Recently a question came up where a writer asked others who formats their e-books. This felt like an odd question to me because I’ve always gone the DIY approach when it comes to formatting e-books. They’re pretty simple to do and there’s very little you can do with them. I follow a few golden rules:
- Don’t specify any fonts. Just use “serif” and “sans-serif”.
- Don’t have a specific base font size, and just make any styles with larger text use relative styles. (Yes I’ve broken this in the past; I’ll correct it in the future.)
- Specify all sizing, like indents and margins, in terms of em units (i.e., relative to font size).
- In the original document, never use tabs; the style should do the indenting.
And that’s it, really. To convert from Word to an e-book I save as filtered HTML, remove a lot of useless crud from the styles and clean it up according to the above rules, and remove any lingering inline styles. Then I run it through Calibre to convert to epub, and convert that to mobi. It’s pretty basic and I’ve never thought much of it. In Below I’m doing a little bit more with the table of contents because the book is broken out into five parts, but otherwise it’s just like any other e-book I’ve done.
But in the thread in question, some authors insisted that professional formatting could make a huge difference and be noticed by readers. Some formatters will deal with issues like extra blank lines and tabs and such to make everything look right, but honestly that’s my job as the author; I don’t consider whitespace patrolling to be a selling point. So what does that leave? That question baffled me, because e-books aren’t terribly flexible. There are certain things you shouldn’t do because they may work in one reader but not another, and many readers are customizable.
I took a look at what one of the authors in the pro-pro camp was doing for formatting in their books. According to Amazon’s previewer, they used images for their chapter headings (black on white, with some blue) so that they had a custom font, and all their body text was bold. I’m hoping the bold body text was a previewer error, but if so then it probably means they tried to do something clever with the formatting that wasn’t widely supported, leading right back to my point that e-books aren’t very flexible. If not, it means they put it all in bold on purpose, which is awful. But the image thing isn’t a lot better; many readers can choose to give their devices a different color scheme, like black on cream or white on black. So I distrust images used for this purpose, and consider them a non-viable option; images should be for the sake of including illustrations.
Font embedding is another thing that can be done in e-books. It shouldn’t—not for body text. Each e-reader has its preferred fonts that are basically chosen to fit that device. If your oh-so-wonderful font ends up looking bad on that device, that’s a problem. Readers may turn such fonts off, and the device may not support using embedded fonts at all. What then?
In the course of this discussion, though, I saw something curious. One formatting company claims they can make pretty, artistic scene breaks—something beyond the three asterisks I normally use—even in an e-book. (In a print book you don’t really need these except at the end of a page; a blank line will usually suffice.) It turns out, however, that they too are using images. And it occurred to me that this is a place where font embedding could actually shine. Font embedding may be able to do more than I thought.
Now here’s the catch: If you embed a font, you absolutely have to be prepared for it not to work. Whatever you do with font embedding has to be built around this concept. That means if I found a font that was nothing but pretty scene breaks, or one with lots of beautiful fleurons (this word was new to me until recently, but I love it), you couldn’t embed use that font as-is. Because if the device can’t support fonts or the reader turns them off, you don’t want a rogue Q hanging around on a blank line where a fallback like three asterisks would be better.
Which leads me to my idea: If you make your own fonts, your own fleurons and scene breaks and other stuff, you could make a font where there was just one glyph: an asterisk that looked like a long decorative line. You could make another font where the asterisk was utterly blank. End result: If you use the blank font for the outer two asterisks and your decorative one for the middle, suddenly a simple three-asterisk scene break becomes a decorative one on devices that support it. This is also completely compatible with print.
A similar idea would be to take a non-printing character like a space—or, in the event of any issues, a non-breaking space—and assign a glyph to that. Then you can have a decorative flourish or fleuron at the beginning of a chapter or other such places, even in an e-book and even when the color scheme changes. This might be limited because you won’t ever have as many placement options as you would with full HTML and CSS, but it’s enough to still do a lot of awesome stuff. And since some e-readers can support drop caps via CSS, imagine what a drop cap would look like illuminated with vines and scrollwork and what have you.
I’ve begun the first stage of testing my idea by creating some decorative scene breaks in Inkscape. I want to import those into a font, but Fontforge appears to be crashy and my alternatives look quite limited. I’ll find a way. I’m curious to see where it will work and where it won’t, because anything that I can use to give my books a leg up on formatting would be great.