Formatting errors in the proof copy?

Okay, I got my proof copy of my nonfiction work yesterday from Amazon and was aghast at the formatting errors. This didn't happen with my first book, and it's probably not uncommon, but wanted to get your input on the matter.

The proof copy has many errors that were not in the uploaded manuscript. Mostly:

- It removed almost all the reference numbers in the body text
- It removed all the corresponding footnote numbers at the bottom of the pages.
- It altered some spacing and spelling throughout the body text

The person working with my editor who handles the formatting and uploading tried to manually put the numbers back in, only on the proof copy the reference numbers are not small and elevated as they should be, they appear as just regular numbers next to the fact they are supposed to reference, so it looks like a typo.

Speaking of typos, the proof copy has hyphenated many common words like "opportunity" and "attributes" which were obviously not that way in the manuscript. There are a lot of spacing errors as well.

So I'm wondering...was/is Amazon just being glitchy? Am I using a bad font that's causing some hiccups? (Its 12-pt Baskerville so it's pretty common.) Are there truncating issues in the manuscript like in chapter title pages that is screwing with this?

We're going to re-upload a fresh copy of the manuscript with the adjustments made but most of these errors didn't even appear until AFTER it was uploaded and the proof copy printed. So I kind of feel at the mercy of Amazon's software...

 

What format did you upload the manuscript in -- a Word .DOCX file? I have submitted PDF files, and the proofs came out identical to what I submitted. Submitting a PDF means that KDP doesn't need to do any formattng -- that has already been taken care of on your end before you upload the file.

 

I submitted a .DOC file to my editor and I assume thats what she uploaded. She may have converted it to a .pdf beforehand. I'll find out. Since it has to be reuploaded anyway, we'll make sure it's in .pdf format.

 

...it was a .pdf they uploaded. So now back to square one.

 

That's surprising, to say the least. I wonder if their (your editors') PDF conversion utility is flaky and they didn't check it carefully before uploading it. I have run the same .DOCX file through several different PDF converters, and the results were surprisingly different.

This probably wouldn't explain why footnote references were changed from superscript to normal but ... did they embed the font when they made the PDF? You commented that Baskerville is a common typeface, and it is. But there are a gazillion versions of Baskerville floating around, and they don't all have the same font metrics.

 

Could be. I could also (should have?) converted it before sending it over to them. But this hasn't happened before, so...not sure what to make of it.

It's just strange because these errors didnt appear until I received the proof copy (I admittedly didn't check the Previewer beforehand) so I had a bit of a panicked, kneejerk reaction when I got the book.

So I went through it cover to cover, and then made some further minor adjustments to the manuscript, and we're gonna upload a fresh copy and see if it glitches up again. Hopefully not.

Aside from KDP's software glitching up the only things I can guess are either it doesnt like the font, or there's just enough truncation/page break issues to cause some errors? But that likely wouldn't explain why it put hyphens in the middle of words, removed italics, and removed numbers from both references and footnotes.

 

This is why I'm leaning toward suspecting a font metrics problem. I have gone the other way -- converting text out of PDFs so I can edit it in Word -- and found hyphenated words in the middle of lines. When I checked the original PDF, those words were hyphenated because they were at the end of a line.

I got into a discussion about -- tada! -- Baskervlle with the tech staff at the company I bought my font package from. One font package, from one font vendor, includes SEVEN (7) different "Baskerville" font sets. (And that's all on top of any Baskerville font that might have gotten onto my system from Windows or from MS Office, or from other sources.) Most of them don't include true italics, which is what led to the discussion. I haven't verified, but their opinion was that Baskerville was originally created as a title/display font rather than as a body text font, and so the originals from which they copied their fonts don't include a lot of things that would be considered "normal" when used as body text. As much as I like Baskerville, that resulted in my switching to another font.

The problem is that if your computer has one particular Baskerville and your editor's computer has a different Baskerville, with different font metrics, you can't ever both see the file exactly the same way. Even if you both have the same Baskerville font, if the KDP computer system has a different Baskerville installed it's going to have to adapt your file to use their font metrics. If you make certain that your PDF has your font embedded when you create the PDF, then KDP's software won't have to make any font substitutions.

 

My first thought is that the superscript (the numbers being small and up high) might not translate across from your word processor. I mean maybe it's done in some way that it only shows up when read in that word processor, and not by other apps.

 

That would make sense. But then my editor said she has used Baskerville many times in the past with no issue. But, somehow, someway, something didn't get copied over into KDP properly.

I'm using the same version of Word on the same computer I did my first book on, and there were no problems with it then. Different font, book size, and it was fiction, but the software/hardware was the same.

 

Yes, but being fiction, it probably didn't include any superscript or subscript. Or footnotes I'm guessing. Those might be handled in some way that doesn't translate well into whatever software they're using. I don't know of course, just guessing.

 

Yes, the superscript/subscript is the main difference. And it took the footnote numbers out of EVERY reference from the body text. And most of the reference numbers too. So there's just random footnotes in there that aren't even referenced in the body text. It just looks like shit. It's correct in the Word doc, but something got screwy between the conversion to .pdf, and the upload to KDP. The original doc is fine...so it's either the 2nd or 3rd step that's causing it.

If I have to switch fonts thats fine. Although since the bulk of the subject matter of the book is stuff that happened a century or more ago, I like the "old" look that Baskerville provides.

 

You have made multiple references to the original file being a .DOC file. Microsoft did away with that file format when they introduced Office 2007. I wonder if the old file format could be part of the problem. What version of Word are you using? If it's 2007 or newer, why aren't you saving in the default .DOCX file format? If you are still using Word 2003, there's a free utility from Microsoft that will patch Word 2003 to both read and write the newer .DOCX format. It looks like the patch is still available. It's called the Microsoft Office Compatibility Pack. You can find it here:
https://microsoft-office-compatibility-pack.en.softonic.com/

How to use it:
https://www.howtogeek.com/howto/microsoft-office/quick-tip-how-to-open-a-docx-file-in-word-2003/

As to Baskerville -- again, there are MANY versions of Baskerville out there. Does your editor have the exact same Baskerville font on her machine that you have on yours? And, when she created your PDF, did she select to have the fonts embedded?

That's why I like Baskerville, too. There are other "classic" fonts, though, such as Goudy Old Style. Palatino is a beautiful font, but I find it a little too light for use as body text. You could also look at Garamond, Sabon (if you have it), Bookman Old Style, Californian FB, or Caslon.

 

I'm using 365.

I keep calling them Word docs out of old habit I'll check on if she had the fonts embedded or not.

 

"docs" as a truncation of "documents" is quite understandable. However, what you posted was:

That's quite another matter. That identifies a specific and proprietary file format -- which happens to be obsolete.

 

Re-saved the Word 365 file, converted it to .pdf and locked it, reuploaded, and that cleared up 99.9% of the glitches. There were a handful of reference numbers/footnote numbers that were changed (a "6" instead of a "1" and so forth), but otherwise it's vastly improved from the first upload. I agree I think the font had a lot to do with it, and Amazon being a little glitchy. Hopefully it's all good now though. Looks okay on the previewer, but won't fully know until a new proof copy arrives.

 

I'm sorry that you are going through this, but it's a valuable lesson for me in that I had not regarded the printed proof copy as worth getting as long as the on-line preview looked okay. I will try not to make that mistake again.

How do you "lock" a .PDF file?

As to embedding fonts: Unrelated to this discussion, I had reason a couple of days ago to look into font embedding. It turns out that the program ("app") I use for creating .PDFs (a stand-alone program that installs as a virtual printer) doesn't embed the complete character set of the fonts used in a document, it only embeds the characters used. That shouldn't be an issue when printing, but it could potentially be an issue with editing the .PDF directly.

So then I checked into Word 2016 (which should be functionally pretty close to Word 365), and what I read suggests that when saving a Word document as a .PDF or exporting to .PDF, you are taken to the same options dialogue box and it doesn't allow you to choose any options relating to embedding. And it appears that Word also only embeds the characters actually used.

My virtual printer is a freeware program called DoPDF. Its commercial big brother, NovePDF, does allow the user to embed the complete character set of each font used in a document. This will, of course, affect the file size.

Could this have affected your file? I don't know, but it's a remote possibility.

 

It's because of issues like this that I fork over the monthly fee and use Adobe Acrobat Pro. It's the only way I could find to get my pdfs created and saved with all the fonts and formatting preserved.

@Damage718, if you ask very nicely, you may be able to send the wonky proof copy back for a refund or credit. I say "maybe," because KDP may argue the problem is with the file you sent them, and they may well be right. I just know I got credited when my proof copy came in with the cover printed cockeyed and the whole book cut at a weird, non-90 degree angle. Nope, try again!

 

Curiously, I have Acrobat Standard, and I've had better luck preserving format when using DoPDF7 than I have using Acrobat. Microsoft is generally a disaster.

 

Not at all. IIRC, to do an e-book through Amazon KDP you don't need a PDF at all. You upload a Word file for the book, and a JPEG for the cover. I don't use the same Word file for the e-book that I use to create the PDF for print-on-demand. For the e-book, I save a copy, convert the body text style to 11-point Times New Roman, remove all headers, footers, and page numbers, and go on from there.

If you don't intend to do a print version, just write in Times New Roman and call it good.

 

It'll be both in print and e-book versions, but we're still hung up on literally FIVE reference/footnote numbers. So it's equally as strange as it is annoying at this point. But if a font change will fix it 100%, so be it.

I also was not positive what she meant by "locking" the .pdf file either at first, but all she did was disable editing so nothing could be manually changed. However it worked, it fixed 99.9% of the problem. But alas, still a few tiny details. In all honesty, most readers wouldn't even pick up on it, but...I do lol

 

So just a quick update for fun...

The errors in the proof/previewer had only appeared after the manuscript was uploaded. Knowing THAT was formatted correctly, I thought to circumvent any Word/pdf conversion issues that may arise between two different computers or versions of the software and just save the manuscript as a pdf on my machine, then upload everything.

It's almost good to go. At first there were a ton of errors with the first proof, then only 4-5, and now just one...just ONE LITTLE INSTANCE where it changed a reference/footnote number from a 1 to a 2. It also knocked the front and back cover off kilter just a bit, but hopefully resizing that will take care of that issue.

So it just seems amazon is being goofy, and of course you can't see if there are any mistakes until after everything is uploaded. It's frustrating and really random and strange. Hoping this final attempt fixes all because I'm frankly sick of these little hiccups