The world is full of typos!
While reading two separate posts on Forbes and The Verge in the last two days regarding the annoyance of typos in e-books, it was reassuring in some respects to note that the brunt of the complaints was against large publishers. However, as I read the long list of comments on these posts, a few new pieces of information caught my attention. Although there was one very important consideration that was missing.
The new information I found interesting was contained in one comment. I quote:
GAP e-books: an explanation for why they are so shoddy
I’ve just learned that the production AND PROOFREADING for Bantam/Spectra e-books is done…in India. By people who barely speak English. (Apparently this applies to every e-book published by the conglomerate which includes Bantam.)
If this is true, it says a lot about how quality control in publishing is being outsourced to the cheapest contractor.
Another practice being used to convert manuscripts to electronic files is by using OCR. (Optical Character Recognition). Anyone who has had any experience with OCR will know it is a process that is far from perfect. A quote from the Forbes article:
Unacceptable or not, that’s what someone has done. Simply OCR’d the printed text and not subbed it through again.
I shudder to think what the result would be, as my experience with OCR has been that the result always contains character errors or typos on every page.
While these two areas gained a mention, along with the obvious blame on poor editing, sub-editing and proofreading, the one missing reason for ebook errors is the process itself. No matter in what form a final text is prepared, be it html, doc, rtf, pdf or any number of other file formats, it will then be converted yet again into the file used by each ebook distributor or retailer. As Kindle, Apple, Sony and all the others use a variety of e-publishing formats, even the most perfect text needs to be ‘crunched’, ‘auto vetted’ or ‘converted’ to this new file type. In other words, the words of the text are converted into ones and zeros, and then back again into text. And rarely perfectly.
This differs completely from the technique used in a printed book, where all text is basically reproduced using photographic processes and therefore reproduced exactly as intended.
From my own experience, when I have download copies of one of my ebooks from different sources, they are NEVER exactly the same. The most common problem are changes in formatting, removal of italicised text, removal, replacement or misinterpretation of accented characters and random changes in fonts and paragraph styles. Quite honestly, some conversions are quite good, while some I could only call a dog’s breakfast. It also makes a huge difference if a file has been prepared using Apple programs or Microsoft programs. Most ebook conversions programs will not work with Apple word processing programs, or if they do, they add random characters and spaces and totally change formatting styles.
So the grand ebook debate about typos will continue I am sure, but it is worth noting that the complaints against sloppy authors, poor proofreading or lousy formatting could, and perhaps should really be aimed at the ebook process itself, and not necessarily towards those who work hard in the preparation of the texts themselves.