k2pdfopt: optimizes PDFs for viewing on e-readers

[mikeww]

hey willus, we spoke over email a few months back. great job on the new version, i like the new -native feature.

however, i'm having a problem with a pdf
http://rghost.net/42224460

ive tried both -mode fw and -mode fw -ls- and both seem to cut off the left side of the file. can you help?

the only way it seems to work is if i turn off native mode -mode fw -n- -ls-

[willus]

Originally Posted by mikeww

hey willus, we spoke over email a few months back. great job on the new version, i like the new -native feature.

however, i'm having a problem with a pdf
http://rghost.net/42224460

ive tried both -mode fw and -mode fw -ls- and both seem to cut off the left side of the file. can you help?

the only way it seems to work is if i turn off native mode -mode fw -n- -ls-

This PDF has a non-standard MediaBox for each page which apparently my algorithm doesn't handle correctly. For now, like you figured out, you can't use native mode on this PDF (unless you don't mind the clipping). I'll take a look at it and try and fix the issue for the next release. Thanks for letting me know.

[mikeww]

ok thanks man!

[gadd]

hi willus,

thanks in advance for such a hillarious program.

i am having problems every time when i try to use the command options such as mode, native pdf outcome or ocr. after i convert the pdf files to the k2opt format, i would like the pdf files to be scannable. otherwise, it would be impossible to manage the texts , make highlights and use in-text dictionary function in kindle 3. i have tried all three command lines for getting a scannable version of a pdf file, however, my each attempt resulted with a corrupted version of k2opt formatted pdf file (the letters were conflated and the orginal texture of the file was blurred). which commands should i use respectively to get a solid scannable version of a pdf?

[willus]

Originally Posted by gadd

which commands should i use respectively to get a solid scannable version of a pdf?

OCR only works in English right now--what language are your PDF files in? Can you attach an example PDF file?

[gadd]

I know that the content of the texts should be in english. Nevertheless, I happen to get exactly the same results in english written pdf files. You can see one example attached to the post.

[willus]

Originally Posted by gadd

I know that the content of the texts should be in english. Nevertheless, I happen to get exactly the same results in english written pdf files. You can see one example attached to the post.

I'm not sure what the issue is. It looks like you are using

-ocrvis st

when maybe you just want to be using:

-ocrvis s

(which is the default) so that the OCR text stays invisible (it can still be selected / highlighted in most readers). I'm not sure if there is any other issue you are having. I was hoping you'd attach the source document, not the converted document.

[gadd]

oh, i got that wrong. sorry.
here's the original one.

p.s. i chose the tesseract option when i applied ocr mode to the file.

[gadd]

i used the ocrvis s. the output is proper but the file can not still be selected.

[willus]

Originally Posted by gadd

i used the ocrvis s. the output is proper but the file can not still be selected.

I believe I have had confirmation from another Kindle 3 user that the OCR text from k2pdfopt is selectable on that device. The converted file you originally posted has selectable text using my PDF reader (Sumatra) on my PC. Can you select the PDF text in your converted files using any other PDF viewing software? Can you attach an example file where you cannot select the text?

For the source doc you posted, this worked for me, and allowed selection of the text on my PC:

k2pdfopt -mode fw -bp file.pdf

(You might want to limit the page range since it's 400 pages.)

[markom]

Originally Posted by gadd

hi willus,

thanks in advance for such a hillarious program.

i am having problems every time when i try to use the command options such as mode, native pdf outcome or ocr. after i convert the pdf files to the k2opt format, i would like the pdf files to be scannable. otherwise, it would be impossible to manage the texts , make highlights and use in-text dictionary function in kindle 3. i have tried all three command lines for getting a scannable version of a pdf file, however, my each attempt resulted with a corrupted version of k2opt formatted pdf file (the letters were conflated and the orginal texture of the file was blurred). which commands should i use respectively to get a solid scannable version of a pdf?

I still do quick OCR with external applications like Adobe Acrobat, Abbyy FineReader etc. like I used to do when there was no OCR in k2pdfopt.

[Shaffer]

Hellos! Just registerd & introduced myself, so, kind of the new face here!

I've been using K2 for a bit and being totally a non-expert I am finding some problems into preparing my wargaming manuals for use into my PRS-T2.

I've sent a PM to Willus with the files...let's see what he has to say!

[willus]

K2pdfopt v1.63 is released. It now has Unicode-16 copy/paste support for more languages when using the Tesseract OCR engine, along with some bug fixes and minor improvements. See the web site for details.

[mideel]

sorry, canceled..

---------------------------------------

please help me with the settings to read it 7" inch tablet... thanks,

also why is sometime the images split into two pages??

[willus]

Originally Posted by mideel

sorry, canceled..

---------------------------------------

please help me with the settings to read it 7" inch tablet... thanks,

also why is sometime the images split into two pages??

What 7-inch tablet are you using? Try using -w 4.2in -h 5.6in if your aspect ration is 4:3 (or just use whatever the width and height of your tablet screen are in inches). If the resolution isn't satisfactory, use the -dr option to improve it, e.g. -dr 1.5 or -dr 2. Use -f2p -1 to always fit images/figures to the device screen (for tall images). I'll add your "split image" question to my FAQ page [edit: added to FAQ page 12/21.]. Does your tablet not have panning and zooming with finger gestures?