Acrobat X Pro, Windows 7 ... working with 500+ page PDF generated with structure tags from Word 2010, as usual, tags then manually inspected and cleaned up, also as usual.
Normally, Read-Out-Loud and JAWS do not read artifacted text, do read alt text. In this PDF the opposite happens, and when saved-as Accessible Text, all artifacted text comes through (including page #s), all alt text does not.
On the other hand, PAC 2.0 Preview does show the alt text and not the artifacted text. PAC 1.3 checks it ok, but won't preview it, gives an "unhandled exception" error.
All pages have Tab Order set to Use Document Structure. Under Tags panel options, "Document is Tagged PDF" is checked. PDF passes Acrobat's accessibility validations, passes PAC 1.3, passes the usual items in PAC 2.0.
In Contents panel there is a single level of containers from the root, no sub-containers. All text to be ignored is in Artifact containers. Figure containers only have graphic Paths and Objects, and Figure tags have alt text.
Any ideas what's going on, how to fix?
Sounds bizarre. Could share the file, or maybe a few pages of it?
Yeah, it was bizarre, probably file corruption ... after I posted this question I tried extracting all pages to a new file, and, guess what?! ... the new file is about 300K smaller and, more importantly, behaves correctly - doesn't read artifacts, does read alts - so whatever problem the original file was having did not extract to the new file. Extracting the pages brought along all the tags and links intact, just had to export/import bookmarks & destinations, re-enter doc props, re-number thumbnails. Whew!
oh - and, PAC 1.3 opens it in preview now too, that got rid of the error.