k2pdfopt: optimizes PDFs for viewing on e-readers

[Objective]

Glad I found this. It works well and is quite simple. Purchased an ebook version but it didn't perform well converting it in Calibre-text was indented and spaced after every two lines. This did the trick blocking the text nicely. using Calibre again tidied it up. However, you have to provide your own metadata - book cover, etc, as Calibre can't read it.
Thanks a million

[willus]

Originally Posted by Objective

Glad I found this. It works well and is quite simple. Purchased an ebook version but it didn't perform well converting it in Calibre-text was indented and spaced after every two lines. This did the trick blocking the text nicely. using Calibre again tidied it up. However, you have to provide your own metadata - book cover, etc, as Calibre can't read it.
Thanks a million

I'm glad you find k2pdfopt to be useful. I am curious what you mean by "using Calibre again tidied it up." Can you post examples of the document before and after "tidying up"? How does Calibre improve it?

[Objective]

Originally Posted by willus

I'm glad you find k2pdfopt to be useful. I am curious what you mean by "using Calibre again tidied it up." Can you post examples of the document before and after "tidying up"? How does Calibre improve it?

Using Calibre after using the formatted k2pdfopt file blocks the text by reducing the font size leaving it nicely justified. Here are examples you asked for.

Before (after using k2pdfopt)
Doc2 EBW.doc

After (after using Calibre)
Doc2EBW(2).doc

I guess this works just as well by dragging or pasting the k2pdfopt file straight into your ereader without using Calibre, so I assume Calibre doesn't make a great deal of difference, although I haven't tried it.

I've experimented with another document. However, either due to the font scale or lack of margin customizing, some of the lettering is cut or omitted at either side of the k2pdfopt formatted document. Are there any ways to resolve this? Thanks

[willus]

Originally Posted by Objective

...I've experimented with another document. However, either due to the font scale or lack of margin customizing, some of the lettering is cut or omitted at either side of the k2pdfopt formatted document. Are there any ways to resolve this? Thanks

Yes. Use -m 0 (or select 'm' at the interactive menu and then put in 0).

[willus]

Originally Posted by Objective

Using Calibre after using the formatted k2pdfopt file blocks the text by reducing the font size leaving it nicely justified. Here are examples you asked for...

I've sent you a private message about this.

[Vanilla]

Hi, first let me tell your k2pdfopt is the greatest pdf modifier for mobile readers that can be

I don't know if this seems to be a useful feature (or implementable one), but could the vertical space between page breaks be made so that it would depend on the font size of the text after the break, compared to the font before the break (or, instead of font sizes - the height of those lines)? This would be very helpful when there is a page break between chapters, contents page etc. so that:

a) no unnecessary space would be added between normal text at page break (lines continue to be 'merged' so to speak in the output pdf)
b) some vertical space would be added when a new chapter starts, for example (otherwise, it looks pretty annoying)

Example for b)

...
lorem ipsum some text from chapter 1 ends here.
CHAPTER TWO. ANNOYING LACK OF SPACE ABOVE

And thus begins the new chapter with new text.
...

And TY again for a great software program!

[willus]

Originally Posted by Vanilla

I don't know if this seems to be a useful feature (or implementable one), but could the vertical space between page breaks be made so that it would depend on the font size of the text after the break, compared to the font before the break (or, instead of font sizes - the height of those lines)? This would be very helpful when there is a page break between chapters, contents page etc. so that:

a) no unnecessary space would be added between normal text at page break (lines continue to be 'merged' so to speak in the output pdf)
b) some vertical space would be added when a new chapter starts, for example (otherwise, it looks pretty annoying)

Example for b)

...
lorem ipsum some text from chapter 1 ends here.
CHAPTER TWO. ANNOYING LACK OF SPACE ABOVE

And thus begins the new chapter with new text.
...

The default mode already does this. See the two attachments (I used the -c option to get a color conversion since the pages are color-coded to make the output easier to decipher). The spacing between contiguous text across pages needs a little work--it's not quite right. I'll have to look into that. But it does put a bigger space in front of "Chapter 2". Do you have an example where it's not working the way you expect?

Edit (9 Jan 2013): Okay, when I convert with "-mode fw", there is no gap put in front of "Chapter 2," so I see what you're talking about there. That is due to the setting of -vb -2 that is applied by -mode fw. That setting causes each page region to be treated as a "black box"--that is, I don't look inside of the page regions to try to determine font sizes, line spacing, etc., so I don't have any information on what spacing to put between regions. I'll think about ways to improve this.

[Objective]

Originally Posted by willus

I've sent you a private message about this.

I didn't use any options to customize the document from the interactive menu. I just pasted it into the .exe and pressed 'enter.'

I can easily send you the entire k2pdfopt doc. The formatted Calibre version, however, is somewhat difficult. It only shows up as a pdf unsupported document. I can still send you it, I think, if you know how to access it as a readable pdf doc. Maybe you're computer savvy in this way. If you know how to do it, it would be useful to know how to unblock unreadable documents.

I'm sending this from here because I haven't made an effort to find your private message after re-logging in. Hope you don't mind.

[Objective]

Originally Posted by willus

Yes. Use -m 0 (or select 'm' at the interactive menu and then put in 0).

I've tried and it works. Thanks a lot

[willus]

Originally Posted by Objective

I'm sending this from here because I haven't made an effort to find your private message after re-logging in. Hope you don't mind.

It's fine--I was trying to avoid having too many back-and-forths in the public thread. You can attach the docs in a reply to this thread or send them to me privately. I'd like to see the Calibre file even if it's in a strange format. FYI, to get to old messages after you login, just go to the upper left of the page, select "User CP" (User Control Panel) from the left side of the top blue bar, and then, on the left-side menu of the control panel page, under "Private Messages" (towards the bottom), select "List Messages."

[Vanilla]

Originally Posted by willus

Edit (9 Jan 2013): Okay, when I convert with "-mode fw", there is no gap put in front of "Chapter 2," so I see what you're talking about there. That is due to the setting of -vb -2 that is applied by -mode fw.

Hi and thanks for answering! I'm not explicitly using this fw mode, but perhaps somehow that vb option gets activated. What i normally use to convert pdfs for my kindle (these settings seem to work for most pdfs, even poorly scanned) are these options: 1 column, landscape mode, no wrapping, 0 margin, 0 output margin (i eliminate page numbers etc. beforehand using another software); sometimes, if the original page width is small, i use portrait mode. I set the options from the menu (interactively), not from the command line, after dragging the pdf to the k2pdfopt icon (which has set the "-bpc 4" option only).

And an example:



I'll try to explicitly set the vb option, to see if changes the output. TY.

[willus]

Originally Posted by Vanilla

Hi and thanks for answering! I'm not explicitly using this fw mode, but perhaps somehow that vb option gets activated...

Any chance you can attach the source file, or attach just a couple pages from it (involving a chapter change)? Then I can try out your options with that file ane make sure I understand what's going on. K2pdfopt might be treating the "Chapter One" piece as a figure since it has that black bar which will prevent a line break. It's hard to tell without having the source.

[Vanilla]

Originally Posted by willus

Any chance you can attach the source file, or attach just a couple pages from it (involving a chapter change)? Then I can try out your options with that file ane make sure I understand what's going on. K2pdfopt might be treating the "Chapter One" piece as a figure since it has that black bar which will prevent a line break. It's hard to tell without having the source.

Sure, here it is
(and i presume it's seen as image, but i think i saw the same behavior in other pdfs; also, perhaps some space before an image does no harm or even helps, at least if the image still fits the page):

sample.pdf

( Perhaps treating an image like a "tall line" i.e. no matter if it's text or not? Just speculating of course)

[willus]

Originally Posted by Objective

Using Calibre after using the formatted k2pdfopt file blocks the text by reducing the font size leaving it nicely justified....

Thanks for sending me the files. The k2pdfopt output PDF file consists of images which are 563 x 739--the default output size from k2pdfopt, optimized for the (older) kindle 6-inch screen. The Calibre epub file consists of images which are 512 x 672. These images are simply the images from the k2pdfopt PDF file re-scaled to 512 x 672 with no special processing. So if the Calibre epub file looks better on your reader, it's either because (1) 512 x 672 is a better resolution for your reader, and/or (2) the epub reading software on your e-reader does a better job than the PDF reader on your e-reader at re-scaling bitmaps. Or maybe your e-reader cannot display PDF files at all? I don't know.

If the reason the Calibre epub file looks better is due to (1) above, and your e-reader can directly read PDF files, then you can avoid having to process the k2pdfopt output with Calibre by adjusting the width and height values when you use k2pdfopt (select 'd' from the interactive menu or use the -w and -h command-line options, or maybe just increase the output resolution by using a value for -dr that is higher than 1, e.g. -dr 1.5 or -dr 2).

[Vanilla]

Made some more tests (on that sample) - basically, only when wrapping is enabled and effective (e.g. in portrait mode, when it actually splits the lines), there is some nice space before the chapter title.