Converting books to epub

[Pam R]

If you have a book in a different format you can use any of a number of web sites to convert it online

[j.p.s]

Originally Posted by Pam R

If you have a book in a different format you can use any of a number of web sites to convert it online

Most people here would use calibre.

Or for amazon format books, kindleunpack

[Pam R]

Originally Posted by j.p.s

Most people here would use calibre.

Or for amazon format books, kindleunpack

Thought it was worth mentioning

[Deskisamess]

I've used a couple of online conversion sites when not at home with access to my computer. They are great in a pinch, but it can take several hoops with CC and Dropbox, and back and forth.

At home, Calibre is easier and better for management.

[Bookstooge]

Originally Posted by Pam R

If you have a book in a different format you can use any of a number of web sites to convert it online

How are the results?

[jackie_w]

That site is actually using some version of calibre to do the conversion. Who knows how old the version is, though?

[Pam R]

Originally Posted by Bookstooge

How are the results?

It is all I have ever used and never had a problem with it .

[DNSB]

Originally Posted by jackie_w

That site is actually using some version of calibre to do the conversion. Who knows how old the version is, though?

I ran a Gutenberg mobi of Three Men in a Boat through it and the content.opf says calibre 4.8.0 was used for the conversion.

<dc:contributor opf:role="bkp">calibre (4.8.0) [https://calibre-ebook.com]</dc:contributor>

One issue I noticed was that using either the website or calibre locally, the code for 67 images was preceded by an invalid <a> and included an invalid tag and the 4 footnotes had a similar issue:

<p class="c"><a tag="{http://www.w3.org/1999/xhtml}a"><img alt="Decorative graphic" title="Decorative graphic" src="../images/00001.jpeg"  tag="{http://www.w3.org/1999/xhtml}img" class="calibre5"/></a></p>

<a class="citation" tag="{http://www.w3.org/1999/xhtml}a" href="part0005_split_003.html#footnote287">

[Quoth]

I re-convert all the Gutenberg books via Calibre. I only download the mobi + images, then make epub2 and KF8/AZW, with the deliberate Gutenberg paragraph spacing replaced by 1.4 em indent, my favourite body font embedded (some ancient ereaders otherwise limited) and smarten punctuation.

We also use Calibre for epub2 and dual mobi for uploads to publish our own titles (variously edited in Word or LibreOffice Writer by Authors/Editor) via export/Save As to Docx.
In both cases conversion to epub2 first, with then any extra css editing or automated changes by Calibre, then all other formats produced without changes from the epub2.

Gutenberg has their own style guide and also I've found the margins & justification on some of their epubs strange. As I need KF8/AZW (works on all from Kindle Keyboard to PW4) and epub2 (all other makes) and Gutenberg style guide isn't what I like, it makes sense to save time & bandwidth only downloading the "mobi + images" version.

Archive.org is terrible. Usually no proofing. I don't bother downloading epub/mobi if it's only their own OCR of the pdf. If it's from Microsoft or Google Books scan, then PDF is best. If it's clean I might OCR, otherwise I import to The Gimp as layer per page setting new resolution, export in reverse order as mng and then adjust margin crop, contrast, brightness and change to mono in ImageMagick (you need to add PDF write to security, don't enable PDF read). Then the PDF will be MUCH faster in an eReader or App/Program. If it's still unsuitable for the Kobo Libra (FAR better for PDF than original Kobo H2O, Kindle DXG or Kindle PW3), then I read on a 10" Lenovo Android tablet in Pocketbook. I used to use a 10" Win10 Atom tablet, but the Lenovo about x5 faster and easier to use without a keyboard/mouse. The best Win10 PDF readers don't work well with touch and don't use volume keys as page keys like almost every Android PDF or ePub reader app.

[Quoth]

Some gutenberg ebooks are or were version 3.x Calibre. Or even something else. I've been downloading them since maybe 2007 or 2008, long before I bought my first Kindle. I was very temped to a Sony ereader when I first saw it in Dublin. My wife also thought it better than LCD, even though it was very grey (compared to models even 5 years ago). Couldn't afford it!

[pdurrant]

There are so many on-line ebook conversion sites, all using calibre in the background, that I'm unwilling to allow links to any particular one.

Especially as they all involve sharing books with an unknown third party. Which isn't a good idea for in-copyright ebooks.

[Quoth]

Personally I wouldn't use any such site even for Public Domain.
Calibre is "free" (please donate) and has been trouble free for me on XP, Win7 and Linux.
Runs on MacOS, Linux and Windows.
It's far better and more flexible than any online site.
It also is the best way to manage more than 50 books, one device, several and works with Kindle, all makes & models of ePub based readers I've tried, two Android phones and two Android tablets I have, all different versions of Android.
It also does a better conversion from docx to epub2 and upload to Amazon for Kindle, than uploading doc or docx or mobi direct to Amazon.

Also this site hosts Kovid's excellent support for Calibre.
https://www.mobileread.com/forums/fo...play.php?f=166
He's courteous, helpful and unlike Microsoft, Google or Apple, he actually listens and fixes stuff!

[DNSB]

Originally Posted by FrustratedReader

Personally I wouldn't use any such site even for Public Domain.
Calibre is "free" (please donate) and has been trouble free for me on XP, Win7 and Linux.
Runs on MacOS, Linux and Windows.
It's far better and more flexible than any online site.
It also is the best way to manage more than 50 books, one device, several and works with Kindle, all makes & models of ePub based readers I've tried, two Android phones and two Android tablets I have, all different versions of Android.
It also does a better conversion from docx to epub2 and upload to Amazon for Kindle, than uploading doc or docx or mobi direct to Amazon.

Also this site hosts Kovid's excellent support for Calibre.
https://www.mobileread.com/forums/fo...play.php?f=166
He's courteous, helpful and unlike Microsoft, Google or Apple, he actually listens and fixes stuff!

↑ ↑ ↑ What he wrote.

[hobnail]

Originally Posted by FrustratedReader

Archive.org is terrible. Usually no proofing. I don't bother downloading epub/mobi if it's only their own OCR of the pdf. If it's from Microsoft or Google Books scan, then PDF is best.

I don't know if everyone else knows this but PDFs can have this clever thing (to me anyways) where there are 2 layers. The visible layer is what you see when you open it in a PDF viewer or whatever, and the invisible layer is the OCR'd text. You can tell if it has the OCR'd text if you click and drag your mouse over the text and it selects stuff, probably not exactly lined up with the visible text. If your PDF viewer supports it you can open the PDF and then do a Save As and select text as the output format and save the OCR'd text. It's very likely the exact same text that you get when you download the .txt file from archive.org but sometimes it can be helpful to access that text as you're looking at the scanned image of the page.

[Quoth]

There can be any number of layers of bitmap image, vector, postscript, java, text, fonts etc. PDF is an envelope with instructions about the page size(s), index/TOC, and how to display it all on a screen so it exactly matches the paper printout.

The text layer on most image based PDFs is just poor unproofed OCR, maybe by Finereader/ABBY etc. It's only so text search will work, which is all Google cares about.