Accessibility for the deaf, especially YouTube - Videos need accessible captions to support people with hearing impairments.

Written By: Peter Abrahams
Content Copyright © 2010 Bloor. All Rights Reserved.

In the early history of accessibility of the Internet, video and audio were really not an issue, little existed and what did exist was not vital. At that time accessibility concentrated on the needs of people with vision impairments and, to a slightly smaller degree, to those with muscular-skeletal issues. The requirement was for websites to be compatible with assistive technologies such as screen readers, screen magnifiers and various alternatives to keyboard and mouse input.

At this stage people with hearing impairments did not have accessibility issues as they could read the text on the screen. This is not the whole story, as British Sign Language (BSL) is the first language for deaf people in Britain and written text is a second language; this means that deaf people would prefer BSL. I discussed the issue of BSL on websites in ‘Should websites include sign language‘. My major conclusion was that it would not be practical to convert all content into BSL and therefore web site owners would have to decide what content was important enought to convert into BSL.

But today, with the increase use of audio—and especially video—the deaf community is finding that increasing amounts of content is inaccessible. What is needed is captioning. The vast majority of video does not include captioning at the moment.

Adding captioning has the obvious benefit of making the content availalbe to the deaf and hard of hearing; but, like many other assistive technolgies, it has a variety of other benefits:

  • It can be used when it is difficult or inappropriate to listen.
  • The text is crawled by search engines so captioned videos should get more and better hits.
  • YouTube has a beta function to translate captions into other languages so the videos become accessible to a much larger audience.
  • The caption could be an actual translation where having a high quality translation can be justified.
  • The caption could be a video description so enabling users with vision impairments to better follow the action.
  • Captions can be a learning tool as the viewer can relate the spoken word and the written word.

You Tube is the main purveyor of such content and does include the ability to add captions; but most user-created content does not include captions. This is hardly surprising because most creators are not even aware of the issue and, even if they are, may not be motivated to go to the extra effort to create captions.

This may be inevitable for privately created content but is not acceptable for content created for inclusion on commercial websites that should be accessible to all. The difficulty is that creating captions for YouTube clips has been hard and costly. Two recent innovations have made it more feasible and are discussed here:

YouTube now have a beta test version that transcribes the audio in real time using Google speech recognition. I have tried this on a few audios and at first sight it is amazing how good it is, but unfortunately at second view you become aware of the mistakes it makes. The problems are that it has to work on any voice; some are clearer than others, also it has to work in real time so the level of processing available is limited and, finally, there is no correction facility so the system does not learn. I believe it will be some years yet before this technology can provide an adequate solution. At present it is what I describe as a band-aid facility; if there is no captioning and a person with a hearing impairment wants to know what the clip is about the transcription will give them a good clue. This is similar to me asking for an automatic translation of a web page, which is inaccessible to me because it is in a language I do not understand, reading the translation will give me a good indication if it is of real interest to me but I know there will be errors in the translation, some of which may be serious, so I would not quote from it without having it translated by a person.

On the other hand Videocritter is a free tool that enables captioning to be created for YouTube. It was written by Ken Meyering as a college class project but is of a standard that you would expect from a commercial product. The process is very simple:

  • Log on to VideoCritter.
  • Connect to the video on YouTube.
  • You then have controls to listen to a portion of the video and immediately type the caption, then listen to some more and type some more.
  • There are also functions to review and correct.
  • When the caption file is complete you upload it to YouTube

I have tried it and it is very easy—you just need a little practice to decide when to stop listening and to start typing.

There are other similar tools available but I have not had a chance to do an in depth comparison.

All of them require you to type the text so I have two requests for extra functionality to reduce the need to type:

  • If there is a pre-prepared script then it should be possible to upload this and then use the tools to sync it with the video.
  • The YouTube Beta transcribe function should produce a caption file that could then be edited and corrected using a tool.

One final issue that needs to be resolved is that the standard YouTube Player is not fully accessible. Easy YouTube, which provided a much more accessible player, does not support closed captions. It really is time that YouTube recognised the importance of accessibility and provided a comprehensive solution.

Given the need to be accessible and the other benefits that accrue from captioning I would strongly urge anyone, and especially commercial organisations, to start using these tools on the videos on their websites. Further, I would encourage users of the websites to complain when captioning is not included.