Examples and conversion testing
Learn how to build a useful conversion check, choose source examples and record what actually happened. Real source documents make the reading task concrete; conversion claims require a saved source, output and reproducible check.
Separate missing words, incorrect order, picture problems and reading-app behavior. Keeping those observations distinct helps you make a clear decision about the result.
Know which kind of example you are reading
Source-page images show the original words, tables or artwork. Reading exercises explain what to inspect. Actual conversion screenshots show what a particular tool produced from a saved PDF. Check whether a result was edited before treating it as an automatic conversion.
The seven-converter research-paper comparison includes default downloads and EPUBTurn’s reviewed result. The source and output library provides the real paper files and a nine-page scanned cookbook excerpt. The cookbook examples now include EPUBTurn’s unedited measurement sample and a three-page photograph and recipe sample. Calibre converted that same three-page recipe input with default settings. The separate full nine-page cloud attempt remains incomplete. The illustrated-document guide also shows a saved NASA hardware sample; only three of its six source pages were selected.
Record a small result that someone else can inspect
Save the source, output, settings and exact location of the observation. For reading order, follow a complete passage across a column boundary. For figures, inspect edges, caption and the paragraph that mentions the figure. Counting archive images alone cannot tell you whether the useful figure survived: covers and raster source pages can also appear as image files.
Use a reading-order record, a caption relationship check and the actual image inventory for those distinct questions.
Keep unanswered tasks visible
A sample that omits a footnote has not tested footnote navigation. An EPUB imported into one app has not passed on every device. A blank questionnaire is a study plan, not participant feedback. Record not covered or not tested instead of filling those gaps with assumptions.
The neutral reader protocol separates factual task answers from layout preferences and supplies no participant outcomes. Use the comparison record to retain a clear scope and avoid replacing an earlier raw output silently after editing.
Explore the guides
Check an EPUBTurn sample before paying for full conversion
Check paragraph order, missing content, figures and adjustable text in the actual EPUBTurn sample, then decide whether to proceed or stop.
Read guideInspect real PDF and EPUB examples: source, raw output and edits
Inspect a published research PDF and its saved EPUBs. Compare column order and complete figures, with manual corrections and access limits disclosed.
Read guideBuild a reproducible PDF to EPUB comparison record
Build a repeatable conversion record with licensed source files, raw outputs, settings, observations and corrections.
Read guideIllustrated PDF to EPUB: actual diagrams, charts and reading results
Compare an actual published research figure with the reviewed EPUBTurn export. Its three plot panels survive, but lower axis labels are clipped; three equation crops needed correction.
Read guideDownload real PDF sources and actual EPUB conversion outputs
Download real published source material and inspect the saved reading files with their documented limitations.
Read guideBuild a reading-order test for multicolumn PDFs
Distinct passages make column-order errors easy to spot.
Read guideTest whether images stay with their captions
An image count can miss swapped captions or unreadable labels.
Read guideDownload a reader-compatibility checklist for your exact EPUB
Record import, text size, figures, navigation and offline tasks for the same EPUB in each app and version. Download a blank checklist with no invented device results.
Read guideConversion cost and correction time: download a measurement sheet
Record actual cash, credits and setup, waiting, inspection and repair time for a defined reading task. Download the worksheet without an unperformed cost ranking.
Read guideRun a fair EPUB reader task: questionnaire and protocol
Download a neutral reader questionnaire and protocol. Separate factual task answers from layout preference and report participant limits without invented outcomes.
Read guide