ELLAUntagged PDFs
A tagged PDF has an invisible structure that says what is a heading, a list, a table or a picture. A screen reader uses it. An untagged PDF has text and nothing else, so a reader gets a stream of words in the order they were drawn.
How to tell
Many PDF readers show "Tagged PDF: Yes" or "No" in the document properties.
What ELLA does
When a PDF has tags, ELLA uses them. When it has none, ELLA works out the structure from how the page looks: large bold text may be a heading, a column of numbers may be a table. That is a good guess and not a fact, so the reply flags structure that ELLA is less sure of.
The result can be an accessible PDF, a Word file, HTML, EPUB, braille or large print.
What can go wrong
- Reading order in complex layouts such as newspapers, forms and multi-column pages.
- Tables with merged cells or no lines.
- Headings that look like body text, or body text that looks like headings.
- Headers, footers and page numbers mixed into the text.
Check what matters
Read the reply, then open the result. Check the headings first, because they make the document navigable. Then check every table. See accessible Word and PDF.
Better still: use the original
The Word, PowerPoint or other file the PDF came from has the real structure. ELLA is more reliable from that, and fixing the original means every later copy is better too. Ask whoever made the PDF for the source file.
If the PDF is a scan
If the PDF is pictures of pages, it has no text at all. See scanned PDFs.
Still stuck?
Browse the help centre, or contact support and say what you were trying to do.