I have copied and pasted the details, below, from a previous message I have sent out to an assistive technology listserv.
I am encountering this problem with various builds and versions of Adobe Acrobat:
Acrobat X on a Windows Vista 64-bit build (note that this is the best Acrobat I can install on Windows Vista, Adobe Reader XI is not even supported)
Acrobat X on a Windows XP build, sorry, do not know the bit count (it is my work computer so I can't upgrade anything)
Acrobat 9 on a Windows 7 64-bit build.
I have also encountered and checked into whether or not the PDF reads (and does what I want it to) in JAWS 14, latest update (Feb 2013). We encounter the same problem with JAWS and this morning found out that there is a conversion error between Word and Adobe. (See this post, where the ---- are).
Summary, I have spoken with JAWS, put out a support ticket with NVDA via e-mail, spoken with Microsoft, and tried to communicate with Adobe about this issue for help and Adobe is refusing to help. The other places tried everything they knew but they could not get it to work. I even tried to strip all the page numbers from Word and number using Adobe's page numbering feature, but that was not successful either because I cannot use Roman numerals.
I would appreciate any feedback the Adobe Forums can give me; please note that I cannot post a sample document at this time as the only document I have is a private work document (cannot share those by policy, so I will have to make up a fake document if I get the time and energy).
Thank you very much for your time and expertise. The pertient info is above and below in "Original Message" and "Activities section" - the other information is there if you'd like to look. And to provide feedback to Adobe about things I think they could improve, since they will not let me email them directly.
My question concerns the reading of Footers / Headers in Microsoft Word 2007 and Adobe Acrobat (headers/footers designed in Microsoft Word 2007 and then converted to PDF with "Bookmark" checked in Adobe Acrobat conversion settings).
The document is a word document containing 50+ pages. There are 12 pages of "front matter" that are marked with Roman numerals. Subsequently, the body text of the document has an Arabic number (1, 2, 3, 4, etc.) in the bottom right corner. Footers are used properly in the document. All Table of Contents links, which do link to both headings in the "front matter" and to headings in the body text, are picked up properly. Literately, the only thing not reading on the document are page numbers (the Word Status bar is reading its page numbers, but the page numbers in the footer are not reading).
I then need to convert the document to Adobe Acrobat PDF. Granted, I have a full copy of Adobe Acrobat X available to me, and my various assistive technologies have always functioned better when opened directly in Acrobat (rather than Adobe Reader). So, I set the accessibility options (from Word) properly, asking it to bookmark Headers and Footers as appropriate, thinking that this will pick up the page numbers and make them read as the bottom right corner of my PDF.
All links work in Adobe to navigate by section/heading. All figures are alt tagged. Everything is perfect - but the one thing that NVDA won't read are the page numbers. Visually, the page numbers are there on the document, but NVDA won't pick them up. Additionally, I also changed the appropriate page numbers in Adobe (Page Thumbnail Pane, on the Navigation Pane I believe it is called) - to reflect the section where it is Roman numerals and Arabic numerals. This did not help, and for what it is worth, I cannot get NVDA to voice when I am in this pane, so I don't seem to have a chance of getting the page number info from there.
NVDA reads the document just fine except for the above-mentioned snare. However, when it reads, it will go to the next page, and say, "Page 2 of 54, (page text), Page 3 of 54 (page text)," etc. when I will need it to say Page ii of 54 (page text), Page iii of 54 (page text)," and so on, changing to "Page 1" when the Arabic numbers are used.
I know that the Page Label issue is part of Adobe's issue, with them not releasing the PageLabel aspect. (I have looked through some NVDA tickets). I do not know how to put a "changeset" into my copy of NVDA though, or even if it would help (I have no computer scripting skill). The easy answer would be to upgrade to Adobe Reader XI and see if that helps, but I don't have the ability to put it on every computer I use, and *I need this file to read consistently across multiple versions of Adobe* (including Adobe Reader/Acrobat 9, X, and XI). (This file is also going out to people who may not have the latest version of Adobe, but are also running some copy of NVDA, either a portable or a full install).
Is there any way to pick up the Footers with page numbers voicing in NVDA, and/or have it read the user-editable page number box that is on Adobe's Toolbar (next to the 1 of 54 parentheses). This user-editable box, to jump to page numbers, reflects the Roman numerals I have loaded into the pages of Adobe. When I press Ctrl+Shift+N in Adobe, I can also go to the appropriate page (if I type in iii, it will take me to iii, if I type in 34, it will take me to the Arabic number 34 -- NOT the 34 of 54, which would land me on a different page. And I want to land on the page that has the Arabic 34, so that's functioning fine.
I just need the page numbers in the footer to voice, "Page iii," or "Page iv," or "Page 29," etc. If I could get NVDA to do that, I could say to viewers of this document, "If you are using NVDA, remember to Ctrl+Shift+N to get the GoTo Box, then type in the page number you want if you cannot follow a link or a bookmark." This document is *very* accessible in my opinion with lots of ways to navigate...the only aspect of navigation that isn't being picked up are those footer page numbers!
I do have my Document Formatting settings on the NVDA menu set to "Report Headers," but that does not seem to help in either Microsoft Word 2007 or Adobe Acrobat. I have even switched the page numbers from the footer to the header to see if that would help and it didn't.
----Activities I have done today----
I spent over an hour on the phone today with Freedom Scientific (makers of JAWS) trying to troubleshoot this. We discovered that Footers will not read very well in JAWS and Microsoft Word 2007 (only solution is to stop reading document text and read Virtual Viewer text briefly, then go back out to document text, then back into Virtual Viewer which is NOT an acceptable or accessible solution whatsoever -- too much work for someone trying to read the document) -- and then we also discovered that:
Upon conversion from Word 2007 to .txt file (.docx to .txt), there are no alt tags for the figures in the document, or page numbers.
Upon conversion from Word 2007 to Adobe Acrobat X then to .txt file, every insert of a page number and footer is replaced with the same alt numberpad numerical code: the one that generates female. ♀No wonder JAWS and NVDA are skipping this, neither understand how to communicate it.
On the advice of JAWS Technical Support I contacted Microsoft Accessibility Technical Support and spent an hour on the phone with them. They say that unless Adobe can find a solution, it appears that it is impossible for Microsoft products to read the footer if JAWS was unsuccessful doing what I wanted it to do. The document/footer in question includes about 12 pages of front matter (numbered in Roman numerals) and 44 pages of body text (numbered in Arabic numerals).
I kindly explained to the Microsoft Support Agent that I found this issue hard to believe, although I understood. Sighted folks have the ability to make their documents look quite professional and that is the caliber and quality of documents anticipated from everyone, especially college students graduating from school, or job applicants. That a coding issue prevents the footer from being read properly, except in Edit view, is disappointing. The representative was with me 100% of the way. She completely understood where I came from. And yes, when I am designing the document myself, I know what I put there (or I pretty much do, anyway). But if I recieve a document, it is much harder to tell what is really there or not, or how the page numbers really lay. And this is confusing as heck, believe me.
I then called Adobe Technical Support, after having a brief online chat with them. Granted, I had to leave the chat in the middle because I got interuptted by something that was high-priority, but the woman chatting from Adobe says that chat is only for installation issues, and I will have to pay to open a support ticket. Excuse me? I have paid a hefty sum of money for Adobe Acrobat 9 and Adobe Acrobat X (work paid for the other Acrobat X copy). I deserve this problem to be troubleshooted for free. There was even a statement on the Adobe Acrobat X website that said Acrobat X users didn't have to pay for support, but Reader users did.
----Slight bit of rant and constructive criticism----
In speaking with Adobe Technical Support, I had to wait on hold for 30 minutes before my call was answered, and then I had to consent to being on hold countless times and had to answer, "Yes, I am converting from Word 2007 to PDF using Adobe Acrobat X" at least 6 times. The representative would put me on hold, then ask me the question, then put me on hold again, repeated 6 times over 45 minutes. At the end of 45 minutes (1 hour and 20 minutes of wasted time by now, that I must justify to my employer) - I told the representative I wanted to speak with his supervisor immediately, I did not care if he was trying to fill out a support ticket (and putting me on hold at least 4 more times in 10 minutes while he claimed he was filling out the support ticket) - and finally he put me on hold again and gave me to a very helpful supervisor. Within 6 minutes of speaking with the supervisor the supervisor had my support ticket filled out and I should recieve a call from Adobe within the next 30 minutes if his estimation was right (he estimated two hours, but it's been an hour and a half).
I have yet to know what can be done about my document. Freedom Scientific Technical Support is closed now for the day and I need to work tomorrow and Friday away from the phone. As I stated to Microsoft Accessibility Support, I truly feel that Adobe should read the page footers, every document says that Adobe should be able to read the page footers, and I feel slighted that as a person who uses assistive technology, there is some glitch that isn't making this happen. I am not blaming a specific place per se, except I am a little frustrated with Adobe especially since they have not gotten back to me, and I am frustrated with their customer service. JAWS customer service and Microsoft customer service were both exemparly, and I cannot say enough good things about them. My wasted time with Adobe however, that was disrespect in my opinion. And I certainly will not pay to open a support ticket, and my workplace has a site license. They've paid enough already.
If you've read this far, thank you. I look forward to replies, or at least understanding. And if anyone has ideas, I'd appreciate it very much (yes, I have tried taking the page numbers off and using Adobe's page number feature, but that would not allow Roman numerals. Only Arabic.).