Creating accessible PDF documents

Find out how to create accessible PDF documents suitable for publishing on our website, and check they work as intended.

Creating accessible PDFs

To create an accessible PDF, follow the process of creating an accessible document using Word, InDesign or Excel and choose the appropriate export settings.

See full details on:

• creating an accessible PDF from Word
• creating an accessible PDF from InDesign
• creating an accessible PDF from Excel.

When creating an accessible PDF from Word you may need to set table headers and the document title using Adobe Acrobat.

This can be done during the checking process.

Checking PDFs for accessibility

Once created, all PDFs need to be checked using Adobe Acrobat.

Checking colour contrast

Checking for colour contrast is useful because it alerts you to instances where text is not readable against its background (for instance, where a user is colour-blind).

This happens when text is not different enough in colour to the background but also if text isn’t large enough. Large text is defined as 18 point (typically 24px) or 14 point (typically 18.66px) for bold text.

It is recommended to check colour contrast first as this will need to ammended in the parent document (word, indesign etc). If you check and fix other elements first, you may have to repeat these steps if any colours need changing.

To check a PDF for colour contrast, you want the hexadecimal or RGB values of the colours used in the document.

If these are available then the online tool WebAIM colour contrast checker is an easy to use tool.

Copy the text colour code into the foreground colour box and the background colour into the background colour box.

If the colours result in a green pass next to WCAG AA then the colours are good to use.

If you don’t have the hexadecimal or RGB codes, you can use the Paciello group Colour Contrast Analyser.

This tool has a colour dropper which can be used to sample the colours used in the document. The text colour should be input to the foreground colour box and the background colour should be input to the background colour box.

If the colours result in a green pass next to WCAG AA then the colours are good to use.

Paciello group Colour Contrast Analyser can be installed through a request to ITS.

Setting up Acrobat

Make sure you have downloaded the latest version of Acrobat Pro DC to access all the accessibility options.

Follow these steps to set up Acrobat ready to check documents:

• select Tools in the taskbar at the top of the screen
• scroll down to the Protect and Standardize section
• drag the icon for accessibility over to the toolbox pane on the right hand side of the screen. This will add accessibility tools to your toolbox.

Skip this if you have already set up Acrobat on your computer.

Using Acrobat to check a document

To check your document for accessibility:

• select the Accessibility tab from the right hand tool box or select View > Tools > Accessibility > Open. This will bring up the Accessibility menu
• click Accessibility Check to open the Accessibility checker options panel
• ensure the Create accessibility report box is not ticked
• the report will appear on the left of the screen and highlight any issues with the document, any areas with issues will be highlighted in bold and have a bracketed number after them
• right-clicking on an issue will open a dialogue box with guidance to fix the issue

Addressing reading order issues

A common issue with PDFs is the reading order.

To check this, follow these steps:

• With the accessibility tools selected click the fifth icon down on the left hand side of Acrobat (it is shaped like a Z with four squares behind it). This will open the Order panel
• A list of the pages in your document will be provided. The numbered elements show the order that a screen reader will read the elements. If the order is not correct, or missing entirely then you will need to read the next section to correct this

Fixing reading order issues:

• select Reading Order from the right hand Accessibility menu open the Reading Order panel
• scroll through your PDF, one page at a time, click clear page structure on every page. It is easier to remove all structure and start again than to ammend existing structure
• highlight the first text, or area of the screen, that you would wish the user to read
• tag this area using one of the options from the Reading Order panel
• continue through the document. The order that you tag items will set the order in which they are read by a screen reader and also provide them with an identifier, e.g. heading, figure, text, form field

Tables will need to be fixed using the additional technique described below

Tables

Follow the steps in the previous section to set up and perform the initial checks on the document.

Tables need a further check for accessibility. They can be difficult to get right so it is recommended that they are formatted correctly before creating your PDF document.

In particular you need to bind the header of a column to the column data. This allows most screen readers to identify each cell with the title of the column it relates to.

Here’s what to do:

• in the Accessibility menu, select Reading Order to bring up the Reading Order panel
• select the table you wish to check
• in the Reading Order panel, click Table Editor. This will highlight the table and indicate headers (TH) and data cells (TD)
• If any cells are incorrectly tagged this can be addressed easily. Right-click the cell and select Table Cell properties. Set the Type to either header or data. If the cell is a header, set the Scope to either Row or Column
• left-click in the left-most header cell and drag to highlight all of the header cells. Right-click on the selection and choose Auto-generate header cell IDs...
• right-click any header cell and select Table Cell properties. Check that the cell now has an ID in the Attributes box. Close the Table Cell Properties panel
• left-click the top data cell in a column and drag down to highlight all of the cells in the column
• right-click the selection and select Table Cell properties to bring up the Table Cell Properties panel again
• in the Attributes box, click the green + next to Associated Header Cell IDs. This will provide a pop-up window with the option to assign a header to a column. Select the ID that relates to the highlighed column. If it is not obvious at first, it can be worked out using the RXCY part of the ID which relates to the row and column location of the header cell
• This last step must be repeated for each column in the table. If the header is blank, the column of data will still need to be bound to the blank header cell (this will still have its own ID). Failure to do this can cause columns to be linked incorrectly.

Using a screen reader

Checking your document with a screen reader allows you to identify any issues that could affect the experience of a visually impaired user.

The document should be checked with more than one screen reader (ideally Voice Over for Apple devices and NVDA for Windows). You should:

• open the PDF using Adobe Acrobat
• open the screen reader
• attempt to navigate the document. Use both the mouse and keyboard commands (Ctrl+Alt+Arrows on Mac, just the arrow keys on Windows).

To navigate tables an extra step is required for Voice Over. The command Ctrl+Alt+Shift+down arrow is required to enter the table. The command Ctrl+Alt+Shift+up arrow is required to leave the table.

If you have created your document using our accessibility how-to guides, there should be no issues when checking your PDFs with a screen reader and all elements should be identified and read to the user.

Apple VoiceOver is freely available on all Apple devices. However, NVDA screen reader will need to be installed on a Windows machine. This can be done via a request to ITS.

Apple have provided a complete guide to Voice Over which includes a list of all the keyboard commands required to use the screen reader. WebAIM have produced a guide to using NVDA.

Note – NVDA screen reader has these known bugs:

• When reading Word documents, or PDFs created using Word, it will read out any blank lines with the word "blank". This is an issue known to NVDA and they are working to address it.
• NVDA also ignores any shapes created using Word when reading .doc or .docx files. This does not happen when reading PDFs created from Word files.