Internet accessibility standard and technique

Improving accessibility for websites

This paper is written to explain a technique to improve accessibility for websites. Accessibility for a website is important. A lot of people today can’t visit the most internet webpage’s because they are not accessible for them. There are a lot of techniques to improve the accessibility, this letter will explain a new ‘invented’ technique to improve the accessibility for blind persons and people with bad eyes.
This technique is developed in 1999-2000 in Belgium by Pascal Vyncke at the SeniorenNet Research Department (SNRD), the technique is used and produced for the website SeniorenNet and is released to the public in 2001.

This technique is made in cooperation with the Belgium Blind League and with the cooperation of real blind people to work out and test the new accessibility technique. There has been also a cooperation with blind associations in The Netherlands and the United States of America and with an professor doctor in the communication science at the university KUL, Belgium.

Introduction
There are too many websites on the Internet that a blind person cannot visit. A blind person has the problem he can’t see anything. There are however techniques to help these people on a computer: Braille keyboards, Braille readers,… The problem is that these software and hardware can’t follow the very fast new techniques of website development and because of this blind people can’t read everything on the internet. It is also a problem for blind people to read long texts, it is very intensive to read Braille.
A third problem is that all the navigation information is also read by the Braille equipment, every time again the while menu, copyright information, ads,…
The global idea of this new technique to improve the accessibility is to read aloud the content of the website. In this way it is possible to only tell the real information and not all the unnecessary information like the menu, navigation, copyright,…

The idea
To make a website totally accessible to a blind person, a person with bad eyes or people that can’t read long texts, especially older people.
The best technique is to read aloud everything. In this way the person doesn’t have to feel the Braille on his Braille equipment for hours, but has to just listen to the text.
Almost nobody has problems to listen a long time to someone. In our evolution the last thousands of years, communications was for people very important. Listening and talking is for all the thousands of years the only way to communicate with each other. Only the last tens of years everybody has to read. It is naturally that not everyone can concentrate a long time to read and it is totally normal that some people have problems to read long texts or have problems to keep looking at a computer screen, especially if the person is older. We have however no problems to listen to someone. That is the reason why the best way to make it accessible is to read aloud the texts.
Blind people can hear the information, people with bad eyes doesn’t have to effort to read the texts and people with reading problems and also older people can also hear the information.

What you need
An important condition for the new developed technique is that it may not cost a lot of money and that there is no billion dollar equipment necessary.
The only need for the technique is a microphone; software to edit, record and save sound and someone to read aloud.

Protocol
To implement the technique in practice on a website it is necessary to make a sitemap and make a table of contents of the website for all the pages that has to be accessible.
The next step is to search a narrator. You need a computer with software that can record sound. You plug in the microphone and the narrator can start reading the texts.

Every time the narrator reads the whole page in one time. It is important not to make an apart sound file for every chapter on the page. Everything that is on the page and that is necessary to read, is read in one time, in one sound file.
If the reader says an error, he or she just stops to read and remain silent for about 10 seconds. Now the narrator can restart saying the last sentence where the error was in and reads the rest of the text.
In this way it is possible for the editor of the sound file to quickly find the places where the narrator has made a mistake and remove the wrong sentence.
If it is not possible for the narrator to read the whole text in one time, he or she can stop and read further a next time. The editor of the sound file then has to put together the two sound files. It is important that at the end of the process, there is exactly one file for one page.

It is important that the narrator respects the punctuations in the texts. The best results are obtained if the narrator waits 1 up to 2 seconds after every point, question mark and exclamation mark. To wait a half second up to 1 second after every comma or semicolon and to wait 3 seconds after a 3 dots (…). After a paragraph wait 3 up to 4 seconds.
These recommendations make it pleasant to listen and makes it the least tiring to listing.

Which software is used to edit or record the voice is not important, almost all the software can do the needed things.

It is important to use a good microphone and not one of a cheap store. Cheap microphones will not record the voice exactly as we can hear it and will force a distortion of the voice. Also have cheap microphones the problem to have a lot of background noise.
It is also important to record the voice in a silent room. Try to avoid as much as possible of background sounds and noises. The best is a soundproof room, but if it is not possible to rent a studio, try to find a room that is silent. Notice that a refrigerator, computer, television, air conditioner,… can make a lot of noise and can interfere with the signals of the microphone to boost the background noise extra.
Stop recording if a plane fly over or other sounds are recorded. Just let the narrator read again the sentences where the noise was into and let remove it by the editor of the sound file.

The best format to save the sound files into is MP3. This is a worldwide standard that can almost every computer, independent of the operating system, play to the person that wants to hear the sound file. Moreover, it is also a format that is very little and small to save on the web server.
If your company decides to spend enough web space for this project, the best is to save the sound files in a good MP3 format with good quality, but not too much. We recommend in our standard to use 80 Kbps, 32.000 Hz, Mono with CBR Constant Bit rate. This is not CD quality, but that is not what we need, it is enough for voice for this purpose.
If your company wants to spend as little as possible web space and bandwidth for this accessibility technique, you can use less quality.
We recommend to use then 16 Kbps, 11.000 Hz, Mono with CBR Constant Bit rate. This is not the best quality and will distort the voice, but will be for people enough to perfectly understand what is read aloud.

To publish the sounds on your website you have to use a figure that is placed on all your pages that are narrated and every time at the same place. We recommend it to place it in the right top corner of the webpage. For totally blind people it doesn’t matter where it is located, but for people that can still see something, it is the best and most accessible if it is located at the top right corner of your site.
It is very important to ALWAYS place the figure at exactly the same place at all your pages.
To be visible to the right people, use enough contrast in the image and it has to attract attention. (We do not say it has to be ugly!)

Our standard says that you always place an “ALT”-attribute in the <img> HTML tag in your website, with each time EXACTLY the same text, something like “to read the page, click here”. This text is read by a blind person so he always can find the button to hear the narrated text. Without this, the blind person is not capable to find the button to hear the voice.

Our protocol recommends to save all the original sound files to a good format without any loss of information. You can use the WAV or RAW format, but also MP3 on a high quality (320 Kbps, 48.000 Hz, Stereo, CBR Constant Bit rate). This because it can happen that your company decides to use another quality (to improve the quality from the low quality to the good quality).

Read only what is necessary. Do not narrate the advertisements, the menu, the copyright information or something else. Only that information that is interesting for the person that is going to listen to the voice.

The website also has to avoid any advertisements with sound in it, especially Flash banners, skyscrapers or IMU’s can play a sound. It is very frustrating and unpleasant to listen to the narrated text while there is on a regular base every time the same sound in the background from the advertisement.

The protocol that we have developed also asks to narrate a table of contents in the begin of the text if it is a text that is longer than 3500 characters (+- 1 A4 page) and there are more subtitles. In this way the listener knows what he is going to hear and can follow the whole text. It is also proven scientifically that this helps to remember the whole text and better understand the connections. It is not necessary, to read the table of contents again at the end of the file.

Recommended is to not protect your website to save the MP3 sound files to the hard disk of the person. If the person wants to save a copy on his computer, it has to be possible. In this way the blind person can listen to the text again in the future if he wants, without searching it again on the website.

Technically for the website it is not hard to implement the MP3 sound files. The website developer can just put all the files in one directory, for example called “Audio” and then upload all the MP3 files.
Every time a person listens to a sound file, this has to be downloaded. This will cost bandwidth, but this can be reduced by reducing the quality of the MP3 files from high quality to low quality.

If you have not enough resources to have a narrator to read aloud every text, you can always do it yourself. It has not to be a professional narrator. If someone listens to someone else, like a sister, a neighbor,… they are neither professional narrators, and still we can listen and understand them and even listen to them for hours. That is why it can be just yourself, an employee of your company or a student.

If it is not possible to find someone to narrate the texts but you want to still use this technique, you can use special software that can narrate digital texts into your language. The problem of this software is that the pronunciation is very bad and it is hard to listen to. But it is better than nothing. You can use this software to generate the MP3 files and upload them to the website. You can find such software by searching software categorized as “text to speech”.
Note that the best is to find a human narrator, only if it is not possible, use such “text to speech” software.

The technique to read aloud a text for blind people already existed on the moment that our team developed this technique. But the difference is that no one implemented this for the web. All the narrated software was for using on the computer and in normal software, never on the Internet. Our team has made this technique, a standard, for implementing this on the Internet, which makes it accessible for millions of people.

About the developer
Pascal Vyncke lives in Antwerp, Belgium (Europe). He is the webmaster of www.SeniorenNet.be and the head of the SeniorenNet Research Department (SNRD). His work is used all over the world for better accessibility for websites on the Internet.
He won up to now already awards for his work in the computer science and Internet.
UPDATE: Since recently he is also writer of a bestseller for Internet learning for seniors in Belgium , called “Internet na 50” (ISBN 90-209-5622-1, 496 pages), publishing company is Lannoo.

UPDATE:
Use of the accessibility technique/protocol
In the years after the technique is released, a lot of websites used the technique. Sites in Europe, but also in the United States of America and even in other countries implemented the technique in there website or at least the most important part of there site.
Hundred thousands of people could use the more accessibility websites up to now, and still counting.
In the years after publication the technique, it also got a lot of press attention. Especially in Belgium and The Netherlands was it national television news, national newspapers, magazines and even the radio, but also press and websites all over the world wrote about the technique for more accessibility for websites.