IPEDIS14 place Georges Pompidou
PDF files (Portable Document Format) have a lot of accessibility problems because they cannot usually be interpreted by speech synthesizers. For several years, the PDF creator, Adobe, has made it possible for XML markup to be used to give the information on PDF files, a hidden structure that makes them accessible to speech synthesizers.
When a document is put together, certain functions in desktop publishing packages like InDesign or Quark Xpress should be used (inserting alternative texts for images, using heading levels to structure the texts). After the PDF is created using Acrobat Distiller, XML markup should be incorporated (the use of tags to indicate titles, images or alternative texts, tables, lists with bullet points, etc.).
The accessibility criteria should be adopted in the source document before it is created with Acrobat Distiller as this reduces the size of the PDF and the time needed to create the markup.
PDF accessibility is needed for the visually impaired with the use of a specific software or equipment (speech synthesizers, Braille, etc.).
A speech synthesizer can only read text, but cannot read images containing text, even if the content is simple. This is the case for PDF communication files made using Quark Xpress or InDesign. These complex PDF files are not accessible using a speech synthesizer unless they have been edited first.
PDF files were not originally designed to be read using computers, but to be printed. Today, companies increasingly use PDF files to exchange and send out digital information in a more professional manner (presentation, security and data protection). Ten years ago, PDF files were not accessible for the visually impaired when they were using tools such as screen readers.
About five years ago, with the growing use of PDF files, Adobe created a technique to allow PDF files to be made accessible and to improve their structure.
You need to make sure that a PDF file acts like a classic HTML page in relation to the speech synthesizer, without its visual appearance changing.
This is where XML markup comes in.
You need to use this markup language, which is similar to HTML, to make a PDF file accessible. The goal is to do it while meeting accessibility standards PDF/UA (recognised ISO 14289-1:2012).
The XML markup is hidden so that users and ordinary readers cannot see it in the PDF.
To begin with, you must force the reading order of the PDF file so that a Jaws-type speech synthesizer reads it, in a similar manner to the way that one would naturally read, starting with the footer after selecting the corresponding page number.
You must then assess the content of the PDF itself. Examine and find alternatives to the images, and tag the changes in language that are needed so that the speech synthesizer can read them correctly, as well as any acronyms or abbreviations that are found in the file.
Click on the titles below to learn more :
In order to make your PDF file more accessible, it is important to use your word-processing or desktop publishing software correctly (Word, Open Office, InDesign, etc.).
You should begin by establishing the structure of your file (heading levels, sections, chapters, footers in a specific style, etc.).
If the text is to be set out over several columns, you should not create the columns by using the tab key on your keyboard; instead, you should use the designated function on your word-processor.
To create an accessible PDF, you need to establish a logical reading order (order of the paragraphs, headers, footers) in a structured manner. Do not forget to use the right XML markup with Acrobat.
If you are making a list of important points, use bullet points as they will be correctly converted when your document is turned into an accessible PDF. Do not make paragraphs by using several carriage returns.
If you have photos, maps, diagrams or images (non-textual elements that contain information), they must be marked with an XML <figure> tag.
When you use links, a clearly explained title should be provided.
If you have complex forms in your PDF file, you should respect certain rules to make sure that your file is accessible: the visually impaired should be used to use the fields, so they must have descriptions. In an accessible PDF file, a visually impaired person will use the tab key to navigate, to make sure that it works in a logical order. You must use Acrobat Professional for this and add the appropriate tags after creating your PDF file.
If you have tables in your PDF file, you must create tags for the headers, columns and lines in order to make them accessible. You should also avoid using tables to create text in columns or to create images.
In order to make the navigation of your PDF file easier, you can create bookmarks using Acrobat Professional.
To make PDF files accessible, you should never need to fully rewrite the information hidden in the PDF. You simply need to add XML tags to the content.
The only case where a part of the text will need to be rewritten for accessibility is when acronyms or abbreviations have been used. This procedure allows you to limit the number of mistakes arising from the PDF encoding. It is also possible that a “lang” attribute is forgotten, but this type of mistake is unusual.
It is not currently possible to automatically render a complex PDF file accessible as one must continually consider which contents of the file are to be made accessible and above all the best way of doing this so that it can be read as easily as possible by the visually impaired.
These days, it is possible to design a PDF with different software, but to make it accessible, this software must allow you to structure the PDF document with XML markup. If a tree structure already exists, some PDF creators are able to transfer the structure. This will allow speech synthesizers to access the contents. If you are using word-processing software for example, you will need to use heading levels to create your PDF file, but that is not sufficient to make the PDF accessible. To do it properly, you will also need to use Adobe Acrobat Professional to insert XML markup into any PDF which is not tagged. This is a more difficult process because you will need to know XML markup language and have a good understanding of it. This process can also be very time-consuming.
There are different tools which can be used to convert a document into PDF format. To create accessible PDFs, it is best to use Adobe Acrobat and to convert them using Acrobat Distiller.
If the PDF file includes an image of some sort (diagram, pie-chart or histogram), the first step is to decide :
If the image is an informative part of your PDF file, should you simply add an alternative long or short description, or should you put the image in the form of a table or list (if the image is of a pie-chart or histogram, as these are the only elements that a synthesizer would be able to read)? If you choose a short description, you can use the “alt” attribute as with HTML, as it is level A on the WCAG 2.0 international “success criteria 1.1.1 [A]” guidelines.
Some Jaws-type speech synthesizers can change the language of an HTML page of an accessible PDF document, in other words, the speech synthesizers automatically change language when a specific word requires it.
An example of a change of language for internet accessibility would be that when you read the French word “Bonjour“, the speech synthesizer should read the word correctly in French and not in English.
To change the language in this particular case, you would use the “lang = fr” attribute, as with HTML. A change like this is level AA of the WCAG 2.0 “success criteria 3.1.2 [AA]”.
Accessibility sometimes requires the meaning of certain acronyms or abbreviations to be indicated.
The tag does not exist in XML language, and so for PDF accessibility you must create an alternative with the exact word.
For example, for the abbreviation “MJ“, the word “megajoules” must be rewritten using a tag. This is a WCAG 2.0 level AAA success criterion (success criteria 3.1.4 [AAA]).
N.B It is sufficient to respect level AA criteria to make a PDF accessible.
To evaluate the accessibility of a PDF, you can use the accessibility checker that comes with Adobe Reader or Adobe Acrobat. Adobe Reader can indicate any serious errors in the structure of the document. However, the alternatives behind the images and the direction of reading cannot be checked, and the accessibility of tables, links, images, form fields, and notes must be done manually. You should check these elements using a screen reader such as Jaws.
To make the most of accessible PDF documents, you should use Jaws (version 6 or above) and have Adobe Reader (version 7 or above). If the document has been correctly marked up, it should read well with Jaws. In Jaws 7 and above the heading levels can be read. In Acrobat, you need a page by page reading method, otherwise it will take too long for each page to load up and this can lead to problems when using Jaws. You can navigate from one link to another using the tab button and activate them with Jaws 8 or above. With tables that have good markup for PDF accessibility, you can navigate from the cells of a table as in an HTML document and you can use the same keyboard shortcuts as Jaws to read the tables. For complex accessible PDFs made using Quark Xpress or InDesign, Jaws uses the order of reading defined by the XML tags that you have added. From version 7 upwards, Jaws can recognize the language defined by the document and can also change languages as indicated by the XML markup.
You can find below methodology to tag your 'Word' 'PDF', and 'InDesign' documents