DjVu to DOCX Converter

Why convert ebooks?

Every e-reader and reading app takes some ebook formats and not others. Kindles read Amazon's formats and EPUB, Kobo, Apple Books and Google Play Books read EPUB, FB2 is the format of many Russian and Eastern European libraries, DjVu holds many scanned books, and PDF and Word files are how books get printed, shared and edited. So a book often needs converting before it opens where you want to read it. The formats also hold more or less of a book: some keep its chapters, table of contents, notes and pictures, and some only its text.

The styles are bold, italic, underline, strikethrough, superscript, subscript and code. FB2 has no underline, and plain text has no styles at all.

Tool description

This tool converts ebooks between EPUB, FB2, PDF, DOCX, TXT and HTML, and from MOBI, AZW3 and DjVu to any of them. Every ebook converter here works the same way, from the general Ebook Converter to the ones for a single pair, such as EPUB to FB2, FB2 to EPUB, EPUB to PDF, MOBI to EPUB or DjVu to PDF: it starts on the formats in its name, and the Source and Target menus switch to any other.

Pick or drop one book or many. Browse files lists only files in the source format, but files dropped on it can be in any of them, and each one converts from its own format, so an FB2 file converts even on the MOBI to EPUB converter. Each book converts as soon as it's added, and again whenever you pick another target.

Books are converted on your device and never uploaded.

Convert many ebooks at once

Pick or drop as many books as you like, and add more at any time: this is a batch ebook converter, so a whole library converts in one go.

• The books convert one after another, in the order they were added, and the progress bar counts them.
• The converted books show up in a grid, each with its own Download button, and keep their original file names with the new extension.
• Download All saves every converted book in one ZIP file, or a single converted book as it is.
• When two books would get the same name, a number is added to the later one, as in book (1).epub.
• If a book can't be converted, an error message names it, and the others still convert.

How it works

Each book is read into its title, authors and other details, its chapters, its notes and its pictures, which are then written in the target format. So whatever the two formats are, a book converts through the same steps, and what carries over is what both formats can hold.

Reading:

• A book is read as the format its content is in, whatever its extension says: an EPUB renamed to .zip reads as EPUB, and a zipped FB2 as FB2.
• EPUB books are read in the order of their spine, a chapter for each of its files, titled by the table of contents. Bold, italic and other styles carry over whether the book sets them with tags or with the classes of its stylesheets, and footnotes marked as such become the book's notes. Fonts, colors, sizes and layout are left out.
• FB2 books are read in the encoding they declare, such as Windows-1251, which many older files use. Each section of the body becomes a chapter, the notes body becomes the book's notes, poems, epigraphs and citations become quotations, and a subtitle like "* * *" becomes a scene break. Genres, annotation, publisher, year and ISBN carry over too.
• MOBI and AZW3 books are read with the KF8 part of the file when it has one, as newer Kindle books do, else with the older Mobipocket text.
• DjVu books are scanned pages, and their text is the text layer that OCR adds to them. Its words are joined into lines, paragraphs and headings the way a PDF's are, from where they stand on the page and how tall they are, and the first page becomes the cover. The book is titled after the file. A DjVu book without a text layer converts to PDF only.
• DOCX documents are read through mammoth: headings in Word's heading styles, the Title style and quotes in the Quote style keep their meaning, and footnotes and endnotes become the book's notes.
• PDF files are read from their text: lines are joined into paragraphs, words hyphenated at the end of a line are mended, page numbers and running heads are left out, and lines in a larger font become headings. A PDF with bookmarks gets a chapter for each top-level one, and others get chapters from lines such as "CHAPTER IV." or from their largest headings. The first page becomes the cover picture. Pictures and fonts inside the PDF aren't read, and a scanned PDF has no text to read until it's run through OCR.
• TXT files are read in their encoding, as UTF-8, UTF-16 or the legacy encoding they're most likely in. Paragraphs are told apart by blank lines, indents or line breaks, whichever the file uses, and lines such as "Chapter 12", "Глава 3", "Kapitel 2" or "第三章" start chapters. The book is titled after the file.
• HTML files are read as one page, with the pictures it holds as data URLs.

Writing:

• EPUB books are EPUB 3, with an EPUB 2 table of contents for older readers too. Each chapter is a file of its own, the table of contents lists the chapters and the headings inside them, the cover gets a page of its own, and notes go on a page at the end that the note marks in the text link to.
• FB2 books get a section for each chapter, nested under its headings, so readers show them in their table of contents. FB2 has no lists, so list items become paragraphs starting with "1." or "•". Pictures are written as JPEG or PNG, which FB2 readers show, and others are converted to PNG.
• PDF files from a DjVu book keep its pages as they were scanned, each the size of the original, with the words of its text layer laid invisibly over them, so the text can be searched and copied.
• PDF files from other books are A5 pages, the size of a book, in Liberation Sans, a font with Latin, Greek and Cyrillic letters; letters it has none of show as "?". Each chapter starts on a new page, the cover fills the first, pages are numbered, and the PDF's bookmarks list the chapters. Links to web pages can be clicked, and notes follow the last chapter.
• DOCX documents start each chapter on a new page, with headings in Word's heading styles, which show in Word's navigation pane. Notes become footnotes, quotations use the Quote style, and pictures are scaled to fit the page.
• TXT files have the title and authors at the top, unless the text starts with them, then the chapters, with a blank line between paragraphs. List items start with "-" or a number, table cells are separated by "|", notes follow at the end, marked like [1], and pictures are left out.
• HTML files are one page holding the whole book, its pictures inside it, its notes at the end.
• Every converted book keeps the title, authors, language and description of the source, as far as the target holds them.

Tips

• To read a book on a Kindle, convert it to EPUB: Send to Kindle takes EPUB files.
• If a PDF converts into a single long chapter, it has no bookmarks and its chapter headings are no larger than its text. Converting a PDF made from an EPUB or Word file usually goes better from that file.
• Books bought with DRM copy protection can't be converted, and show up in the error message with the books that failed.
• To keep a scanned DjVu book as it looks, convert it to PDF. Its text layer gives the other formats text that can be resized and edited, but with whatever mistakes OCR made.
• Comics and other fixed-layout books are made of pages to look at rather than text to flow, so they convert poorly: the converter works with a book's text.