Image tags outside of a table tag when converted to PDF from Word.

[Paul-BT]

Hi Folks,

I have recently encountered an anomaly with the tags structure in a PDF document when converted from Word in Office 365. It's a booklist document produced in a table with two columns and numerous rows, and in each row in the first cell (column 1) there is an image of a book cover. Regardless of whether alt text is applied or noted as null ("") for the image, when the file is converted to PDF, all the image tags are placed outside of the table tag. This means I have to sort the reading order first and then place the tags back in the relevant TD tag in the table. Does anyone know if this is because Acrobat has a bug, or it is forcibly doing so on purpose because perhaps you shouldn't really have images in tables - even if the document is made fully accessible?

Thanks

[Karl Heinz Kremer]

Are you converting using Word's on PDF generator, or are you using Adobe Acrobat to convert to PDF? Do you get different results depending on which method you use? You should be able to have images in tables represented in the tag tree. 

[Paul-BT]

Hi Karl, I would not have thought to try the other option of 'Save as' PDF (wouldn't use print for obvious reasons), as I have always been led to believe that if you have Acrobat installed, you should use the 'Create PDF' option from the ribbon. Add to that, I'd expect both options to produce the same result - how wrong I was! Having just carried out a couple tests using both options, the resultant tags structures are noticeably different (see the image below). It's clear that using the 'Save as' option not only preserves the images in the table as expected, but it also excludes the extra unnecessary <P> tags you get when using 'Create PDF'. I have also tested both results with NVDA and behold the 'Save as' option works with that as expected, so all in all - result!! I am however now questioning whether anyone should use the 'Create PDF' option at all - perhaps you could clarify why they differ and if that's the way forward? Thanks again, Paul

[Karl Heinz Kremer]

One of these uses Microsoft's PDF generator, the other one uses Adobe's PDF generator. There are no hard rules about how to tag a document, so both of them decided how they would want their tag tree to look like. Now that you have options, pick the one that fits your expectations. 

[a_C_student]

Usually the PDF generator works better for me, but not always. If I get screwy results I try "Save As". Also, sometimes converting w/o tags then auto-tagging in Acrobat works best. And, sometimes it is easier to hand-tag from scratch. 

[Paul-BT]

Thanks all for your helpful responses!