Create accessible PDFs

The Challenge
People who use screenreading software rely on it to "read" text out loud to them. However, many documents posted on the Web are either completely inaccessible or just very frustrating for screenreader users.

A common and vexing problem is when documents are scanned and posted on the Web as a PDF. A scanned document is actually a photographic image, which the screenreader can't read. Essentially the scanned document is a blank slate.

Another problem occurs when a Word document has been converted to PDF but does not contain structure, such as headings, alternative text for images, and lists. This formal document structure enables a screenreader user to do such things as scan the document, navigate to certain sections, interpret images, and access Web links. Without document structure, a screenreader user may be faced with listening to many pages of text to find one pertinent paragraph and may not even be able to access much of the material, including images and charts. The key to creating an accessible PDF is to take the appropriate steps to build in structure when creating the source document and before converting it into a PDF file.

Solution

  • Never post a scanned file. These are images; screenreaders can't access them. If you only have the scanned file and no source document, use Adobe Acrobat's OCR Text Recognition feature to clean up the document and convert it into a PDF with text that's accessible to screenreaders (although it won't contain document structure).
  • Ensure that Word files are created with embedded structure before they are posted on the Web or converted to PDF. Such formal structure benefits everyone, not only screenreader users. Structured documents are more searchable and navigable; upon conversion to PDF, structured documents automatically generate bookmarks, which can serve as a table of contents.
  • Question whether it is necessary to post a PDF or Word document, or if an HTML page (as long as it's accessible!) is sufficient.