WCAG: Web Content Accessibility Guidelines
The World Wide Web Consortium’s Web Content Accessibility Guidelines (WAI) are the leading international, nonpartisan standard for accessibility on the Web. Version 1.0 of WCAG, published in 1999, is still in effect, while Version 2.0 is under development.
Multimedia are considered “Web content” and must be made accessible under WCAG 1.0. Checkpoint 1.4 states: “For any time-based multimedia presentation (e.g., a movie or animation), synchronize equivalent alternatives (e.g., captions or auditory descriptions of the visual track) with the presentation.” That’s a Priority 1 requirement; all sites claiming to meet the Web Content Accessibility Guidelines must meet the requirement if they provide multimedia.
(Priority 1 is the base level of accessibility, with Priority 2 and 3 as higher levels. You can’t meet the higher levels without meeting the immediately-lower level. Hence all sites with multimedia that claim to conform to WCAG 1.0 must provide captioned multimedia. Exceptions – as with silent videoclips – could be imagined, but are rare.)
Techniques to meet Checkpoint 1.4 include the following:
Auditory presentations must be accompanied by text transcripts, textual equivalents of auditory events. When these transcripts are presented synchronously with a video presentation they are called captions and are used by people who cannot hear the audio track of the video material.
The long-anticipated update to the Web Content Accessibility Guidelines, namely version 2.0, had not been completed by the end of the TILE project. One draft that was current at the end of the project accurately represents the trend of WCAG 2.0 regarding captioning:
Level 1 Success Criteria for Guideline 1.2
- [...] Captions are provided for all significant dialogue and sounds in time-dependent material. [I]
- Descriptions and captions are synchronized with the events they represent. [I]
A text transcript or other non-audio equivalent does not need to be synchronized with the multimedia presentation if all four of the following statements are true:
the content is real-time and
the content is audio-only and
the content is not time-sensitive and
the content is not interactive
This exception applies to both success criteria 2 and 3 above.
- If the Web content is real-time video with audio, real-time captions are provided. [I]
If the content is a music program that is primarily non-vocal, then captions are not required.
- If a presentation that contains only audio or only video requires users to respond interactively at specific times during the presentation, then a synchronized equivalent presentation (audio, visual or text) is provided. [I]
Exception: If content that is rebroadcast from another medium or resource meets accessibility requirements for that medium, then the rebroadcast satisfies this checkpoint if it complies with other applicable sections of WCAG 2.0
Level 2 Success Criteria for Guideline 1.2
- Synchronized captions are provided for all real-time broadcasts.
Strengths and weaknesses of WCAG 2.0
Both WCAG versions have the indisputable strength of requiring that multimedia be accessible. The exact methods that authors may use, though, are a weakness. WCAG 1.0 defines captions as a certain kind of transcript, but states that a transcript is all that’s required. Authors can meet Checkpoint 1.4 by simply providing a text or HTML document (or any kind of document) that transcribes whatever multimedia they provide. Synchronized transcripts – that is, captions – are not actually required.
In WCAG 2.0, the loophole to provide transcripts is removed, though the final 2.0 Guidelines may authorize providing nothing but a separate transcript under limited circumstances (e.g., video of very short duration, as yet undefined).
WCAG 2.0’s clause stating that multimedia that meets some other requirements automatically meets WCAG 2.0 is open to abuse. A broadcaster with a requirement to caption less than 100% of its programming can simply select uncaptioned programming and provide that online. (After all, that programming meets the requirement that only some of the broadcaster’s programming be captioned.)
- Captioning is still mandatory for all levels of compliance.
- Real-time captioning is explicitly required for the first time.
- Authors have to provide a transcript for certain audio Webcasts. Exactly which kinds of Webcasts are meant to be covered by the guideline is not clear, but it seems to exempt Web radio stations. Lectures or one-time Webcasts might be exempted, or might be exempted only as they are first transmitted (to meet the “real-time” requirement); archived Webcasts might need to be captioned. The exemption is ambiguous.
- Programs with non-vocal music are routinely captioned, including dance, opera, and musicals. Wordless music videos have been captioned. Music captioning is poorly understood and even more poorly done, but it is certainly possible. The guidelines also fail to distinguish among:
The rationale seems too thin, and there are too many exceptions to the rule, to support this exemption.
- programs that include nothing but music (as in Web radio stations)
- hosted programs with a spoken (or even signed) introduction and/or closing
- streams that are usually nothing but music except for unusual times of day or days during the week when DJs are on the air
- “Interactive audio or video segments” have not really been explained by the WCAG Working Group. Perhaps the intent is to refer to online quizzes or games. Those indeed should be captioned.
- The exemption for “content that is rebroadcast from another medium or resource” could result in most of that content being uncaptioned.
- The intent is apparently to capture Webcasts or Web rebroadcasts of TV programming. But very few broadcasters have 100% captioning requirements, and even in those cases there are exemptions. (CBC Television and Newsworld have 100% captioning requirements, though outside commercials don’t have to be captioned. The Astral-owned movie networks in Canada also have those requirements, but subtitled films aren’t also captioned, and no commercials, promos, or interstitials are, either.)
- Even the larger Canadian broadcasters, with their 90% captioning requirements, have the exemption of overnight programming, not to mention the 10% of programming that simply doesn’t need to be captioned.
- Other broadcasters, with their lower requirements, have even larger quantities of uncaptioned programming available. That programming could be selected for Webcasting. Because it falls into the proportion of TV programming that doesn’t need to be captioned, it would meet the original TV requirements and would comply with the proposed WCAG 2.0 even without captions.
- The Working Group’s warning that video should be presented so that viewers don’t have to watch the program and captions simultaneously would only be applicable if the video paused or froze with every caption. It’s possible with many kinds of recordings (VHS, DVD, extended captions in online video), but that simply is not the way people watch captions. Eye-movement studies by Jensema et al. show that even hearing viewers previously unaccustomed to captioning quickly adapt to reading captions. Most time is spent looking at captions, not struggling to balance the use of “a single sense to follow two or more things at the same time.”
Proposed captioning requirements
The WCAG Working Group held a meeting in Toronto in September 2003 at which some suggestions were made concerning WCAG 2.0’s captioning requirements. What follows is an update of those recommendations.
- It’s unrealistic to expect authors to caption all their videoclips right away.
Expertise is an issue. With no known training programs, it’s also hard to develop.
- Player complications are significant, and if the author provides video in multiple formats, then multiple incompatible caption formats will also be required.
- Captioning itself is difficult and captioning software makes things harder.
- Sending video out of house for captioning costs hundreds or thousands of dollars per program hour. For many small and medium-size Web publishers, that constitutes undue hardship right there.
- In some languages and countries, there is no tradition or practice of captioning at all, even on television.
- Transcription is not the way to make video accessible. The correct accessibility method is captioning.
A transcript is an artifact separate from the original, just as a talking book or large-print or Braille edition is separate from a printed book. But audiovisual media can and must carry their accessibility features with them.
- If an online video segment is provided with captions, a transcript can be added and offered later. A transcript must never be the sole method of accessibility for deaf and hard-of-hearing viewers. At best it can be supplemental.
A “combined” caption transcript plus audio-description script have been attempted only rarely – for only a tiny handful of known projects (including DigNubia). These combined transcripts can be manually assembled, but there is no automated method to combine those two sources – due, among other reasons, to a lack of interchange formats and the difficulty of autoconverting into valid, semantic XHTML.
Reuse of existing captions and descriptions
- Authors should be required to reuse any captions or descriptions that were created for the original program.
- That means all TV programming captioned or described for television must carry its accessibility over to the Web.
- It’s technically straightforward, inexpensive, and almost immediately achievable, even for audio description.
- It’s already being done now.
- For a video presentation that was closed- or open-captioned in another medium (e.g., television, video, DVD), retain and reuse the captions.
- Use a caption decoder and digitize a copy of the video with captions decoded (open or burned in). If necessary, produce an open-captioned submaster tape and digitize that.
- For presentations with bitmap captions (e.g., DVD or DVB originals), digitize a copy with the bitmaps open or burned in.
Same-language subtitles intended for second-language speakers are not captions and do not qualify as such.
- If an uncaptioned version is available immediately and a captioned version available only later, provide the captioned version as soon as reasonably possible after it becomes available.
- Captioning on a video may be optional or required (closed or open). Authors may provide separate captioned and uncaptioned streams or a single captioned stream.
- For an optional stream, give the visitor an accessible method of selecting captioned or uncaptioned video.
- For any stream with captions, provide a statement, in an accessible form, that the video is captioned.
Standards analogous to WCAG
The Web Content Accessibility Guidelines are not the only Web-accessibility standard, but all the others currently in place were derived from or are analogous to WCAG.
Section 508 of the Rehabilitation Act of 1973 applies to U.S. federal government departments (with few exceptions) and a small number of federally-funded agencies and organizations. §1194.22 states:
Equivalent alternatives for any multimedia presentation shall be synchronized with the presentation.... Captioning for the audio portion and audio description of visual information of multimedia presentations are considered equivalent alternatives.
Note that the 508 requirements do not permit the use of a separate transcript to comply:
This provision requires that when an audio portion of a multimedia production is captioned... the captioning must be synchronized with the audio. Synchronized captioning would be required so someone reading the captions could also watch the speaker and associate relevant body language with the speech.
(There is, nonetheless, an ambiguity in the 508 specification, since it says that these multimedia requirements are “consistent” with Checkpoint 1.4 of WCAG 1.0, which permits transcripts. Still, that note could be viewed as incidental and purely informative.)
Audio-only files don’t have to be captioned “because it is not multimedia. However, since audio is a non-text element, a text equivalent, such as a transcript, must be available.”
The German BITV regulations require:
Für jede zeitgesteuerte Multimedia-Präsentation (insbesondere Film oder Animation) sind äquivalente Alternativen (z.B. Untertitel oder Audiobeschreibungen der Videospur) mit der Präsentation zu synchronisieren.
(English: “For each time-based multimedia presentation [in particular, film or animation], equivalent alternatives [e.g. subtitles or audio descriptions of the video source] are to be synchronized with the presentation.”)
EU and Italy
The European Union and Italy have adopted WAI standards directly.