Book scan -> pdf -> Kindle Touch - problems

rainsparade · 05-27-2012, 06:09 AM

Hi, I've got a lot of books in pdf form created by feeding actual books into a scanner (sadly destroying them in the process). I'm not trying to change the format or anything as there are too many symbols and parts with different languages for it to come out correctly (I've tried). I'm just trying to send them as pdfs to my Kindle Touch, however, I've run into a variety of problems:

1. File size too large. Most of the books are 500+ pages and this translates into about a 150-250MB pdf with no compression. Kindle's limit is 50MB. Some of the pdfs are 2000+ page textbooks.
- 'Reduce File Size' under Adobe professional's save as settings usually solves this but sometimes the file size is still too big. Also the crispness of the text is reduced.

2. Text is missing letters. This is the main problem, although when viewed on a PC the pdf is fine, when viewed on a Kindle often words are missing letters. I think this happens when the PDF has been OCR'ed.
- I've tested with Adobe ocr and Abbyy rinereader's ocr and the problem is the same.
- 1dollarscan.com's ocr seems to come through okay, anyone know how they do this?

3. Kindle Previewer won't preview pdfs. It's annoying to have to dig out the kindle every time I want to see how a pdf looks on it. Is there any alternative to Kindle Previewer?

Thanks for your time in reading this, I hope I was clear, any help would be appreciated.

frostschutz · 05-27-2012, 11:52 AM

I reduced a 14MB PDF (high res scan, only 20 pages) to 800kb just by trimming borders, scaling all pages to 1024x768 (on the kindle that would be 800x600) and reducing color depth to 16 gray levels (while making the background perfect white). Of course by doing so you lose any and all zoom since there just isn't any more detail to the picture then. But without zoom it doesn't look any different either way on the reader.

Try not to use OCR at all unless you want to go away from PDF. Of course the point is moot if you want/need search, reflow. toc could probably be done manually without ocr...

2000+ page pdf scan will probably not be possible. You'll have to split the book.

rainsparade · 05-28-2012, 01:42 AM

Quote:

Originally Posted by frostschutz

I reduced a 14MB PDF (high res scan, only 20 pages) to 800kb just by trimming borders, scaling all pages to 1024x768 (on the kindle that would be 800x600) and reducing color depth to 16 gray levels (while making the background perfect white).

Hi frostschutz, thanks for your reply.

I'm just curious how did you do this last part: reducing color depth to 16 gray levels (while making the background perfect white).?

frostschutz · 05-29-2012, 11:27 AM

very manually, using ImageMagick & Gimp...

I'm not sure if other software such as scantailor can achieve the same more automated

DSpider · 05-29-2012, 01:55 PM

Scan Tailor will output 600 dpi black and white images (around 3000x5000, depending on the book) which compresses great. But don't resize them, tho. Antialiasing adds unnecessary information and they won't be 1 bit images anymore. I don't like repeating myself very often so just read: https://www.mobileread.com/forums/sho...d.php?t=173214

Anyway, you shouldn't have scanned straight to PDF. Now you'll probably have to extract the images from the PDF because you can't just feed Scan Tailor anything. How would you do that? I don't know. Maybe print to a virtual printer that outputs PNG or TIFF images. You'll have to do some googling.

05-27-2012, 06:09 AM	#1
rainsparade Junior Member Posts: 4 Karma: 500192 Join Date: May 2012 Device: Kindle Touch	Book scan -> pdf -> Kindle Touch - problems Hi, I've got a lot of books in pdf form created by feeding actual books into a scanner (sadly destroying them in the process). I'm not trying to change the format or anything as there are too many symbols and parts with different languages for it to come out correctly (I've tried). I'm just trying to send them as pdfs to my Kindle Touch, however, I've run into a variety of problems: 1. File size too large. Most of the books are 500+ pages and this translates into about a 150-250MB pdf with no compression. Kindle's limit is 50MB. Some of the pdfs are 2000+ page textbooks. - 'Reduce File Size' under Adobe professional's save as settings usually solves this but sometimes the file size is still too big. Also the crispness of the text is reduced. 2. Text is missing letters. This is the main problem, although when viewed on a PC the pdf is fine, when viewed on a Kindle often words are missing letters. I think this happens when the PDF has been OCR'ed. - I've tested with Adobe ocr and Abbyy rinereader's ocr and the problem is the same. - 1dollarscan.com's ocr seems to come through okay, anyone know how they do this? 3. Kindle Previewer won't preview pdfs. It's annoying to have to dig out the kindle every time I want to see how a pdf looks on it. Is there any alternative to Kindle Previewer? Thanks for your time in reading this, I hope I was clear, any help would be appreciated.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
how to scan a book and make a pdf book?	kawaisoonano	Workshop	9	03-24-2013 02:06 PM
Kindle Touch - problems with heat	sparrowlight	Amazon Kindle	11	04-09-2012 08:56 AM
Some problems with kindle touch 5.0.4	bluefire1128	Kindle Developer's Corner	0	03-07-2012 07:45 PM
Kindle Touch odd charging problems	sparklemotion	Amazon Kindle	8	01-25-2012 02:55 PM
Help: Tips & Tutorials on how to debind, seperate pages & scan a hardback book to PDF	thebigalphamale	Workshop	4	04-17-2010 01:41 PM

05-27-2012, 11:52 AM	#2
frostschutz Linux User Posts: 2,279 Karma: 6123806 Join Date: Sep 2010 Location: Heidelberg, Germany Device: none	I reduced a 14MB PDF (high res scan, only 20 pages) to 800kb just by trimming borders, scaling all pages to 1024x768 (on the kindle that would be 800x600) and reducing color depth to 16 gray levels (while making the background perfect white). Of course by doing so you lose any and all zoom since there just isn't any more detail to the picture then. But without zoom it doesn't look any different either way on the reader. Try not to use OCR at all unless you want to go away from PDF. Of course the point is moot if you want/need search, reflow. toc could probably be done manually without ocr... 2000+ page pdf scan will probably not be possible. You'll have to split the book.

05-29-2012, 11:27 AM	#4
frostschutz Linux User Posts: 2,279 Karma: 6123806 Join Date: Sep 2010 Location: Heidelberg, Germany Device: none	very manually, using ImageMagick & Gimp... I'm not sure if other software such as scantailor can achieve the same more automated

05-29-2012, 01:55 PM	#5
DSpider Evangelist Posts: 450 Karma: 343115 Join Date: Nov 2009 Location: Romania Device: PW2 2014	Scan Tailor will output 600 dpi black and white images (around 3000x5000, depending on the book) which compresses great. But don't resize them, tho. Antialiasing adds unnecessary information and they won't be 1 bit images anymore. I don't like repeating myself very often so just read: https://www.mobileread.com/forums/sho...d.php?t=173214 Anyway, you shouldn't have scanned straight to PDF. Now you'll probably have to extract the images from the PDF because you can't just feed Scan Tailor anything. How would you do that? I don't know. Maybe print to a virtual printer that outputs PNG or TIFF images. You'll have to do some googling.