PLUS – 6 ways to beat the z-order problem in the first place!
It’s the content reordering nightmare. You are manually tagging a PDF in Adobe Acrobat. All of a sudden stuff starts disappearing from the page.
Professionals call it the “z-order problem”, but it’s known to most users simply by the howl of pain it so often provokes.
Many ask “is there a way to avoid it”. There are a few key mitigation strategies that can really help reduce the need to go to Adobe Acrobat’s dreaded Order panel.
Why does it happen?
What’s going on is that Acrobat assumes you want to group all the objects you select – of any type – into a single tag. It’s not smart enough to say (if applications could talk): “Hey, my user selected a big block of text that looks like a paragraph, and he’s going for a <P> tag. Maybe I should ignore the fact that he’s also selected a large image with that paragraph!”
The current-generation Acrobat Professional XI software isn’t great at this sort of question. Accordingly, to minimize the z-order problem when manually tagging with Acrobat it’s important to adopt three key strategies. Other software may have similar features – check your documentation!
Select what you want. Unselect what you don’t
The only way to conclusively avoid the z-order problem when tagging by hand using Adobe Acrobat is to be sure that you select exactly and only the items you want tagged. This isn’t always easy. You must avoid non-text objects when you want to add text tags such as <P> or <H2>. Likewise, you must exclude unassociated objects when you select content (text or graphics) that you want grouped in a single tag such as a <Figure>.
Let’s get to the tips!
Tip 1. Get familiar with Acrobat’s selection tools
In order to pay careful attention to your selections you need to be able to see what you are doing! In the Tags panel, be sure that “Highlight Content” is selected. You’ll then be able to rely on the faint blue boxes that surround objects you’ve selected. Verifying this setting is the very first thing I do before I even think of touching a tag. Frankly, I don’t even know they make it possible to turn this feature off.
If you are using the Touch Up Reading Order Tool you’ll get the selection highlights regardless of the Highlight Content setting. Due to its crudeness, however, I don’t encourage the use of this tool, but that’s a subject for another post.
Method 1: shift-click to add content to the selection
After using the mouse to “zone” a block of content, use shift-click to add content to your current selection.
Note that the mouse gets a plus sign to indicate you are in “add content to selection” mode.
Method 2: control-click to remove content from the selection
Alt-click (Windows) or option-click (Mac OS) to zone content for removal from the currently selected content.
Note that the mouse gets a minus sign to indicate you are in “remove content from the selection” mode.
Tip 2. Save your work at 5 or 10 minute intervals
You can’t really undo from content-reordering. Accordingly, be sure to save your work at very frequent intervals – hopefully, right before you make a move that screws up the z-order. Often, ordering is screwed up in several places, or putting it back into the right order is daunting for some other reason – usually because there are lots of objects on the page.
That’s why I prefer not to try “fixing” z-order problems when they occur. It’s usually easier and safer to go back to a PDF from just before you made the move which screwed up the page.
Save yourself the suffering, and dump that file – revert to your last save instead!
Tip 3. Review the whole PDF before starting manual work
If you are starting from an untagged PDF, be careful! Acrobat’s “Add Tags” feature is capable, all by itself, of screwing up the z-order even before you’ve started with manual tools.
Flip through the PDF before and after running “Add Tags”, or use Acrobat’s Compare Documents feature. If you see z-order problems before you’ve started manual tagging, stop and think through how you are going to address the problem.
- If the problem is restricted to only one or two pieces of content, maybe you can fix it in the content panel, but that’s a last resort, especially with complex pages
- If significant numbers of pages are damaged you have a more serious problem. You may need to recreate the PDF, rasterize some content, or otherwise change the affected pages so they survive the Add Tags function in better shape.
Want more? 6 ways to avoid z-order errors in the first place
- I’ll go ahead and say it now: Quark XPress and Microsoft Publisher make PDFs that are really hard to tag. Avoid these applications when tagged PDF is the objective
- Put graphics you don’t want tagged into the page’s “background”. For example, in PowerPoint, put images that should be artifacts (and thus won’t get Alt text) into a template
- Avoid creating overlapping objects. Even if the content appears visually distinct the objects themselves may still physically intersect. Be especially careful of graphics that span significant portions of the page
- If your images or diagrams are comprised of smaller objects, group them in your source application, or if necessary, convert them into a single image. This simplifies the tagging problem for Acrobat – and for you, if you have to do manual cleanup
- Avoid placing one object on top of another to create a specific visual effect (a practice known as “masking”)
- Use PDF creation software that creates tagged PDF rather than adding tags following creation
4 Responses
Some PDFs we get are so heavily vetted, there’s little chance of getting the client to “fix” them for ease of remediation. What you see is what you get. You mention manipulation in the Content panel as a last resort. Understood ‘cause it is slow and tedious. But in some cases and regardless of complexity, we have found it to be the only resort.
And we couldn’t agree more about PDFs made from Microsoft Publisher.
Re: highlighting content in tags tree
Leaving it on keeps the element associated with the tag. Seems useless to turn it off except;
Let’s say you’re perusing the pages and analyzing the tags. You identify an element and want to check its tag. Fine, in the tags tree, you can “Find the tag from Selection”. But now you want to scroll through more pages and do the same. In order for that to work, the tags highlight feature must be turned off or it’ll jump you back to the last highlighted element – (so now where’s that current element I wanted to check?)
Tip for this feature. While perusing, leave your tags opened in the tree. When you come upon another element, place your cursor on the last tag you were highlighted on. Now you can use the “Find the tag from Selection” (it will have changed from blue to a very lightly shaded gray)
Thank you for providing a name to this problem.
Another method I use to amend an error is to use the Edit Text & Images tool to select the overlapping layer, select > right click > Arrange and Send to Back.
Works the majority of the time if there aren’t too many layers involved.
I’ve found that using “Edit Text & Images” to change the layer order often (usually?) then results in an error in the Tag order or the order shows in the Order panel. Then fixing the Tag order or the order in the Order panel forces a layer change that reverts to the original problem of “hidden” text. It’s *very* frustrating.
I totally agree with fixing the source doc, but also understand the “we can’t change content” dictates that we see all too often in the government. Negotiating changes is an art form. 🙂
Also, I wholeheartedly agree that Microsoft Publisher produces horrible PDFs. One such PUB file I got to “make an accessible PDF” initially had over 4,000 failures just in the Acrobat Full Accessibilitty Check. I didn’t have the heart to run the more thorough PAC2 (PDF/UA) check. I ended up having to manually tag each page of the 8-page newsletter.
Comments are closed.