Hey,

As an former Sony Reader I frequently used a program called PDF2LRF which I found to be the best PDF converter too because it specifically increased the contrast for text/ images to make them more readable on our eReader screens (which default converters don't do)

I came up with a straightforward workflow to convert PDFs to Kindle format which also bypasses Amazon's service. Note: this works best for image based/ or text +image based pdfs or if you want to preserve PDF formatted.


PDF to Kindle Workflow:
The workflow I've come up with is this.
1) Use PDF2LRF to input the PDF and output to a .zip file filled with PNG files. (I have to use PDF2LRF in command line format to accomplish this).

link: http://www.mobileread.com/forums/showthread.php?t=13135 (download the command line program). Use command:

" pdf2lrf -i INPUT_PDF_FILENAME.pdf -o OUTPUT_PDF_FILENAME.zip " (which generates a zip file of PNG images from PDF)

2) Then use the program that jbenny posted (PNG2PDF) to generate a PDF file in commandline from the Zip files

LINK: http://www.mobileread.com/forums/showthread.php?t=13135&page=18 (scroll down and download the file PNG2PDF.zip) Type this command in command line prompt:

" png2pdf OUTPUT_PDF_FILENAME.zip "

This generates a file called: OUTPUT_PDF_FILENAME_png.PDF

3) Afterwards, I run Mobipocket Creator to import PDF file and output to .PRC format which is then readable with the Kindle.

download link: http://www.mobipocket.com/en/DownloadSoft/ProductDetailsCreator.asp

Use option: Import PDF file: OUTPUT_PDF_FILENAME_png.PDF in Mobipocket Creator program (free), then after setting the Metadata, click Publish, to generate PRC file.

4) Move the generated .PRC file to your Kindle and enjoy it.
Note: you can fiddle with the PDF2LRF settings but I found the default settings worked pretty well.

Note: The advantage of this is I don't have to email the file to Amazon at all to get it converted and readable on kindle! (Since Gmail's 20 meg limit which I find quite limiting). And best of all the output is split nicely with high contrast, 2 pages/ landscape rotated which makes it as readable as on my Sony device compared to Amazon's default PDF conversion.

Summary: PDF -> PDF2LRF -> PNG2PDF -> MobiPocket Creator -> Kindle.

You don't need to convert the PNGs to PDF. You can just use the hidden Picture viewer to view the folder of pngs as a book. I am doing this because Kindle does not support chinese natively so I am forced to view it as "picture".

Summary: PDF -> PDF2LRF -> PNG2PDF -> MobiPocket Creator -> Kindle.MobiPocket Creator takes a PDF file as input. Do you even need PDF2LRF and PNG2PDF in the flow?
_____
Dennis

MobiPocket Creator takes a PDF file as input. Do you even need PDF2LRF and PNG2PDF in the flow?
_____
Dennis
I think the advantage of PDF2LRF is the reflow control. mobipocket's pdf import function is pretty primitive and it is really bad if there's embedded pictures etc. I guess that's why OP need to PDF2PND2PDF2Mobi

Since most viewers don't understand why it's better to do all this extra process work, I took a few screenshots.

Please look at attached:

PDF options that I've tried:

1)
When I first sent the original PDF directly to Amazon, all image info was lost, while the text converted fine. This sucks because maps are gone (leaving me lost)

2)
The second one, I had to convert the PDF to be image based, (turned it into a 20 meg file) and then emailed it to Amazon to be converted. The result is much better in that the result looks like the original PDF, but as you can see, the text is not optimized for the reader and it is hard to read smaller text. Also all maps (non-standard images) were badly resized to be very small making them unusable. (bad image fidelity)

3)
The third one, which I processed using the PDF2LRF -> PNG2PDF -> Mobi Creator is the best in that it replicates the look of the original PDF (image based) and more importantly, it bypasses having to attach/ email to Amazon, wait 1 hour, transfer back, which is bad for large files > 20 megs. With PDF2LRF, everything including text/borders etc... is optimized to be more readable on ebook readers then default because the program increases the contrast for text.

Also, concerning LoveBeta's response, using the PNG viewer doesn't work too well when you have hundreds of PNG images (due to the number of pages in a PDF). Having a converted .PRC file allows you to bookmark your books while images can't be.

In summary, my basic goal is to view PDFs on the Kindle as similar as the original PDF and this is the best solution I've found so far.

I don't suppose someone could post a Mac version of pdf2lrf?

How do you handle TOC generation? If no TOC, don't you find it too limiting?

Why not use PDFRead 1.8.2 located here (http://www.mobileread.com/forums/showthread.php?t=21906).

It converts directly from .pdf to .prc using a simple GUI! (GUI for Windows users only). It supports TOC, dilation of text, auto-cropping, edge enhancement, landscape/portrait modes, etc.

It is also possible to leave the converted images and .html behind after the conversion is done. Also a .opf is provided to allow further tweaks using Mobipocket Creator.

Give it a try.

nrapallo! pdfread is EXACTLY what I am looking for. Thanks for chiming in.

nrapallo! pdfread is EXACTLY what I am looking for. Thanks for chiming in.

Hate to steal someone else's thunder, but it did seem a good fit for you! Tell me what you like or don't, after you have gone through a few conversions; preferrably over at the PDFRead thread.

I have a pretty big eBook I'm trying to do this with, and I'm getting an error "Illegal System DLL Relocation - pdf2xml".

When I look at the process, it just keeps eating up more and more RAM, then moves on to the pagefile, then I get this message. So I have no way of going from PDF -> PRC right now.

Also, PDFRead is excessively slow it would take a day for this one PDF.

Anyone have any ideas?

guineapiguser, in your workflow, I couldn't even get past the first. I am completely new and I just wanted to read some of my scanned pdfs on my kindle. I used pdflrf through the link you provided. I noticed you kept saying "pdf2lrf" and not "pdflrf," is there a difference? anyways, I typed in pdflrf -i A.pdf -o A.zip, wait for 45 minutes, and then I get a zip file. What's funny is that Winrar doesn't read it, and I tried to jump directly to png2pdf and it gave me error messages. Here's what I got:
Archive: 1orif.zip
End-of-central-directory signature not fo
a zipfile, or it constitutes one disk of
latter case the central directory and zip
the last disk(s) of this archive.
unzip: cannot find zipfile directory in 1o
and cannot find 1orif.zip.zip, peri

C:\PDF2>for %A in (PNGs\*.png) do convert %
Error: Failed to open PDF file:
PDFs\*.pdf
Errors encountered. No output created.
Done. Input errors, so no output created.

So this is the second part of the workflow, I'm not even sure if I did the first part right. Please help anyone that has done this succesfully. I just got the kindle and would like to read scanned pdfs on it. Just scanned, I don't want any ocr because there are just way too many errors on there. Just wanted to read the pictures. Thanks!