Make a fixed-page document readable on a phone or e-reader

Reading a PDF on a phone is an exercise in pinching and panning. The page is a fixed rectangle designed for A4, the screen is a tenth of that size, and no amount of scrolling makes a 10-point column comfortable. Reflowable formats solve this properly.

What reflow actually gives the reader

In a reflowable format, the text is a stream and the reading app decides where lines and pages break, based on the screen and the reader's preferences. That hands control back to the person reading:

• Font size becomes a preference, not a fixed property of the file.
• Text wraps to the screen width, so there is no horizontal scrolling.
• Line spacing, margins, and typeface can be adjusted for comfort.
• Night mode and sepia modes work properly, because the app controls the colours.
• Text-to-speech works naturally, following the document structure.
• Bookmarks and highlights survive font size changes.

The EPUB conversion tool produces a validated EPUB, which matters because malformed files are rejected outright by many reading apps rather than degrading gracefully.

Reflowable or visual

The tool offers both, and the choice comes down to how complex the source layout is.

Reflowable rebuilds the content as flowing text. This is what you want, and it works well for single-column, prose-heavy documents: reports, papers, manuals, articles, manuscripts, and books.

Visual packages the pages as images inside an EPUB container. Nothing reflows, the reader cannot change the font size usefully, and the file is much larger. It exists because some documents genuinely cannot be rebuilt as flowing text — a magazine layout, a densely illustrated technical manual, sheet music, a comic, or a form.

The honest test: if the spatial arrangement of the page carries meaning, reflow will destroy that meaning and visual is the right answer. If the page is just words in order, reflow is strictly better.

What to check after converting

Reflow conversion is an inference, and the same reading-order issue that affects every PDF extraction applies here:

• Column interleaving. A two-column academic paper can convert into alternating lines from each column if the drawing order ran across the page.
• Headers and footers inline. The running header and page number now appear in the middle of the text, once per original page.
• Figures adrift. Images land somewhere in the flow, often separated from their captions.
• Tables flattened. A table's grid was never in the file, so it can arrive as a run of values.
• Footnotes detached from their reference points.
• Hyphenation survivals, where words broken at original line ends stay broken.

Read the first chapter and one in the middle before committing to the file. Problems are consistent once present, so a two-minute check tells you whether the conversion succeeded.

Getting it onto a device

Most e-readers and reading apps accept EPUB directly, usually through a cable, a send-to-device email address, or a cloud folder. A few ecosystems prefer their own format and convert on receipt, which generally works fine from a valid EPUB.

If the file is rejected, the usual causes are that the EPUB failed validation or that the device is enforcing a size limit — visual EPUBs, being full of page images, hit those limits easily.

Alternatives worth considering

• Listening instead of reading. The audio export tool generates speech on the device, which is the better answer for a commute or for a document you want to get through rather than study.
• A dark reading copy. If the complaint is glare rather than layout, the night reading guide covers cheaper fixes.
• The opposite conversion. If you have an ebook and need fixed pages for printing or citation, the ebook to PDF guide covers the typesetting decisions involved.

A note on where the file goes

Personal libraries are personal. Research papers reveal what you are working on, manuscripts are unpublished, purchased books are licensed to you, and internal documents are internal. Converting a reading pile through an upload-based service hands over a fairly detailed picture of what you read. Doing it in the browser avoids that, and for a task this routine, avoiding it costs nothing.