You convert a PDF and the words are there, but the reading experience feels wrong. A sentence stops in the middle. A page number appears inside a paragraph. A sidebar is read before the passage that introduces it. These are structure problems, not just styling problems.
A PDF describes a page, not always a story
A print PDF records how content is placed on a fixed page. That placement does not always supply an unambiguous paragraph boundary, heading hierarchy, or reading sequence. A converter has to recover relationships that a person may infer from the design.
This is a well-known difficulty: calibre’s conversion documentation describes significant limitations with PDF input, including complex layouts. That does not mean every PDF conversion will fail. It means source quality and review effort are part of the decision.
Five patterns worth checking first
1. Line breaks that become paragraph breaks
Compare a few ordinary body paragraphs against the PDF. If every printed line becomes its own paragraph, the text can feel fragmented when the reader changes font size. If separate paragraphs have been joined, a change of subject may disappear.
2. Running heads and page numbers in the text
Read across a page boundary. A repeated chapter title or folio should not interrupt the sentence. Removing repeated furniture requires care: a phrase that repeats may also be legitimate content.
3. Columns read in the wrong order
On a two-column page, read the last lines of the left column and the first lines of the right. Check sidebars separately. Visual proximity is not always a good guide to the intended sequence.
4. Tables reduced to loose text
Find a cell and identify the heading that gives it meaning. If the relationships are lost, adding borders or changing a font will not repair the table. Review the structure and the content together.
5. Notes separated from their references
Follow a footnote reference and check the note text. Inspect whether a note at the foot of one page was incorrectly joined to the body on the next page. Test a way back to the relevant passage.
Build a repeatable review pass
Begin with a representative chapter rather than the simplest page. Include one example of every major content type. Keep a short issue log with a location, the observed problem, the expected result, and the correction.
After editing, inspect the surrounding content as well as the individual fix. Joining two paragraphs can repair one sentence while accidentally swallowing a heading. Then publish again, run the relevant checks, and read the exported file in a target reading system.
Set a practical acceptance criterion before scaling up. For example, require a complete chapter review with no unresolved content-loss findings and a documented plan for the remaining accessibility checks. This is an example of an internal quality gate, not a universal standard.
Sometimes the best fix is a better source
If the editable manuscript is available, compare the work needed to export from it with the work needed to reconstruct the PDF. For image-only scans, choose an OCR workflow appropriate to the source and review its recognition errors. DocSpry does not support scanned PDFs.
For born-digital files, DocSpry’s PDF-to-EPUB workflow gives you a place to review headings, reading order, and other structure before publishing. Test the result on your own content; do not assume a simple sample represents every title in your backlist.
Sources & further reading
General workflow guidance. Product suitability depends on the source document and the receiving system’s requirements.