Going Kindle for less than the price of a tasty sandwich

I am, it has to be said, a huge fan of ebooks. I bought a Sony e-reader two years ago and owned a Kindle the minute they were released in the UK. I now exclusively read Kindle books because of the convenience. In the main the books are also cheaper. That said, having gone to the trouble of writing and actually printing a book, there is something quite satisfying about holding in your hands tangible evidence for all that toil. Especially when the book looks THIS good. The reality is, while electricity is available to power portable devices, the future role of printed books will gradually evolve away from what it is now. The lesson I learned on this debut publishing experience was a printed book takes a lot of time to produce, and almost as long to filter though wholesalers and distributor databases. In future I'll be making the Kindle version available while the printing mechanics clunk through and not afterwards. Compared to printing a book, publishing for a Kindle was almost a breeze.

I did consider using the free e-book management tool - Calibre. It does a damn fine job of making your book Kindle ready and with a little learning you can get very close to a professional finish.  I didn't consider for one minute uploading the HTML or Word book file directly to Amazon and letting them do everything else, as that would have left the book very much in the hands of automation. I could have hired someone to create the Kindle edition, but it's not that difficult if you've ever done a bit of web design. I did it myself so that I'd have total control over layout. My mantra throughout has been to produce a product at least AS GOOD as those available from 'proper' publishers. To this end my baseline for laying out both book versions was Lee Child's - The Affair.

Both digital and printed versions began with my completed draft of the book. Chasing Innocence was written in OpenOffice and then exported to Word for copy and proof editing. Once that was done it was imported to QuarkExpress for typesetting, for the per page layout of the printed book. Then it was exported to a PDF document format and sent to the printers. For the digital versions, QuarkExpress doesn't have a 'save as Kindle' or 'save as epub' option, so the book was exported as Word. And from Word I exported as 'filtered html' - which removes a lot of the Word specific 'junk code' and leaves you with a relatively un-cluttered html file. HTML is a text file, formatted so you can see content in a web browser. It is also the basis for the content of digital books. The Kindle especially wants uncluttered HTML with little formatting. This is because the Kindle expects to handle much of the text layout or digital typesetting itself, including the font style, the font size and line spacing at the readers choice. I decided on three simple paragraph styles. The first forced the Kindle NOT to indent the first line of each chapter or section. If you take a look at the first paragraph of any chapter of any printed book, you will see why. Without this formatting the Kindle will indent the first line of every paragraph. The second style tells the Kindle what size of indent must be applied to all other paragraphs. No other styling was specified because the Kindle supports very little. If you try to get fancy, what looks pretty in a web browser will probably be ignored or worse, confuse the device. The third was for styling chapter headings. I kept this simple too, trying to reproduce the look of the printed book with centred text and each new chapter starting several centimetres down the page, using the top and bottom margin tags. These three styles made the content of the HTML file nice and easy. I also used the top and bottom margin tags to create ad-hoc spacing in the non-chapter pages at the front and back. Then I manually added in page-breaks at every required section, as letting the Kindle decide had some unexpected results.

The Kindle has a menu option, from which you can select 'GoTo...' From here you can select the beginning, the cover, table of contents and the end.  The device works out the end for itself. For the Table of Contents to work you must name each section in the book using standard HTML 'A' tags. As I had ninety-five chapters and several other distinctive sections, I named the sections and used Calibre to generate the table of contents from a copy. I then pasted that into the beginning of my HTML book file. To make sure the Kindle GoTo menu option works the cover needs to have a name tag of: 'cover', the table of contents: 'TOC and the Title Page: 'Start'. Creating the cover itself was a matter of cropping the book's cover to fit the Kindle screen (600 x 800) and linking the image into the HTML file, and giving it that 'cover' name tag. For the Beginning option you make sure the title page is named 'Start' and you're almost good to go.

Apart from the main HTML body of your book and the cover image file, you need two other files for a fully packaged Kindle book. The first is used by the Kindle to navigate backwards and forwards between chapters in the book. Your bundled book won't be accepted by Amazon without one. Once more I used Calbre to create the 'toc.ncx' and then edited it to my needs.

The next file is almost the final stage. The contents file (OPF) is used to compile your book. I used MobiPocket Creator to do this as it was free and recommended by Amazon, editing the contents it generated to match Amazon guidelines for embedding the cover and pointing to the toc.ncx created previously. I then used it to compile the book. Now this is where it gets interesting. You want to check your book looks good on a Kindle right? While you can download the Kindle previewer for your PC/MAC, it doesn't exactly match how your book will look on a Kindle. If you actually own a Kindle you may know it has its own email address which you can find in its settings pages. If you email your compiled book as an attachment to this address, from the email account you log into Amazon with, it will appear on your Kindle. Usually within five-ten minutes.

By doing this and reviewing on the Kindle, I noticed and changed some formatting, and decided it would look more professional with some logos on the title, copyright and final page. So I added these at the correct locations like I did for the cover and after a few revisions we were good to go. A few days later after the final check-through, I logged into Amazon, specified a separate cover image for the Amazon page, linked it to the printed books ISBN and clicked go. As I own worldwide rights I had to specify a price for each region, instead opting to let the books pricing be based on the US dollar value. This works out at £3:49, which was my initial target price. I will of course monitor and amend the price. Amazon then took the file and processed it, allowing me to download the exact file of Chasing Innocence they will send to your Kindle when you ever so kindly buy it. After I tested it was good, I clicked Finished and after a request to prove I was the author, copyright and world wide rights holder, the book suddenly and quite majestically, appeared on Amazon.

Now all you need to do is click on this link to go to the Kindle version of Chasing Innocence, where you will be one more click from seeing what I've been banging on about these last few months, and importantly, you will be mere seconds from being the proud owner of an enthralling, thriller fiction digital novel that cost you less than a tasty sandwich.

Shades of Grey

The real thing

Last week, two days before Christmas Eve, I received the printed proof of Chasing Innocence. The proof is for me to accept everything is correct, that there are no errors in the cover or layout. Once I click 'Accept' on the printer's website, the book becomes available to buy. For those that have bravely pre-purchased, it means the book will be printed and despatched. Accepting the proof is a big thing. I was totally in awe of the proof - from the fact I was actually holding my book as a real book for a start. It also looked sensational and everything inside looked great. It would have been easy to simply accept the proof and get giddy at the prospect of it being packaged and sent on its way. I didn't though, because, well, life is never that simple. I did cuddle the proof some more and alternatively stared adoringly at the front cover, and eventually got down to checking the front and back for errors. I then started on the interior layout. It occurred to me very quickly something was wrong with the print. I cornered Prid, handed her the book and told her to read a page of two. She said all the right things that included lots of wows, and wonderfuls. And then cautiously closed with the statement: 'I can see dots in the text.'

There were indeed dots in the text. You only really saw them while reading in a well lit room, but it screamed: 'Amateur hour'. One of the things I have worked very hard to do, is make Chasing Innocence look like a commercially produced book. Not looking self published is not a matter of vanity for me, its about the authority professionally published books come with. If you trust the book as a polished product  then that allows you to buy into the content. Self Published books come with an expectation there will be a lesser standard. As a consequence you hold back from fully committing to the story because you are looking for the errors. I don't want this AT ALL.

So I didn't want dots. I pondered a lot of causes. Maybe the printer budgets for a certain amount of ink per page? Maybe the quality of print won't be the same as those used by commercial publishers? Except I was using one of the biggest and most respected printers in the industry: Lightning Source. So was there something wrong with the PDF bookblock submitted by the typesetter? As a publisher it is my responsibility to make sure the bookblock, the book's interior text,  meets the printer's requirements. As with every stage of this process there had been complications, three different versions of the bookblock had been produced. For the final incarnation I had checked the chapters were in sequence and that the bookblock contained the whole book, I completed some technical checks and submitted it along with the book cover. A few days later the proof had arrived. The next day a colleague picked the proof off my desk. The colleague is Tim. He has worked in the printing industry and has lived the creation of this book on a weekly basis for the last two years. Once he had confirmed the cover was sensational he flipped it open and took a look inside. 'It's printed in greyscale.' he said after two seconds of contemplation. 'Have you noticed the dots?'

Greyscale text printed in Black and White

Greyscale would be a word that haunted my every thought for the next 7 days. The book's contents are printed in black and white, which can't translate greyscale to the printed page, which caused the dots. But how had I got greyscale in the first place? The typesetters had used QuarkExpress v9 (QE9), an industry tool for formatting the book's contents. I had the source layout file used by the typesetters. A quick download of the trial version of QE9 confirmed the source text was a perfect black. So converting the layout to PDF had somehow changed black text to greyscale. But how, why?

In trying to replicate the problem I hit a brick wall of printing terminology. But I do subscribe to lynda.com, which I can't recommend highly enough. I spent three hours that night watching through publishing printing tutorials. This gave me the basic understanding for producing the bookblock PDF for printers. Creating the bookblock through QE9 showed varied results. Sometimes the output text was greyscale, other times black. So we come back to, why? This is where I learned about the difference between RGB and CMYK. The former is a three colour standard we see every day, it is used to produce colour on our computer and TV screens. It is the colour used by QE9 to produce the black text on the screen. CMYK is a four colour standard used by the printing industry. Lightning Source, my printers, had very specific requirements for bookblock creation. I realised the typesetters had produced a bookblock that exceeded these standards but in doing so QE9 had converted black RGB to greyscale CMYK. The interesting irony was it didn't do this every time. So how did I produce a bookblock that would work?

I spent the next five days juggling Christmas duties and learning everything I could about the different standards required for producing print ready bookblocks that weren't going to surprise me when they landed on the doormat. This included more hours on lynda.com and paging through the venerable Aaron Shephard's website and the numerous books of his on publishing I own. Of course his books contained most of the answers but before the detail had meant little. I now had context and studied the semantics of producing a printer friendly bookblock. Aaron Shephard's key advice was to use the Adobe Distillor for creating Adobe PDF files, and leave as few decisions on output to QE9 as possible. I researched Lightning Sources' exact requirements for printing standards and finally I had output that had black text. I then followed their specific requirements for checking the content is compliant and then I uploaded it. That was last night. The new proof should arrive sometime next week. Failing any further 'learning opportunities', I will be clicking 'Accept' with a huge sigh of relief.