Content features of the resource, such as accessible media, supported enhancements for accessibility and alternatives.


Example 1 — EPUB 3
<meta property="schema:accessibilityFeature">captions</meta>
<meta property="schema:accessibilityFeature">alternativeText</meta>
Example 2 — EPUB 2
<meta name="schema:accessibilityFeature" content="captions"/>
<meta name="schema:accessibilityFeature" content="alternativeText"/>
Example 3 — Audiobooks
"accessibilityFeature": ["captions", "alternativeText"]
Example 4 — HTML with RDFa
<p>Video includes
   <span property="accessibilityFeature">captions</span>.</p>
Example 5 — HTML with microdata
<p>Video includes
   <span itemprop="accessibilityFeature">captions</span>.</p>


The following values are used to identify features of the content that are helpful for accessibility:


Unless the metadata format used supports lists (e.g., JSON-LD), use a separate property declaration for each feature. Do not merge the values into a single property using spaces, commas, or other delimiters.


The alternativeText value is used to indicate that all images and other non-text content (e.g., audio and video) that is relevant to understanding a publication includes a text alternative.

Alternative text provided using the HTML alt attribute.

<img src="rabbit.jpg" alt="Bugs Bunny" />

Presentational images (i.e., that are only used for purely display or ornamentation purposes) are excluded from the requirement for alternative text, so not every image or instance of non-text content will have a text alternative when this property is set.

A presentational image.

<img src="spacer.jpg" alt="" role="presentation" />

See the page navigation knowledge base page for more information.


The annotations value is used to indicate that a publication comes with explanatory annotations that improve the accessibility of the work (i.e., the value should not be used where the annotations are designed for general use, such as in annotated teacher's editions or exercises with explanatory notes).

Examples of annotations with an accessibility benefit could include extended descriptions for images, plain language explanations of complex theorems, and text identifiers for visual cues.

Annotations can either be part of text of a publication or can be authored in a separate technology such as Web Annotations.


The Open Annotation in EPUB specification is not currently known to be supported in any reading systems. A proposal is expected to be adopted to include accessible annotations in ARIA 1.3, however. As a result, this value is rarely applicable to EPUB publications at this time.

Do not use this value for ruby annotations. See the rubyAnnotations value instead.


The audioDescription value is used to indicate that description tracks are included for any video sources that require them.

Audio descriptions differ from captions and subtitles in that they describe the actions that are occurring in the video (i.e., they provide more information that just what is being spoken). Audio descriptions are most helpful when the content cannot be seen.

Audio description provided for an HTML video using a track element.

   <source src="speech.mp4" type="video/mp4"/>
   <track src="description.vtt" kind="descriptions" srclang="en" label="English"/>


The HTML audio element cannot be used to render audio descriptions, as it does not provide a display area for them.


The bookmarks value is used to indicate that a PDF includes a list of bookmarks. PDF bookmarks are like a table of contents but may include destinations beyond the major headings. See the WCAG technique "Creating bookmarks in PDF documents" for more information.

This value is not used in web-based formats like EPUB. Although reading systems provide a different kind of bookmarking feature, these bookmarks lack a common distribution format and are more of a general usability feature than an accessibility one.


The braille value is used to indicate that either that a publication comes in a preformatted tactile form or that an alternative rendition in this medium is included.

This value is primarily used by accessible republishers who produce HTML files using braille Unicode characters for the text.

A braille-formatted paragraph.

<p>⠠⠉⠁⠇⠇ ⠏⠑ ⠠⠊⠎⠓⠏⠉⠇⠇⠲ …</p>

The captions value is used to indicate that a publication includes captions for auditory content.

Captions provide a transcription of the dialogue that is being spoken as well as describe any relevant sounds that are occurring (e.g., music playing, a phone ringing, etc.). They are intended to help when the audio is muted or cannot be fully heard.

Captions are not the same as subtitles, which are strictly a transcription of the dialogue (and typically available in additional languages than that being spoken). They differ from audio descriptions in that they do not describe actions.

Captions provided for an HTML video using a track element.

   <source src="speech.mp4" type="video/mp4"/>
   <track src="captions.vtt" kind="captions" srclang="en"
          label="English for the hard of hearing"/>


The HTML audio element cannot be used to render captions, as it does not provide a display area for them.


The ChemML value is used to indicate that a publication contains chemical information (e.g., molecules and reactions) encoded using the ChemML markup language.

This value is not typically used with publications at this time as there is no support for rendering ChemML in HTML.

Example of ChemML markup.

<cml xmlns="" title="Chem sample" …>
   <molecule id="mol" …>

The describedMath value is used to indicate that a publication includes textual descriptions for math equations.

Text descriptions can be provided, for example, using language-specific technologies like the MathML alttext attribute. They can also be provided using the HTML alt attribute when only images are used. The value is a human readable equivalent.

Math description provided in an alttext attribute.

<math xmlns="" display="inline"
      alttext="(mathml alt-text): y minus y 1 equals StartFraction y 2 minus y 1 Over x 2 minus x 1 EndFraction left-parenthesis x minus x 1 right-parenthesis">

This property is complementary to the MathML and latex properties, as the inclusion of either MathML- or latex-formatted equations does not infer that text descriptions will be available.


The displayTransformability property is used to indicate that a publication does not restrict the ability of users to modify the display of textual content (e.g., to change the font family, font size, line height and word spacing).

In order to set this value, all text content must be encoded as text (i.e., there must not be any images of text). In addition, absolute length units must not be used in the styling (e.g., pixels and points).

It is not required that a publication provide the means of transforming the text. It is expected that users who require this feature will have a reading system that with display transform options.


The highContrastAudio value is used when a publication contains audio with speech and there is low or no background noise. Low background noise is defined as a minimum 20dB difference between foreground speech and background sounds (see WCAG success criterion 1.4.7).

High contrast audio is helpful for users who are hard of hearing and have trouble separating foreground speech from background noise.

The requirement for low or no background noise does not apply to musical works where the background sound is integral to the presentation.


The highContrastDisplay value is used to indicate that there is at least a 7:1 contrast ratio between the foreground text and the background, including for images of text (see WCAG success criterion 1.4.6).

High contrast display is helpful for users with moderately low vision or with color deficiencies.


The index value is used to indicate that a publication contains one or more indexes of the content (e.g., a topical index or index of names).

Indexes allow users who cannot see the text or have difficulty following it to find and return to subjects more easily.


The largePrint value is used to indicate that the text of a publication has been formatted to meet large print guidelines. Large print guidelines vary from region to region, but generally require a point size ranging from 14 to 18, use of sans serif fonts, and increased word spacing.

The use of this value should not be confused with display transformability. The text must be styled by default for large print reading to use this value, not that it can be modified for large print reading.

Large print is still predominantly a print medium, so outside of accessible republishing it is not common to use this value.


The latex value is used to indicate that a publication contains math equations formatted using the LaTeX typesetting system.

LaTeX equations can retain semantic information about math equations that is not always available in other formats (e.g., presentational MathML) making them easier to understand.

The ability to read LaTeX equations often requires being able to export the data to another program, however, as LaTeX does not natively render in Web-based formats like EPUB. As a result, it is not commonly found at this time.

A latex-formatted equation.

\begin{equation} f(x)=(x+a)(x+b) \end{equation}

The longDescription value is used to indicate that all complex content (images, tables, charts, equations, etc.) include an extended description of their purpose.

Complexity is determined by whether alternative text alone is sufficient to convey the full information of the content (e.g., for images and charts), or whether the structure is likely to be understood by someone who cannot see it (e.g., for tables). See the techniques for WCAG success criterion 1.1.1 for more information on when extended descriptions are needed.

See the images knowledge base page for more information.

Long description included in an HTML details element.

<figure id="fig01">
      <p>Figure 1 - Spectral analysis of an earthquake</p>
         <p>The chart shows a typical …</p>

Hidden long description attached using the aria-describedby attribute.

<img src="speheres.jpg" alt="Celestial spheres" aria-describedby="img-desc"/>
<p id="img-desc" hidden="">The 47 spheres are shown …</p>

The MathML value is used to indicate that the MathML markup language has been used to encode the mathematical equations. For Web-based formats like EPUB, Presentational MathML is used to encode the content as it can be displayed in a growing number of reading systems and browsers.

Although Presentational MathML often lacks important semantic information that is only available in Content MathML, it is supported by assistive technologies which allow users to more easily navigate and understand the content.

The MathML value is often paired with the describedMath value when fallback descriptions are provided.

See the MathML knowledge base page for more information.

MathML equation for definition of a circle.

<math xmlns="" display="block">

The none value is used to indicate that a publication does not include any accessibility features.

This value exists so that there is no ambiguity about whether a publication does not contain any accessibility features or whether it was never checked for them.


This value has been proposed for inclusion but has not yet been accepted. See issue 12 for more information.

The pageNavigation value is used to indicate that a publication includes a page list.

The page list is a navigation aid that provides a list of links to static page break locations (i.e., that do not change based on screen size or when the display is transformed). The static page breaks typically represent print page numbers, but publishers sometimes insert static locations in purely digital works to ease coordination of locations.

The inclusion of a page list is particularly useful where mixed print-digital environments (such as in educational settings) are anticipated, as the page list allows digital users to coordinate their locations with print users. In these cases, the printPageNumbers value is also set.

See the page navigation knowledge base page for more information.

EPUB 3 page list.

<nav epub:type="page-list" role="doc-pagelist" aria-labelledby="pglist">
   <h2 id="pglist">Page List</h2>
      <li><a href="intro.xhtml#pg1">1</a></li>
      <li><a href="intro.xhtml#pg2">2</a></li>

The printPageNumbers value is used to indicate that static print page break locations are included in the text.

The inclusion of page break markers from a print source allows users to identify where they are in a digital publication relative to its print equivalent. The page break markers are typically accompanied by a page list that allows users to quickly navigate to a specific location.

See the page navigation knowledge base page for more information.

Page break identified using the DPUB-ARIA doc-pagebreak role.

<span id="pg24" role="doc-pagebreak" aria-label="page 24"/>

The readingOrder value is used to indicate that there is a logical reading order to the text of a publication.

The logical reading order is the progression from one element of content to the next (e.g., a heading for a chapter is followed by the first paragraph and so on). When elements are out of order in the markup and styled to appear linear, assistive technologies will present them to users in the incorrect order.

The logical reading order is also established by ensuring that primary content can be distinguished from secondary. The use of HTML elements like figure and aside helps identify secondary content, allowing users to skip over it if they choose.

Fixed-layout publications tend to be the most problematic due to the fixed positioning of elements and having to represent secondary content that spans two pages, but it is possible for fixed layouts to have a logical reading progression if care is taken.

See the logical reading order knowledge base page for more information.


The rubyAnnotations value is used to indicate that a publication includes ruby annotations.

Ruby annotations are most commonly used to provide pronunciation guides for Chinese, Japanese and Korean texts.

Ruby annotation in HTML.


The signLanguage value is used to indicate that sign language interpretation is included for any auditory content.

Although captions and audio descriptions serve a similar purpose, many readers who are deaf prefer sign language interpretation as it is their primary means of communication so quicker and easier to follow.


HTML does not offer a simple mechanism for synchronizing sign language interpretation with a video. If interpretation is not part of the video stream, authors will have to provide a separate video overlay or play interpretation side-by-side with the video.


The structuralNavigation value is used to indicate that all headings are properly marked up.

The use of headings tags, and their correct numbering, allows users of assistive technologies to quickly move through the different levels of the book without having to traverse the table of contents.

See the HTML headings knowledge base page for more information.

Headings incremented for each subsection using the HTML h1 through h6 tags.

   <h1>Section 1</h1>
      <h2>Section 1.1</h2>

The synchronizedAudioText value is used to indicate that synchronized playback of prerecorded audio with text highlighting is available.

Synchronized playback can help new language learners follow the text and hear how it is pronounced, for example, but is also helpful for users who need only one or the other modality. Audio playback can be used to provide an audiobook experience for users who are blind or have low vision, while text highlighting can help users who have difficulty following lines of text.

Text and audio synchronization is typically provided using the SMIL language (e.g., in DAISY talking books and EPUB 3 media overlays).

See the media overlays knowledge base page for more information.

EPUB 3 media overlay file.

<smil xmlns=""
  	<seq id="id1" epub:textref="chapter_001.xhtml"
             epub:type="bodymatter chapter">
            <par id="heading1">
               <text src="chapter_001.xhtml#c01h01"/>
               <audio src="audio/mobydick_001_002_melville.mp4"

The tableOfContents value is used to indicate that a publication includes a table of contents for navigating the major sections of the work.

The table of contents is an important aid for navigating a publication for those who cannot see the text.

The table of contents is an integral part of formats like EPUB so is always expected to be present. Authors still need to ensure that the table of contents is sufficiently complete (e.g., includes at least all the top-level headings), otherwise it is incorrect to set this value.

See the table of contents knowledge base page for more information.

EPUB 3 table of contents

<nav epub:type="toc" role="doc-toc" aria-labelledby="toc-hd">
   <h2 id="toc-hd">Contents</h2>
      	   <li><a href="s04.xhtml#intro">INTRODUCTORY</a>

The taggedPDF value is used to indicate that a PDF includes accessibility tags to improve its readability.

Accessibility tags are a way of labeling the components of a PDF to try and retain some of the semantics that are otherwise only visually present. They are a key part of the PDF Universal Accessibility (PDF/UA) standard.

This value is not relevant to Web-based publication formats such as EPUB.


The tactileGraphic value is used to indicate that a publication includes access to tactile graphics.

As EPUB reading systems typically cannot present tactile graphics to users, hyperlinks to downloadable graphics are a common way to make these available. It is not required that the graphics be rendered in the publication for this value to apply.

It is not common for this value to be used outside of accessible republishing, however, due to the specialized work required to create tactile graphics.


The tactileObject value is used to indicate that a publication includes tactile objects.

Due to the need to print tactile objects, hyperlinks to downloadable files are the most common way to make these available.

It is not common for this value to be used outside of accessible republishing, however, due to the specialized work required to create tactile objects.


The timingControl value is used to indicate that a publication contains content with timed interactions (e.g., a quiz or game with a time limit) and that the timing can be controlled by the user.

As users with disabilities may require additional time to complete tasks, they need to be able to disable, adjust or extend time limits.


The transcript value is used to indicate that transcripts are provided for any auditory content.

Transcripts allow users who cannot hear the audio or have difficulty following the speech to read what was spoken. They are an alternative to embedding audio descriptions or captions, which may not always be playable.

Adding a link to a transcript

<audio src="speech.mp3"/>
<p><a href="transcript.html">Read a text transcript of the speech</a></p>

Although including a transcript as part of the EPUB publication is the best way to ensure its availability, hyperlinks to external transcripts are also acceptable (e.g., publishers may have a common web-based location for all their videos and transcripts).


The ttsMarkup value is used to indicate that a publication's text has been enhanced with additional phonetic markup to improve text-to-speech (TTS) playback.

Synthetic speech engines often struggle to render complex words and heteronyms (words that are spelled the same but sound different). The result is that it can be very difficult for users who cannot see the text to understand what is being read out.

The SSML markup language allows the phonetic renderings of words to be included in the markup of publications to improve rendering, for example. PLS pronunciation lexicons similarly allow libraries of phonetic pronunciations to be attached to HTML content.


The unlocked value is used to indicate that no digital rights management (DRM) or other content restriction protocols are applied to the content.

DRM schemes are often applied to digital publications to protect them from being copied, but these schemes, or the rights applied to the content, can sometimes interfere with the ability of assistive technologies to access or play back the text.


It is often vendors who apply DRM schemes to publications before distributing them to users. As a result, before setting this value, authors need to ensure that the publication will not subsequently have DRM applied. The pubication has to be guaranteed to be unlocked after distribution.

Related Links