Need some conversion help

[Ticallion]

I know this is a well discussed topic already, but I could use some advice on converting .pdf files or at least what would be the most conventional way of doing so for the ebookwise.

So far I've tried a lot of methods. I even went so far as to torrent some commercial software, which I don't condone as I only needed to test, and still have had a lot of bad results. I tried Mobipocket Creator, Nitro PDF Pro, PDF Transformer, PDF Ripper and PDFRead so far. Nitro PDF had a lot of nice options, but the converted output was always messed up no matter what options were tested. It always had a problem of spacing and squishing certain lines, which seems to be a problem with almost all these tools.

With images intact, I had the best results with PDF Transformer (though it always required lots of tweaking), unfortunately the software is miserably slow even when converting text only so it's not even worth it. Mobipocket seems even better than commercial software for straight converting to html then to the ebw but it only does text.

So I was wondering, wouldn't Adobe Acrobat probably do the best job since it's basically the main software when working with PDF files? I use Foxit, but that's only for reading and it can't do any conversion except to txt format which is bad. I never liked Adobe and any of its bloatware, but would be willing to try it if it can do perfect or near perfect conversions of PDF's that'll look at least pretty good on the ebw.

Also, is it best to convert to RTF format then to imp? Seems to produce the best results.

[nrapallo]

NOTHING complex converts properly from .pdf to non-pdf ebook format! Nothing, as I've found out while developing/maintaining PDFRead.

If the .pdf ebook was created from a Word (or similar) document, then Mobipocket Creator seems to be your best bet (and it's free), otherwise when dealing with just "images" of pages NOTHING works easily.

At best, an OCR solution "may" yield some results, but expect to do a lot of (re)formatting/correcting.

So that leaves "image" based solutions like PDFRead and PaperCrop (using the pi algorithm) to handle and preserve the look and feel of the page. This, however, only works well when your ebook device supports higher resolutions than 320x480 (EBW1150), 480x640 (REB1200) or even 1024x768 (iLiad v2). But to be honest as the resolution increases, there is NO need to convert the .pdf; just display it!

Basically, IMHO, 5", 6" or even 8" ebook devices cannot properly display .pdf ebooks without some type of zooming and/or panning.

Your best bet is to avoid .pdf ebooks, if you have just an EBW1150.

[Ticallion]

That's a bummer. I forgot to mention my experience with PDFread. Basically every PDF I tested in it when converting to IMP, either a first sentence or last sentence would be cut off. So I gave up after some of that.

So basically, any PDF's with images and such will have problems converting in every software out there? Even in Acrobat Pro? It sucks because I have a lot of books in pdf format and I've already converted mostly all of them, I'd just like to have the images intact. There's also a problem with quite a few of them having margins and such squeezed together where there should be spaces and what not.

[nrapallo]

Originally Posted by Ticallion

That's a bummer. I forgot to mention my experience with PDFread. Basically every PDF I tested in it when converting to IMP, either a first sentence or last sentence would be cut off. So I gave up after some of that.

Perhaps you were not using the right profile for the EBW1150? If you have a sample .pdf, I can try to see what settings work best with it and with PDFRead. PM or email me at: nrapallo (at) yahoo.ca if you are interested.

So basically, any PDF's with images and such will have problems converting in every software out there? Even in Acrobat Pro?

It sucks, bigtime! Acrobat Pro is not worth it at all, unless the more recent versions like 9.x have improved a lot. My experience was that it wouldn't do OCR on images nor was able to intelligently extract text from within .pdf files. Just useless; these Adobe programmers!

It sucks because I have a lot of books in pdf format and I've already converted mostly all of them, I'd just like to have the images intact. There's also a problem with quite a few of them having margins and such squeezed together where there should be spaces and what not.

IMHO, the only way to preserve the look of a .pdf ebook is to extract just an image of each page and then somehow slice that image so that it can be (optionally) rotated and enlarged on smaller ebook screens. If there is a large white margin, then the chances improve on getting legible text, but remember it's just an image of text; not actual text that can be highlighted/looked-up!

If copyright is not an issue, can you post an example of your successes and/or failures. Perhaps there's a way to salvage something...

[Ticallion]

Wow, I can't believe even Adobe doesn't do a proper job. Well I guess messing with commercial software is a waste then.

As for the PDF's I'm working with, unfortunately copyright would be an issue. See I just recently got into ereaders, so the first thing I did was basically go around finding ebooks (rapidshare, etc.) of all the books I own in paperback. Whether or not that is justifiable would be up to you, but I don't believe in buying the same product twice just to have it in different formats. I've only downloaded ebooks of ones I own on paperback though, if that makes it seem any more... acceptable? Just the problem is most people upload novels and such in PDF format, rarely ever in just html and those which convert nicer.

Either way, thanks for the help here. I'll just stick with mobipocket then I guess and do without images, no big loss.

[nrapallo]

Originally Posted by Ticallion

Wow, I can't believe even Adobe doesn't do a proper job. Well I guess messing with commercial software is a waste then.

That's exactly what I found out and now only use Mobipocket Creator or PDFRead (or similar freeware).

As for the PDF's I'm working with, unfortunately copyright would be an issue. See I just recently got into ereaders, so the first thing I did was basically go around finding ebooks (rapidshare, etc.) of all the books I own in paperback. Whether or not that is justifiable would be up to you, but I don't believe in buying the same product twice just to have it in different formats. I've only downloaded ebooks of ones I own on paperback though, if that makes it seem any more... acceptable? Just the problem is most people upload novels and such in PDF format, rarely ever in just html and those which convert nicer.

My main reader, REB1200, supports color and a bit bigger screen than the EBW1150. The extra size increase makes it is almost legible when dealing with Letter-sized .pdf ebooks, but only when I rotate and slice them. As an example what PDFRead can do, I offer below just a sample of THE .pdf ebook that STUMPED me when trying to convert from .pdf to non-pdf ebook format for reading on my EBW1150 or REB1200. That ebook is freely available from here and caused me to become interested in the PDFRead project!

Either way, thanks for the help here. I'll just stick with mobipocket then I guess and do without images, no big loss.

Well, I too prefer ebooks with images, and Mobipocket ebooks available in our E-Books Upload forum do have many such examples. Just use Mobi2IMP to properly convert them into .imp ebooks and you'll be somewhat sated.

Please note that the .imp ebooks ending with '_1150.imp' are for your EBW1150 and the ones wnding with '_1200.imp' are for my REB1200. Both can be viewed on a computer by double-clicking them after downloading them (this assumes though that you have eBook Publisher from ETI installed).

[Ticallion]

That's basically the same kind of output I had when using PDFread. I can see on page 5 that the bottom sentence is cut off as well as on page 6 the top sentence is cut off. That's no problem really with magazine scans like this, but for reading a long novel and having that on every page is just bothersome, at least to me.

With ABBYY PDF Transformer, it didn't do that. I just hated how a lot of the images weren't selected properly before converting so you'd have to go through and make sure they're outlined properly. Was just hoping for something simple that had everything done for you before converting.

You know, this kind of thing won't be an issue when Asus releases their ereader. That is if it supports PDF which it probably will since I guess supposedly it'll double as a mini-laptop. Seen this yet?

http://www.wired.com/gadgetlab/2009/09/asus-eee-reader

[nrapallo]

Originally Posted by Ticallion

That's basically the same kind of output I had when using PDFread. I can see on page 5 that the bottom sentence is cut off as well as on page 6 the top sentence is cut off. That's no problem really with magazine scans like this, but for reading a long novel and having that on every page is just bothersome, at least to me.

NO, that's by design! It's an overlap area that repeats on the next page so that there is NO "cut" in the middle of a sentence. Read the bottom of page 5 and you'll see that the top of page 6 starts a line or two before the text you read at the bottom of page 5 and is NOT cut on page 6. See, no information is lost!

With ABBYY PDF Transformer, it didn't do that. I just hated how a lot of the images weren't selected properly before converting so you'd have to go through and make sure they're outlined properly. Was just hoping for something simple that had everything done for you before converting.

You know, this kind of thing won't be an issue when Asus releases their ereader. That is if it supports PDF which it probably will since I guess supposedly it'll double as a mini-laptop. Seen this yet?

http://www.wired.com/gadgetlab/2009/09/asus-eee-reader

I have an ASUS eeePC 10" netbook but won't use it as an ebook reader when a dedicated model like my REB1200 can read the same ebook format. I did buy an iLiad to better allow me to read .pdf ebooks natively but even it needs to zoom and pan to be legible. It's best to get a larger device (with a 10" screen) to read those pesky .pdf ebooks.

[sassanik]

I used Amberlight PDF converter on a few books. The results were less than stellar. It basically turned all quotation marks into ? and did some other funky things.

Depending on how the PDF document was generated it is sometimes possible to copy the text and paste it into Word.

A good friend of mine suggested Open Office for PDF editing. Though I have not tried it out personally.

I have a Cool-er and a Pocketbook 360 and both display PDF books more or less correctly.

I believe that Adobe intentionally has made it hard to convert PDF's so that technicial documents cannot easily be changed.

Sorry the news is not better!

Amy

[Ticallion]

Originally Posted by nrapallo

NO, that's by design! It's an overlap area that repeats on the next page so that there is NO "cut" in the middle of a sentence. Read the bottom of page 5 and you'll see that the top of page 6 starts a line or two before the text you read at the bottom of page 5 and is NOT cut on page 6. See, no information is lost!

I can understand why you designed it that way, but for a novel it comes out looking a bit messy. I did try one more piece of software today called Able2Extract and it gave one of the worst outputs for conversions. Even so far as to mix up page orders and insert these huge blank spots. I really only have one PDF book with lots of images so it's no big deal, most others just had covers.

As for the Asus, this new device is an ereader meant to compete and squash Kindle. It isn't meant as another eeePC. Looks incredibly nice for the price it's being advertised at, supposed to come out this spring so we'll see if it follows through. If it's as nice as it looks, I'd take it over any ereader out there. Having dual page, touchscreen, and large displays at that price is incredible.

Hi sassanik, I didn't know Open Office supported PDF's. Unfortunately Word 2007 and such don't, even Microsoft themselves suggest some of the software mentioned here for working with PDF's. Surprised they don't have anything on the market to compete with Adobe and Foxit.