How big should an ebook file be?

(Originally appeared on teleread).

How big should an e-book file be?

I asked myself this question after trying to troubleshoot a problem reading a Plucker file. The Plucker desktop lets you spider URL’s to a certain depth, so that means you could potentially deal with humungous files if you’re not careful. (Plucker Desktop actually offers lots of options for controlling file size).

But what is a manageable file size? How big can an ebook be without taxing the constraints of the device or software or user experience? This question is complicated because ebooks are viewed on two different kinds of platforms: miniature devices and desktop/laptop based devices.

For the laptop, RAM and disk space are no longer hardware constraints. I’ve opened pdf files that are several MB without problem. On my main desktop, I have nearly one terabyte of storage and 3 gigs of RAM. It seems like overkill right now (I’m editing video with it), but in two or three years it will probably seem underpowered.

Portable devices still face significant hardware constraints, at least in comparison to desktops. My Nokia 770 has 1 gig MMC memory card but only 64MB of onboard memory that the user can use. My Dell Axim had 128MB.

In the days of dial up, the optimal size for a web page used to be 30-40KB, although nowadays blogs and longish articles frequently exceed 100 KB (for the record, the main teleread page is 178KB + 253KB for graphics). Actually, for me, the limitation is not size of web page but how many tabs in firefox I can leave open.

E-books are typically binary files for offline reading. Look at the list of formats and file size for Cory Doctorow’s Eastern Standard Tribe . Some of the more esoteric formats can exceed 500 KB, but the popular ebook formats are in the 200-300 KB range. (The pbook version is 224 pages and contains 49,000 words).

But Eastern Standard Tribe is a relatively small book, and my .imp version only has a single grayscale graphic. Many Gutenberg books are 1000+ pages, and each volume of the Mahabarata on blackmask is 2-3 MB. That’s without graphics. But what if you dealt with an ebook containing lots of multimedia?

First, images. How many images should an e-book contain? How big should the graphic files be? Second, audio, embedded flash and video. PDF and apparently dotReader have the capability to embed these multimedia objects. Good, yes, but if dotReader is ported to a PDA/phone, I’m guessing the bloat would render the e-book unusable. Still, the dotReader site notes for dotReader software “a notable memory management feature is the ability to display content only as it is being called up (as opposed to loading an entire document and then displaying it).” That sounds promising (and that may address hardware limitations), but it still does not help trim file size.

Ok, we can agree that sound/flash-enabled e-books are useful only in special contexts. But we’ve already grown accustomed to an online reading experience saturated with all kinds of graphics. Indeed, one significant advantage e-books would have over pbooks is its ability to display graphics without significantly adding to the sticker price of the book. If a 224 page p-book had an illustration every three paragraphs (which is typical for blogs), you’re probably talking about coffee table book prices. If the e-book version had an illustration every three paragraphs, it may not cost more; but how much would it increase the size of the binary file? (and how convenient and practical is it for the user to save it on a memory card?)

I am a content creator working on a DIY e-book, and have to admit I am clueless about how many graphics ought to be in an e-book or how compressed the graphics ought to be or how big the final binary ought to be. Content creators need better guidelines (or benchmarks) about ebook output and file size. Not only do they need to figure out things like accessibility and screen real estate, they need to know what file size to aim for and how to how to optimize for low-end devices without them looking crappy in higher-end devices. Obviously, these benchmarks are going to remain a moving target. For web designers, though, it was relatively easy to work within low-bandwith constraints. You knew to fit everything on a page to 40KB, while providing a link to high resolution versions of the graphics. The difference between designing for a browser and designing for an ebook reader is that a browser loaded individual resources. All the graphics and multimedia remained available on the web even if they didn’t automatically load. But would it make sense for a zipped ebook to contain high resolution and low resolution versions of the same graphic?

    To summarize, here’s what I’d like to know:

  1. When creating e-books, what should the optimal file size and resolution be for raster graphics?
  2. Do certain e-book platforms (such as PocketPC/Microsoft Reader or Palm/Plucker or Nokia/FBReader) have practical limits on the file size it can open?
  3. How big will a typical Sony Reader e-book be? (and how will it vary according to graphics).
  4. How many e-books is a typical reader comfortable dealing with on his portable device? 5? 10? 50? 100? How many is the minimum the reader would expect to have on his memory card?

Posted

in

by

Tags: