Level Access Accessibility Consultant James Thompson hosted the fourth of an eight-part webinar series on PDF Accessibility Basics last Thursday: Non-Text Elements.
There were a few questions during the webinar that we did not have time to address, so we’ve provided the answers in this post, as well as listing out the questions posed in the webinar that James also answered.
For more information, you can access the webinar slides, transcript, and recorded presentation here: Non-Text Elements Webinar Resources.
Q: What would be best source for mathematical formulas or equations in PDF?
A: If you have the fonts that would allow the formula to be read correctly, ensure they are embedded. Otherwise I would tag it as a figure and write out how the formula would be read in the Alternate Text.
Q: Clicking “Background” in reading order makes it an artifact? What about using “null” as alt-text?
A: Clicking Background in the Touch Up Reading Order panel will make that element an artifact. Null is used as a description for an empty value (In HTML a null alt tag would be alt=””) In a PDF instead of Figure’s having blank Alternate Text they should just be made an Artifact.
Q: Would it be good practice to tag the “figure 1” caption as an artifact and then properly alt text the image?
A: Recognized text should be a higher priority than a figure’s Alternate Text. So given the choice the text should be tagged and the figure should be made an artifact.
Q: If a logo is in the header should it be identified or artifacted?
A: The first instance of the logo should be identified and all subsequent instances should be made artifacts.
Q: What is the best tag for leader dots in a TOC. reader is reading dot dot dot
A: There is no one way of dealing with them. I prefer to artifact them as they only serve visual users in associating the name with the page number. Another option is to tag the name and dots separate from the page number so the dots can be skipped when the user moves to the next element.
Q: Can you show how to fix a character encoding fail in pdf?
A: Your best bet is to go back to the original document and re-export it to PDF. If you can’t do that you can go to the Print Production tool and open the Preflight option which may allow you to embed the font. Adobe has a good explanation of how to use Preflight on their site (https://helpx.adobe.com/acrobat/using/analyzing-documents-preflight-tool-acrobat.html)
Q: Is it possible to change the unicode name
A: The unicode name is set in the standard.
Q: If you use windings icon, how would you tag it?
A: I would treat it like an image and make it an artifact if it is decorative and tag it as a figure with descriptive alternate text if informative.
Q: Is it possible to use a wing dings symbol and artifact it?
A: Yes, you can make them artifacts.
Q: When I run the accessibility checker after adding a checkbox to a pdf and if the checkbox is checked I receive a character encoding fail. How do you fix it if it is built into Adobe?
A: If you have run through the preflight fixes and the encoding is still not fixed, listen to the checkbox with a screen reader to determine if it is reading correctly. If the checkbox reads correctly but still has the encoding error then you should be fine as the main goal is to have it read correctly. Otherwise, if it is not reading correctly, go back to the original document to remove and re-add it with a different font. If that is not possible try using the Acrobat text editor to remove the element and add a new font.
Q: when selecting an image to change to an artifact, there’s a selecting of Artifact Type: Page – Pagination – or Layout. What should be selected? Also, there’s an option to Attach to Side(s). What that’s for?
A: For the purposes of accessibility leaving the options as default is the standard way of creating artifacts. For designers they allow for the specification of location to be able to find the artifacts later.
Q: Adobe does not let you change the font it is automatically set to Adobe Pi and grayed out. How would you change the font?
A: Adobe Pi is used when Acrobat doesn’t have the font you are using (usually with checkboxes). You can change the fallback font in the preferences which is detailed in this article: https://helpx.adobe.com/acrobat/using/default-font-pdf-editing-add-text.html
Q: When doing the accessibility checks on a document with a lot if images, including word marks (text images), the OCR process led to making the word marks into unreadable images. In looking at the Adobe Acrobat documentation, there is an option for “Searchable Image (exact)” that is supposed to let images stay in tact. However, even when running OCR in Searchable Image (Exact), the images with words in our document (logos etc.) were modified so that they were unreadable. Two questions:1) How can I still run OCR in the accessibility check process without modifying images?2) How can the default settings for accessibility check for OCR be modified so that the default setting doesn’t modify images?
A: I typically use the ClearScan option instead of Searchable Image (Exact) but I don’t know of a way in Acrobat DC to ignore specific areas of the page. Instead of doing the text recognition of the whole file or page you can go use the Find First Suspect function and run through the document skipping and correcting characters as necessary.
Q: I’m working on a document that has images of SQL statements. literally, just the sql code. We’ve been entering the entire SQL statement in the alt text, alerting the user in the beginning of the alt text that it is an example of the statement needed. Do you have any suggestions for images of text? It’s sometimes a frustrating thing to deal with.
A: Since SQL consists of alphanumeric characters you should be able to run the built-in OCR to have it read as text. https://helpx.adobe.com/acrobat/using/edit-scanned-pdfs.html