EPUB to PDF
An EPUB is markup that reflows. A PDF is pages that don't. Converting between them means inventing the pages, and that is the whole job.
Read in this tab. Never uploaded. Here's why that's true.
Some characters won't survive this.
This is a fixed-layout EPUB.
Every page is already a picture, usually a comic or a children's book. You'll get those pictures, one per page, and the text options below won't change anything.
More options contents page, bookmarks, page numbers
Working…
This runs on your machine, so the speed is down to your machine. Keep the tab open.
Your PDF is ready
Before you print a hundred pages
- Look at page one first. Open it and check the type size and the margins on screen. Wrong choices are cheap to fix here and expensive to fix at the printer.
- Print two pages, not two hundred. Paper is whiter and type is smaller than a screen suggests. 11pt looks fine on a monitor and reads small on A4.
- Set your printer to actual size. "Fit to page" shrinks everything by a few per cent and undoes the margins you just chose.
Reading it on a tablet or phone
Copy it across however you normally would. If the type looks small, convert again at a smaller page size rather than a larger point size — the page shrinks with it, so the text fills more of the screen.
Something went wrong
- 0
- bytes uploaded
- EPUB 2 & 3
- both read
- No DRM
- detected, never removed
What the two formats actually are
An EPUB is a zip full of HTML. There are no pages in it anywhere. The reading app decides where every line and every page ends, at the moment you open the book, using the font and the screen you happen to have. That is why the same novel is 312 pages on your phone and 190 on a tablet: neither number was in the file.
A PDF is the opposite bargain. Every glyph has a position on a page of a stated size, decided once, by whoever made the file. Nothing moves afterwards. That is what makes it printable, what makes page 47 mean something, and what makes it open identically on a machine that has never heard of your reading app.
So converting is not a translation, it is a commitment. Somebody has to decide the page size, the margins, the type size and where every line breaks. This tool puts those decisions in front of you, because the alternative is making them on your behalf without saying so.
What it does with the book
Reading the EPUB is the easy half, because the structure is declared rather than inferred:
- The spine gives the chapters, in order.
- The contents, either an EPUB 3 nav document or an EPUB 2 NCX, gives the chapter titles, which become the contents page, the bookmarks and the running heads.
- The cover image becomes page one, if the book has one and you keep it.
- The markup gives headings, emphasis, lists, blockquotes, code and figures.
Typesetting it is the hard half. Every paragraph is broken into lines measured against the real font metrics and justified by widening its spaces. Headings are kept with the text they introduce, so a chapter title never sits alone at the foot of a page. No single line of a paragraph is stranded on either side of a page break. Figures are drawn at their natural size, or scaled down to the column if they are larger, and moved to the next page rather than split.
What it does not do
DRM. Most books bought from Amazon, Kobo or Google are encrypted. We detect that and say so; we do not remove it. Books from Standard Ebooks, Project Gutenberg, itch.io, StoryBundle and most publishers' own shops carry no DRM and convert normally.
Non-Latin scripts. We set the text in the fonts built into every PDF reader, which is why nothing has to be downloaded and why the output is so small. Those fonts cover Latin and its accents, so French, German, Polish, Turkish and Vietnamese all set correctly. Greek, Cyrillic, Hebrew, Arabic and the CJK scripts do not. The converter checks and warns you before you commit.
Fixed-layout books. A comic or a children's picture book is already a sequence of images. You will get those images, one to a page, and none of the text options will do anything, because there is no text.
The long version of all of this, including the parts it gets wrong.
Questions
What actually ends up in the PDF?
The text, its emphasis, its headings, its lists and its pictures, set into pages at the size you chose, with a contents page, PDF bookmarks, running heads and page numbers. What does not survive is the book's own stylesheet: fonts, colours, drop caps and decorative layout are replaced with our typography.
Why not just keep the book's own fonts and CSS?
Because honouring an EPUB's CSS properly means a layout engine on the scale of a browser's, and the answer would still be wrong. That CSS was written for a page whose size the reader picks at run time. Yours is fixed the moment the PDF is written, so the type has to be set for the page you actually chose.
Does the table of contents work?
Yes, both kinds. The contents page lists real page numbers and each line is a link that jumps there, and the PDF outline appears in your reader's sidebar. Both are built from the book's own contents: its EPUB 3 nav document, or its EPUB 2 NCX if that is all it has.
What happens to tables?
Each row is flattened to its cells' text, separated by a middle dot, and indented. The grid is lost. Drawing a real table means negotiating column widths against content that was never measured for your page, and a table that silently overflows the margin is worse than one that plainly reads as a list.
Why is my PDF a different length to the ebook's page count?
Because the ebook never had one. Reading apps invent a page count from the current font and screen, which is why the same book shows a different number on a phone and a tablet. The PDF has a real page count, and it follows from the page size, margins and type size you picked.