gzipped!

What does Captionator do?

  • Implements the WHATWG TimedTextTrack Specification, complete with the full JavaScript API
  • Supports the <track> element
  • Supports 100% of WebVTT, along with WebVTT v2 proposed features
  • Additional support for SRT, Youtube/SBV, SUB, LRC, and TTML caption/subtitle formats
  • Works in Firefox 3.5+, IE9, Safari 4+, Chrome, Opera 11... basically any browser which supports HTML5 Video!
  • Small, configurable, and under active development
  • Library independent
  • Accessible, with ARIA support
  • Minimal global namespace footprint (written with a closure)
  • Captionator is not a video player (this means you can write your own!)
  • Currently the most comprehensive polyfill available for Closed Captioning!

What can I do with Captionator?

  • Add subtitles to make foreign-language video content available to your audience
  • Add captions for the hard of hearing
  • Add descriptions for the blind or vision impaired
  • Add time-sensitive thumbnails to your custom video seek bar using metadata tracks
  • Display realtime tweets or timed comments on top of your video, like soundcloud
  • Fancy chapter based navigation with the chapters track type
  • Overlay lyrics, interview supers, or explanatory text on your videos
  • ...and much, much more!

You can see a demo of Captionator here: http://captionatorjs.com/demo.html

Using Captionator

After including the library, adding captions to your video is pretty simple:

<script type="text/javascript" src="js/captionator-min.js"></script>
<script type="text/javascript">
    window.addEventListener("load",function(eventData) {
        captionator.captionify();
    });
</script>

This will not only caption your video (this example will caption every element on the page with Timed Text Tracks available to it,) but it will also provide a .tracks property on your video element(s) - which you can use to dynamically manipulate the track data as per the WHATWG specification.

If you've got specific requirements about which videos get captioned, and in what language(s), there are some extra options:

captionator.captionify(videoElementsToCaption,defaultLanguage,options)

The first parameter can be an array of selectors or DOMElements, or a single selector string or DOMElement. The second parameter is a language string.

You can use the options parameter to specify your own render function for captions, if you don't like captionator's inbuilt renderer:

captionator.captionify(["#yourVideoElement1","#yourVideoElement2"],"de",{ renderer: myFunction });

(More on this below!)

Multiple subtitles and custom render functions

It's pretty straightforward to manage multiple enabled subtitle tracks. Take this set of track elements for example:

<track kind="captions" src="subs/english-subs.srt" srclang="en" label="English Subtitles" default />
<track kind="captions" src="subs/german-subs.srt" srclang="de" label="German Subtitles" />
<track kind="captions" src="subs/japanese-subs.srt" srclang="ja" label="Japanese Subtitles" />

In this case, the English subtitles are enabled by default. Unless you specify a custom renderer, Captionator will automatically generate as many separate containers as are required for enabled tracks, set up the relevant events and styles.

Specifying a custom renderer

Should you wish to specify your own renderer, you can use the following syntax when calling captionator.captionify:

captionator.captionify(null,null,{
    "renderer": function(yourHTMLVideoElement) {
        ...
    }
});

The renderer function you define is executed, and passed the HTMLVideoElement whenever it fires a timeupdate event. You can use the TextTrack.activeCues to determine what cues should be displayed at any given time.

The event data for the video timeupdate event is not passed to your function. This is because you are defining a renderer, not an event handler. Should future Captionator updates require caption re-rendering on different events (or in places not triggered by events at all) code which depends on event information will not function correctly.

Enabling and disabling subtitle tracks programatically: A Quick Guide

You can find a demonstration of this feature in the example file.

Getting Tracks

Captionator simply makes a new property (array) available through javascript on the HTMLVideoElement:

var myVideo = document.getElementById("myVideo");
var myTracks = myVideo.tracks;

By extension, getting access to the track you want is as simple as:

var firstSubtitleTrack = myVideo.tracks[0];

Each track defines the following user accessible properties:

  • label - String - describes the track (in plain human language)
  • language - BCP47 language string which describes the track
  • kind - Resource type (one of subtitles, captions, chapters, descriptions, metadata.)
  • readyState - indicates whether the resource is loaded (one of NONE/0, LOADING/1, LOADED/2, or ERROR/3)
  • mode - the most important property (probably!) - determines whether captionator will fetch and render the resource.
  • cues - A TextTrackCueList (functionally, an array) containing all the cues for the track
  • activeCues - A TextTrackCueList containing all the cues for the track which are currently active
  • videoNode - the HTMLVideoElement which the track relates to/extends. (Not in the WHATWG spec.)

Ergo, to access the property language from the third track, you'd use the following code:

var thirdTrackLanguage = myVideo.tracks[2].language;

To enable or disable a track:

myVideo.tracks[2].mode = captionator.TextTrack.SHOWING; // Equivalent to (integer) 2
myVideo.tracks[2].mode = captionator.TextTrack.HIDDEN; // Equivalent to (integer) 1
myVideo.tracks[2].mode = captionator.TextTrack.OFF; // Equivalent to (integer) 0

The track is then enabled/disabled when the video fires a timeupdate event, or when a track mode changes. You can update it immediately (although Captionator handles this itself in nearly every case) like so:

captionator.rebuildCaptions(myVideo);

(Where myVideo is an instance of a captioned HTMLVideoElement)

For a more advanced example, see the subtitle selector in the example file.

Options

The following lists options which you can pass to captionator:

  • enableCaptionsByDefault (Boolean) - determines whether to show captions by default, if a caption language matches the user's UA language or is selected for display according to the rules outlined in the WHATWG specification. Tracks with the enabled attribute set to true will be displayed regardless of this option.
  • enableDescriptionsByDefault (Boolean) - as above, except for description track types instead of caption or subtitle types.
  • exportObjects (Boolean) - instructs Captionator to export its own implementation of the TimedTextTrack objects (TextTrack, TextTrackCue, etc.) and their relevant constants into the global scope. Captionator ordinarily keeps these within its own object. You might find this useful for creating code which utilises instanceof conditionals, or creates instances of these objects itself, which you want to be native-TextTrack-support-agnostic. (Phew, what a mouthful.)
  • renderer (Function) - sets an alternative renderer for captions & subtitles. You can utilise the WHATWG TimedTextTrack specification to manipulate or get information about the tracks themselves.
  • processCueHTML (Boolean) - determines whether HTML/WebVTT cue source is parsed. Defaults to true. If this is set to false, cue source will be retained as unprocessed text, and special WebVTT cue spans will be appended straight into the DOM (rather than perform their function as detailed in the WebVTT specification.) metadata tracks are never processed, regardless of the value of this setting.
  • sanitiseCueHTML (Boolean) - determines whether non-WebVTT-compliant tags are dropped when parsing, thereby sanitising the source of WebVTT cues. Defaults to true. Cue source is not sanitised when processCueHTML is set to false.
  • ignoreWhitespace (Boolean) - By default, line breaks (single) within cues are converted to
    elements in HTML. Set this to true to prevent whitespace from changing processing behaviour. By default, this is false.
  • controlHeight (Integer) - defines an 'exclusion zone' (where cues will not be rendered) at the bottom of the video to allow for video controls. The available area for cues is determined based on the height of the video less the height of the video controls. By default, if the controls attribute is present on the video element, this is calculated automatically based on the user agent. Should the controls attribute be missing, this value is zero. If you want to implement your own controls, use this to tell captionator how tall they are.
  • debugMode (Boolean) - If true, draws a canvas with debugging information for cue positioning on top (in z-space) of the video. The canvas displays vertical, vertical-lr, and horizontal line divisions, as well as Captionator's own understanding of the available cue area (post cue-rendering.) This option is not available in the minified builds of Captionator.
  • appendCueCanvasTo (HTMLElement | DOM Query as string) - Defines a node in the document within which Captionator should render the video cues. This function is intended to allow you to create a wrapper div and have Captionator render cues within it - hopefully easing the process of making a custom video player. If successful, and Captionator is able to find the wrapper node based on your input, it will set the top and left values of its own cue canvas to zero, rather than finding the offset position of the video element itself, and append its cue canvas within the wrapper when rendering. If the query fails, the cue canvas will be appended to the body as normal, and positioned using the offset position of the video element.
  • enableHighResolution (Boolean) - If true, Captionator sets up a 20ms timer for refreshing cues and captions, firing much more rapidly than the default timeupdate event listener on the video element. This option causes Captionator to use a lot more of the user's CPU time - only use this if you have a real need for quick, <250ms response times. This option defaults to false.
  • forceCaptionify (Boolean) - If true, Captionator will ignore checks for native TextTrack support in browsers and press on ahead anyway. It will also not fail if the element/s to be captioned are not videos. If you're not sure about this, DO NOT enable this option. It will only create more bugs. Do not enable in production.

Styling Options

  • minimumFontSize (Float) - Defines the minimum allowable font size with which Captionator will render cues (in points.) Defaults to 10pt.
  • minimumLineHeight (Float) - Defines the minimum line height with which Captionator will render cues (in points.) Defaults to 16pt.
  • fontSizeVerticalPercentage (Float) - The cue font size as a percentage (0 - 100) of the height of a given captioned video. Defaults to 4.5%.
  • lineHeightRatio (Float) - The ratio of line height to font size. Defaults to 1.5.
  • cueBackgroundColour (Array) - An array containing four items, each for: red (R), green (G), blue (B), and alpha (A), in that order, which define the background colour of cues. Defaults to [0,0,0,0.5].
  • sizeCuesByTextBoundingBox (Boolean) - Instructs Captionator to set the cue size by the default bounding box of the text, rather than size them to 100% of the available rendering area (the WebVTT specification's method, and Captionator's default method.) False by default.

Video and Audio tracks (MediaTrack)

A/V track support has been removed from Captionator as the specification changed.

Building Captionator.js

Captionator.js comes pre-built in both unminified and minified versions, but you can build it yourself.

You'll need node.js, jake, jshint, and uglify-js. You can use this guide to installing node.js if you get stuck.

Once that's done, install the dependencies with npm:

npm install --global jshint
npm install --global jake
npm install --global uglify-js

All done? Now, do whatever you need to do to Captionator by modifying its various components in the /source directory. When you're ready to build, you can just:

jake

The default jake task builds and lints the file, minifies it, and runs through the test suite. That's it! All files created are placed in the /js directory, ready for production use.

You can also independently run each of the following tasks:

  • jake test - runs the Captionator test suite
  • jake build - builds the unminified Captionator.js
  • jake minify - generates the minified captionator-min.js
  • jake lint - runs JSHINT over the unminified file.

Autobuild

If you use a Mac, and Safari, autobuild.pl is a little script which monitors the source files, and then rebuilds, minifies, and lints captionator.js - and reloads your Safari tab whenever it detects changes. Using it is as simple as running:

./autobuild.pl

Notes about newer browsers

As browers start to implement WebVTT and the TextTrack JS API, Captionator.js will step back and your captions will be delivered by the native capabilities of those browsers. In some cases though, the implementations are buggy or incomplete. Currently, both IE10's and Chrome's WebVTT & JS API implementations are incomplete. Captionator.js will not run in these browsers. Instead, it quietly quits. You can detect whether this is the case by checking for a false return value from the captionator.captionify() function.

Currently, Chrome's implementation of TextTracks breaks Captionator, but doesn't actually work on its own yet. I advise that you turn off TextTrack support in chrome://flags.

IE10's implementation is not complete, but it does work.

If you understand that you'll probably break things but want to try it anyway, you can disable Captionator's TextTrack support checking by setting the forceCaptionify option to true.

New Features

  • Supports WebVTT proposed features such as DEFAULTS, STYLE, and COMMENT cues
  • Now supports TTML and LRC
  • Advanced layout engine ensures cues are appropriately sized and positioned regardless of how many are enabled at any one time (with limits, obviously!)
  • Optional auto cue sizing algorithm, which sizes the cue to the text bounding box
  • Brand new WebVTT renderer, with new styling options!
  • Performs automatic validation of WebVTT cue spans and HTML!

Thanks

Thanks to @silviapfeiffer for her knowledge and assistance with some of the hairier aspects of these specifications! Thanks also to @ourmaninjapan for his welcome assistance with Japanese text/line breaking algorithms!

Licence

Copyright (c) 2012, Christopher Giffard All rights reserved.

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

  • Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
  • Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Something went wrong with that request. Please try again.