Is there any easier way to do this?

[fozzy40]

Hello all!

I just got my Kindle 3 and love it! I read a lot of scientific papers with 2 columns. To get around this, I've been copying the text from a PDF, pasting it in Word, saving it as a new PDF (without pictures and columns), and then using Calibre to create a mobi.

Is there any easier way to do this?

fozzy40

[JaneD]

I think creating a second PDF is an extra step you don't need. My K3 reads word docs when I send them via email from Google Docs, so maybe you could just save the Word Doc and grab it with the "Add Books" button on calibre, then send to your k3. Or convert to mobi and then send to your k3.

If that doesn't work, post your question on the calibre forum. Those guys will know what to do. PDFs are notoriously hard to convert.

[Leep]

Originally Posted by fozzy40

Hello all!

I just got my Kindle 3 and love it! I read a lot of scientific papers with 2 columns. To get around this, I've been copying the text from a PDF, pasting it in Word, saving it as a new PDF (without pictures and columns), and then using Calibre to create a mobi.

Is there any easier way to do this?

fozzy40

I'd recommend installing the free version of Mobipocket Creator Professional. I find that it converts PDF to mobi as well as anything else that I've found.

cheers

[robertlz]

My experience:

Send PDF to your free kindle email account and use "Convert" as the subject. (without quotation). I found it works well, through some images might be ignored.

[curstpriest]

Acrobat reader (Free edition) -> Export Text (gives you text file without images, text flows and is well formatted), then import that to calibre to turn to mobi.

If you have Acrobat PRO you can go -> Export HTML, and then import that directly into calibre. It does a decent job if the images do not contain SEARCHABLE text. If the text is raster, you're fine. The mobi will contain images and proper formatting/tables if you do it this way.

Good luck.

Oh
Do not use mobi pocket creator, it uses pdf2html and butchers complex pdfs and destroys tabular formatting
again: do not use mobi pocket creator do not use mobi pocket creator do not use mobi pocket creator

[Ortep]

Originally Posted by curstpriest

Oh
Do not use mobi pocket creator, it uses pdf2html and butchers complex pdfs and destroys tabular formatting
again: do not use mobi pocket creator do not use mobi pocket creator do not use mobi pocket creator

Mobi Creator is the ONLY one I know that does a decent job. All other make a mess of the conversion. And beleave me I tested a lot of them. For example: I never got a decent conversion with Acrobat

[curstpriest]

Try acrobat pro extended 9

I can explain a little.. of how this software works.
Mobipocket creator hacks up PDF file in an unintelligent manner. If it is a SIMPLE file, it will get the job done, in an OK manner. If the file is very old and simple and THROWN into a PDF container, it will work, but other versions will also be available, because it was batch converted into PDF as well as RTF, DOC, MOBI, EPUB, etc.

Mobipocket does this.
1) Extract all image files (raw image files dumped to directory)
2) Text reflowed logically from containers into columns to form ONE solid body of text
3) Images are reinserted into body

Acrobat does this (HTML + CSS).
1) Screenshot images (in context) containers near text (may capture text too, charts/graphs/tags/labels)
2) Text reflowed logically from containers into columns to form many paragraphs and blocks of text
3) Tables and layout tagging is generated
4) Images, tables, paragraphs combined into body

Mobipocket will give you PDF -> HTML, but you won't get much formatting.
Acrobat will give you PDF -> HTML, and you will get size, centered text, tabbing, tables, etc.

Acrobat may handle images strangely in older PDFs or batch converted PDFs. If you have a graph with axis labels, it will TRY its best to incorperate them into the image, but you may have some digits cut off, so your paragraph may look like this:

THIS IS A GRAPH:
10
9
[image that shows a graph with vertical axis that has part of 9 cut off, and no 10]

Mobipocket will show this:

THIS IS A GRAPH:
10
9
8
7
6
5
4
3
2
1
[Image that shows no text labels on either axis]
1 2 3 4 5 6 7 8 9 10

You can experiment. Ultimately it is the lesser of two evils. I have found that acrobat does a better job on complex documents. But only Acrobat 9 Pro can handle images, and for that you have to pay $$ or obtain it through other means.

[oecherprinte]

Acrobat Pro Extended 9 costs almost 600 bucks (http://www.amazon.com/Adobe-Acrobat-.../dp/B0018R0AY0) as opposed to Mobipocket Creator which is for free.

So for the average ebook reader, the acrobat solution is not really feasible. Or am I missing something here?

I used Mobipocket Creator and it destroyed a lot of formatting but it was capable of handling two columns. I do not know what happens to equations as I was just dealing with legal texts.

[curstpriest]

There are alternatives to acrobat that do what it does with HTML conversion, I just don't remember the names. I found one good one a while ago that cost 15$ or so for single license. Exported files as HTML3 as single pages or combined, images were not whole background images/page.. Can't remember the name of it now. Was one of the many PDF to HTML converter programs out there (There are so many!).

Also, you can probably get Acrobat Pro 9 for like 90$ if you can grab the student version. I would only suggest paying for it if you are in need of using it for other things, such as scanning books and converting them (OCR) or actually creating/reformatting PDFs.

NitroPDF is also a decent program that has some good features, but its OCR has issues.

When I have time later I'll look into working on writing a C/C++ PDF -> HTML converter if I can't find one if there is enough interest. Probably can just grab the old pdf2html source code that has been around forever.

[emellaich]

Originally Posted by fozzy40

Hello all!

I just got my Kindle 3 and love it! I read a lot of scientific papers with 2 columns. To get around this, I've been copying the text from a PDF, pasting it in Word, saving it as a new PDF (without pictures and columns), and then using Calibre to create a mobi.

Is there any easier way to do this?

fozzy40

So, I know you don't want to hear this after just having bought a K3, but I find an iPad to be much better for pdf's. I prefer my K2 for fiction, and my iPad for journal articles.