10 Steps to Accessible PDF
This article appeared in Issue 130 of InDesign Magazine.When I mention the term “accessibility” to most InDesign users, I typically get a blank stare. As designers, we’re visual people and we focus on the appearance of what we create because that’s what we’ve all been taught to do. But appearance isn’t everything. We also need to consider how people with low-vision, blindness, or mobility impairments can consume our content. Those folks often use assistive software commonly referred to as “screen readers” that read content (usually a PDF or website) aloud. In an era where even household appliances can understand and respond to human speech, this may sound simple enough. But unfortunately, not all PDF files are created equal. Having actual live text in the document is a very important factor in making a file readable, but there are several other factors that also come into play.
. . . .
This article is for members only. To continue reading, please sign in, or sign up for a membership today. Thanks for supporting CreativePro! CreativePro membership keeps you up-to-date with the technology, solutions, and resources to strengthen your professional development.
BECOME A MEMBER
For just $6.50/month (billed annually), you’ll get access to valuable benefits, including:
 |
Sadly, the 2020 version broke many, many things for Accessible PDF (Article panel, faulty empty tags for hyperlinks, hyperlinks not read, Alt text warning for hyphens and forced line breaks in TOC, wrong nesting…). It is/was a big mess.
The latest pre-release version, not yet available for the public, fixes things – but not all.
That said: QuarkXPress 2018/19 can create a PAC3 compliant Accessible PDF/A-2u in one go!
For PDF’s from InDesign we have to fix always the same standard thing, and for PAC3 PDF/A(2u) you have to run a lot of fixes after that (tables, TOC link fixes, artefacting)
Users here needing to create Accessible PDF are all ‘downgraded’ to 2019 as the mess that 2020 made (and still does!) was so bad…
I agree Frans, 2020 somehow broke several things many of which appear to be fixed in an upcoming released. I don’t fully understand how that happens. And the added accessibility feature that seemed like a great improvement (yet not even promoted in the release), but was broken out of the chute. Possibly had we known to test the feature we could have caught it.
Still, I continue to strive to get the accessible output from InDesign to improve. I haven’t used Quark in quite a few decades but it’s interesting that they’ve put the effort into getting accessibility output to be quite good.
By the way, no idea if this is a language thing, using Dutch Acrobat, but when I create a new tag in the structure, not all tags are show in the pop up list to choose, just 4 standard ones. I have to create a paragraph tag then right click for Properties, and then I see all tags to set.
Any idea why I do not see this complete list in the pop up when creating a new tag?
And, did you also notice: if we take a PDF and define tags (for a non tagged PDF) with the reading order panel, as soon as you have only one thing anywhere made an artifact, after that you can no longer create other tags? This is a strange bug, it means that Artifacting always have to be done at the end. An Acrobat bug…
Hi Frans, I’m not seeing the limited list of tags when creating a new tag in the English US version. Initially it only shows about 5 items but if you scroll, the rest show up. Not sure if it has to do with your dutch version or not.
The artifacting problem is a long standing one that I’ve discovered how to fix if/when it happens. I’m happy to show you offline if you’d like. To avoid this altogether, I’ve just stopped artifacting using the Reading Order panel and instead do it using the Content pane. My workflow is that I tag everything using the Reading Order panel, then go back and artifact everything that I don’t want read in the Content pane. Because the elements I haven’t tagged don’t have a container, they’re easy to find in the Content pane.
No, I can’t scroll in the initial list, only when choosing Properties. It is only in the Dutch version. Strange huh? Yes, I use the Content pane to artifact, the bug I see is when using the Reading order panel indeed: drag, choose Artefact, then try dragging some text to tag: nothing happens. bad bug.
Anyway, I’ve created a special Acrobat Preflight Profile Action that fixes a lot of things in one go: annotated links failure, Artefact non-tagged items, Set PDF/A identifier etc. Works like a charm on, of course, PDF’s that have been properly prepared in InDesign ;-). That preflight action catches many thing for PAC3 check.
Still, all Tables are considered Anchored items, within a enclosed P-tag, and PAC3 shows a warning for that (the yellow sign), same for Anchored images. To be really, really following guide lines those have to be dragged out of the P-tag.
QuarkXpress all does this fine.
Yeah, I get a lot of calls on this issue and as I said, I’ve figured out how to fix it when it happens but this has been happening for a few years now. Very frustrating.
It’s refreshing to see that Quark has put some serious energy into their Accessible PDF output which is awesome. The problem is, I haven’t seen, been sent, or encountered anyone using Quark for 10 years. The files I work with already exist and don’t need to be built from scratch. So even if Quark is that good, it’s a no-go for me.
What we really need is for the InDesign team to take accessibility more seriously and get some of these issues resolved. And they can be resolved, no question. I was able to get them to add several features but that was quite a few versions ago now. Time for more accessibility features!
I actually talk about that artifacting bug in my LinkedIn Learning course here https://bit.ly/2Owqyco.
Well, here is another bug: when you have an auto numbered Paragraph, since 2020 you can change it from Automatic (that will tag thing as list and list items) to lets say just P or more likely a H3 or H2 etc. when it is more a auto-numbered heading.
That will however NOT set the tag to H3 P or H2 or whatever, but instead will give the tag a name that is the same as the name of the Paragraph style!
I listed this today as bug.
This is the latest RC Pre release version from yesterday…
Well, it seems that some documents do this. I tried creating a document from scratch, and there it works oke for numbered lists. But I have an example where it doesn’t. Roundtripping to IDML did not solve it there…
If I’ve followed your instructions “To continue reading, please log in above, or sign up for a premium membership today!”, then why don’t I see the article?! Feel mislead. Is the InDesign Secrets’ real s objective a paid subscription? If that’s the case, be honest and upfront. Disappointed with your ambiguity.
Lee, I’m sorry about the confusion. You appear to have a free membership, not a premium membership. Articles from InDesign Magazine (as well as other premium content) is only available to paid members. If you believe you’re supposed to have a premium membership, please contact [email protected] ! Thank you for being part of the InDesignSecrets community.
I feel the same.
Just one question: How do you set it to be pdf/UA so that it will pass Pac3?
Ah, you will have to fix a lot of things in Acrobat: annotate links (like TOC), artefact everything that is not tagged, set the PDF A identifier bit, drag anchored frames out of their P-tag, and a few things more ;-)
Oh, and of course convert to PDF-A(2u) with preflight in Acrobat ;-)
Hi F van der Geest, If your are talking about giving TOCs Alt text, this is now possible in InDesign 2020. As for artifacting objects, this has been a thing for some time with the object export option. You can also use the Articles panel where everything you don’t include gets artefacted… or placing the thing you want to artefact on a master page.
Hi Britta are you talking about PDF / UA identifier xmp file? in Acrobat DC go to files> properties> Additional metadata> select Advanced> and press the Append button. find the XML file and press OK.
Download the XML file at Kenny Moors website taggedpdf.com here: https://taggedpdf.com/508-pdf-help-center/pdfua-identifier-missing/
correction: XMP-file not XML
No, an auto generated TOC still has this problem in PAC3, you still have to run the annotations fix from Acrobats Preflight tools. You can however provide a Alt text now for hyperlinks, the implementation however is flawed and hopefully fixed in the coming bugfix update. For the missing PDF A/U identifier, you can also run the simple fix in Acrobat Pro from the preflight tools, no need for complicated XMP files or things like that ;-)
Jakob Petersen: No, an auto generated TOC still has this problem in PAC3, you still have to run the annotations fix from Acrobats Preflight tools. You can however provide a Alt text now for hyperlinks, the implementation however is flawed and hopefully fixed in the coming bugfix update. For the missing PDF A/U identifier, you can also run the simple fix in Acrobat Pro from the preflight tools, no need for complicated XMP files or things like that ;-)
I’m wondering about the plug in called MadeToTag. I’ve used the demo and had a lot of success exporting PDFs that pass PAC2021 on the first try. Has anyone else tried it, found any shortcomings?
It’s awesome. It essentially overcomes all of the limitations (or at least most of them) that exist in InDesign natively. I use it all the time.