The making of ebooks is a difficult matter!

It isn’t just one of your holiday games. You may think at first I’m as mad as a hatter … —The Naming of Cats by T.S. Eliot

And you’d be right!

Long post, but with “under-the-hood” sneak-peek images!

Making ebooks is not for the faint of heart. It’s a brain-lethal combo of geeky nerdiness, creativity, concentration, and obsessive-compulsiveness that often makes me want to … pound things. I’ve heard it referred to by other epub-making hobbyists as a particularly frustrating game of “whack-a-mole.” When you think you’ve solved a problem, another pops up.

An EPUB is a group of XHMTL files, images, metadata, and fonts zipped up in a special archive. So anyone can make one. Making one that has accessibility baked in, that works across platforms, is trickier.

There is a standard for a valid epub file. There is NO set standard for the epub readers and rendering engines. A reflowable-text ebook will look different depending on what device or app is used to read it. With more robust mainstream apps and devices, the differences are fairly minor. Yet most apps and devices have individual quirks.

The ebooks I make are open format EPUB 3.0, and what you, the reader, will see, won’t be exactly what I see. We each will choose our own preferred settings, devices, or apps and that’s OK!

I learned how to make epub by reading about the process online. I started my education in 2010, shortly after getting my first e-ink reader, a Nook. My first publication was for a nephew, to whom I ended up giving the Nook.

I’ve spent time looking under-the-hood at other people’s epubs. I borrow code like crazy. I test. There’s little consensus online as to how best to go about ebook making. There’s good advice and bad. Arguments over how to handle any given element abound. Add in accessibility, and you have yet another arena of disagreement. Which means that one does one’s best and accepts that perfection is not reachable.

It’s been a rocky road!


The Bones of an Epub

My first priority when I make epub is to build good bones! The bones are a combination of the XHTML files and also the HTML5 code that makes up those files.

This is a simple, linear children’s novel. Each chapter is a separate file. Each chapter has a level one heading. Each section of front and back matter has its own file and a level one heading. This makes creating a full table of contents easy in Sigil-EPUB Editor (free).

Good looks never hurt!

Then, I use CSS (Cascading Style Sheets) to give the ebook a look like its print counterpart. Sometimes, the print layout will not be suitable for reproducing. I preserve the vintage look when I’m able.

The “look” results from experience or eyeballing it until I like the result, usually testing on several devices. Most of my public domain books don’t have fancy styling and that makes life easier.

Performance is the end goal!

I try not to get in the way of end-user settings. I don’t hard-code left and right margins, fonts, font sizes, or line-height for main body text. I do set up body text as justified with hyphenation. Justified text without hyphenation is awful at larger font sizes.

If you prefer your text left-aligned, MOST e-reading apps and devices can override my setup easily. There are exceptions, but mainstream readers are generally fine.

Sharp-eyed folks will notice the chapter styling in the image below from my PC desktop looks plainer than in Sigil. Quite normal. This is a circa 2010 version of Adobe Digital Editions. It is a good test for the epub engine that runs on all modern Kobo e-ink devices. Older engines don’t support all modern code; what they don’t support is usually ignored. If a book works correctly in ADE, it’s a good sign!

The code being ignored by ADE is a couple of pseudo-classes:

p.chapter-lead::first-letter {
font-size: 1.60em;
line-height: 0.75; /* Set line-height by using default value of 1.2 divided by font-size. */
vertical-align: baseline;
}

and

p.chapter-lead::first-line {
font-variant: small-caps;
}

The first code chunk is responsible for the larger first letter of the first paragraph of the chapter. The second bit tells the first line of the chapter to display in small caps. Neat stuff!

It’ll show up if an ebook is loaded to Kindle via Send-to-Kindle (at least for modern Kindles), or if loaded as a kepub on a Kobo. On a modern Pocketbook device, just load the epub.


So, what steps take a vintage hardcover book to an ebook?

  1. Acquire the book! I try for beat-up, damaged, cheap copies. The ones that got scribbled on, or water-damaged, or are broken-spine poor specimens.
  2. Cut, trim, and run through a document scanner to produce a searchable PDF.
  3. Inspect the resulting PDF; make sure all pages are accounted for!
  4. Load the PDF into Optical Character Recognition (OCR) software and run it. I have an old, non-subscription version of ABBYY Finereader.
  5. Correct the OCR job by comparison with the scan as needed.
  6. Output a clean file. My preference is text format, but with mark up during my OCR comparison to preserve italics and curly quotes. These get added back later. HTML output is easier, but my program is old and won’t output the cleanest HTML code.
  7. Then, I open the text output file in a text editor. I join broken sentences.
  8. And then, AGAIN, checking all text against the PDF scan for paragraph structure. Here, I also watch for scene breaks, italics I didn’t catch earlier, or special characters I missed.
  9. When done, I convert my text file to HTML. This is when I do my search and replace to add my curly quotes and italics back in.
  10. I then drop my HTML chapters into my own Sigil-created epub 3.0 template. For each one, I make sure my heading structure is correct.
  11. Once that’s done, I auto-generate a table of contents based on my heading tags.
  12. Then the CSS styling starts. HTML5 is the structural and accessible bones of the book; CSS is the visual presentation.
  13. Now I’ve got a book that can be read. But it won’t have niceties like a cover, title page, or any illustrations yet!
  14. While it seems that getting the text done would be the most time-consuming part of the process, it’s not. It is image work that is the main TIME-SINK! Even the initial scanning takes amazing amounts of time.
  15. Few affordable 96-year-old+ books have dust jackets! Sometimes an interior illustration can be made into a cover resembling the original dust jacket. Other times, a cover must be concocted from scratch! … Artist hat on!
  16. Title pages are laid out using Inkscape (free). Finding exact or similar fonts to the original is not always possible. I create variants until one looks “right.”
  17. Interior illustrations get background cleanup.
  18. Once there are clean images, they can be added to the book. And then it’s time for some accessibility, ALT TEXT for each and every image that is non-decorative!
  19. Final proofread is done on an e-ink device for eye comfort! Left aligned, non-hyphenated text and a larger than normal font size. There may be issues from the original print copy that need correcting! Mostly, the latter will be typos or inconsistencies with spelling, hyphenation, or capitalization. Editor hat on!
  20. I make further corrections in the epub as necessary from my proofreading notes.
  21. Are we done yet? Nope.
  22. I like to acknowledge the work of the author and/or artist by providing some biographical information. But all that needs to be researched, written, re-written … Writer hat on!
  23. Testing is next. Even using a template, each book can have individual code that needs checking. Round one is on the computer, round two on e-ink devices, round three in apps on Android and iOS.
  24. Are we THERE yet? Where’s the ebook?
  25. Final checks through CSS validation, EPUB validation, Ace by DAISY accessibility checker, and sub-setting any embedded fonts. Done. Backup time!

Rev your e-readers! Ebook incoming!

I’ve finished PC and device testing today, having loaded the book on my Kindle, Kobo, Nook, and Pocketbook devices. I hope to tackle app testing tomorrow or next day.

The incoming ebook is not 96+ years old like most! It is one of those rare cases where the copyright was not renewed. I ran searches up the absolute kazoo!

Now you’ll understand some of the process I go through. I DO like making these forgotten books live again. Feels like a calling. Keeps my brain ticking over in my old age!

Leave a Reply

Your email address will not be published. Required fields are marked *

4 × three =