PDF conversion problems.

[qprfan99]

Hi gents

Now I know everybody says that there are often problems converting pdf files & they explain why & sometimes suggest remedies. Unfortunately I usually have absolutely no idea what they're talking about.

The manual, most of the time, is just complete gibberish to me (I just read something - "This expression is rather complex" - I just had to laugh; rather complex? it's a foreign language!)

Now I love Calibre & I love my Kindle & for reasons best kept to myself, .pdf files are the ones I usually have to deal with.

I'm generally not bothered too much about the outputted format (spaces etc) but lately I have had some books which give just one line a page, occasionally just a letter per page sometimes a part of a para. Some others just print it in alternate lines i.e. has to be read Line 1, then Line 3, then Line 2, then Line 4 (which removed the middle portion of the line to Line 5) & then on to Line 6.
Even I couldn't read a book like that!

So could somebody please take pity on a helpless old man? Is there a quick fix that a Luddite like me could understand & cope with?

Thanks for any help you can give
Larry

[itimpi]

I suspect that means the original PDF file was a multiple-column per page variant. At the moment Calibre cannot handle such files properly. There is new PDF conversion engine under development that should do better PDF conversions, including handling multiple-column files, but I have no idea what the ETA is for making it available.

[troll05]

Originally Posted by qprfan99

Hi gents

Now I know everybody says that there are often problems converting pdf files & they explain why & sometimes suggest remedies. Unfortunately I usually have absolutely no idea what they're talking about.

The manual, most of the time, is just complete gibberish to me (I just read something - "This expression is rather complex" - I just had to laugh; rather complex? it's a foreign language!)

Now I love Calibre & I love my Kindle & for reasons best kept to myself, .pdf files are the ones I usually have to deal with.

I'm generally not bothered too much about the outputted format (spaces etc) but lately I have had some books which give just one line a page, occasionally just a letter per page sometimes a part of a para. Some others just print it in alternate lines i.e. has to be read Line 1, then Line 3, then Line 2, then Line 4 (which removed the middle portion of the line to Line 5) & then on to Line 6.
Even I couldn't read a book like that!

So could somebody please take pity on a helpless old man? Is there a quick fix that a Luddite like me could understand & cope with?

Thanks for any help you can give
Larry

If it really is a multi-column Pdf as itimpi suggests try the Mobipocket creator.
It is freeware and quite easy to use
-Download here:http://www.mobipocket.com/en/downloa...oadCreator.asp
-Install the professional version
-On the home page select "import from existing File -> Adobe PDF"
-Select input and Output folder, then press import
-On the next window "Publication files" select "Build" in the tool bar
-Next window "Build publication" select "no compression" and "no encryption" then hit the "Build" icon and wait until conversion is finished

The results were quite satisfactory in my case, each 10 or 15 pages there are one or two lines that got mixed up, but even these were traceable as they were quite close to each other

If Mobipocket doesn't work as well, I would suggest the PDFtoEPUB Converter, which -despite his name- can also create MOBI files,also with mostly good results.This tool is so selfexplaining to use, should be no problem at all.
PDFtoEPUB was downloadable for free in a limited quantity here:http://www.pdftoepub.com/, but I don't know if still available.

If all this fails or is too complicated, set all your hopes on Kovid Goyal, who is presently working on a new PDF engine for Calibre, but will take approximately another half year or more to finish.

Hope this can help

[ldolse]

If your book looked that crappy after conversion I doubt it's even a multi-column issue, it's probably a source with some OCR. If my guess is right there's basically nothing you can do except load it up in OCR software like ABBYY Finereader or Acrobat Professional - but that's going to be a lot of work.

Did you read the sticky at the top of the forum?
https://www.mobileread.com/forums/sho...d.php?t=118605

I don't believe that is written in an overly complex/technical manner, but if it is let me know which parts are confusing and I'll try to make it simpler.

[dwig]

Originally Posted by qprfan99

...Is there a quick fix that a Luddite like me could understand & cope with?...

Quick fix? no. The original document structure is so altered, and is so many different ways, by conversion to PDF that there is no simple or universal method for converting from PDF to any format with linear flow( e.g. true ebook formats, convention document formats, ...).

Coping with the problem? The nursery rhymes we learned as children, or should have learned, often provide the necessary help. In this case its "Humpty Dumpty" (see: http://en.wikipedia.org/wiki/Humpty_Dumpty). In this case, use the following associations:

• Humpty Dumpty = the original document structure
• the Great Fall = the original conversion to PDF
• King's Horses = the various conversion tools
• King's Men = the good souls here at MR that offer help and advice

[BensonHedges]

troll05 - Does anyone know if the Calibre pdf engine will be a pdf to ePUB converter or a rasterizer like Papercrop and PDFLRF?

I am really disappointed with the pdf software that is out there. I want to be able to read scientific papers on my nook and there is just no way an ePUB can get all the tables, equations, and weird columns right. The rasterizers sort of work but are very clunky, crash a lot, and I haven't found any that work on Mac. It would be awesome if Calibre had a built in rasterizer that actually worked and was native with Calibre.

Does anyone know 1. if this is in the works and I can be hopeful? Or 2. is there a decent rasterizer like Papercrop that works for Mac?

[itimpi]

The Calibre engine will be targeted at converting to HTML based formats such as ePub and mobi.

If you want to be able to read complex technical documents then in many cases PDF is likely to be the only viable option. Having said that I have started seeing such papers authored in ePub in the first place and then they are fine. Perhaps for what you want you should be asking B&N to improve PDF support in the Nook.

[troll05]

Originally Posted by itimpi

Perhaps for what you want you should be asking B&N to improve PDF support in the Nook.

Yeah, and when you have convinced them to do so, you might ask Osama to be a nice guy again
I can't see any of the major players support PDF stronger than they do now especially when they have their own proprietary formats like B&N or Amazon

[jackie_w]

Originally Posted by BensonHedges

I am really disappointed with the pdf software that is out there. I want to be able to read scientific papers on my nook and there is just no way an ePUB can get all the tables, equations, and weird columns right.

You could take a look at Briss which is a PDF cropper and chopper with a visual interface. It is also still under active development. It produces another PDF so you should be able to retain formating. I don't know about technical papers, but it also retains the TOC in novels. It is written in Java should run under Linux/Windows/MacOSX.

[trisha.moore]

Hello all,

I too have noticed of late that many of my PDF's are suffering the same issue, ie alternating lines, random placing of letters and numbers. For a week or so I thought I had just had some bad files come my way, although one particular book I've been trying to read was very very messed up, so I opened the original PDF imported it into Calibre that I had luckily not deleted yet and quickly realized that there was nothing wrong with the original file. No alternating lines, misplaced letters, all was well.

Somehow, just by pointing Calibre to see the file it has converted it, but I never asked it to convert. Is there a way for Calibre to "See" my files and add them into the database/catalog without stripping it down and turning into some frankenfile? I'm perfectly ok with calibre being unable to currently convert certain pdf's to other formats, but I'm not trying to convert them...

Any help would be appreciated!

[ldolse]

Unless you told Calibre to do a pdf->pdf conversion (a really horrible idea) it would never do anything to your original files - Calibre doesn't attempt to manipulate pdf in any way.

It does have the ability to create a pdf though using conversion, which is why converting to pdf->pdf would be a really bad idea, as it would overwrite your good pdf with a much poorer version. Maybe you accidentally converted all your old pdfs to Calibre created pdfs?

[trisha.moore]

I haven't changed any pdf settings, nor do I see any settings that effect importing pdf's, or converting them as you mention. The only settings I see are line unwrapping on input (haven't touched it) and metadata stuff. All I am doing is adding them to the library, ie database, where no conversion should be happening until I say so.

Unless you told Calibre to do a pdf->pdf conversion (a really horrible idea) it would never do anything to your original files - Calibre doesn't attempt to manipulate pdf in any way.

Unfortunately, this is far from the truth. I tried it several times. Raw pdf is perfect, no problems. Go to calibre and add to library> open pdf from library> hardly legible. Calibre is changing everything about my files.

It's too bad I didn't notice this until after I imported almost all of my 800 or so book library, which is now decimated. I've definitely closed the book on Calibre until this is resolved, as 80% or more of my books are trashed.

[Manichean]

Do you have any non-copyrighted PDF files that you can attach for us as test cases?

[Perkin]

What are the 'open pdf from library> hardly legible' files being viewed in, if it's calibre's viewer than that's why they're illegible (the viewer is converting them to epub before viewing).
In Preferences->Behaviour, make sure you can unset the 'Use internal viewer for' pdf, the when you try to view a pdf from calibre it will open your default pdf viewer - which should show it properly.

[Starson17]

Originally Posted by trisha.moore

It's too bad I didn't notice this until after I imported almost all of my 800 or so book library, which is now decimated. I've definitely closed the book on Calibre until this is resolved, as 80% or more of my books are trashed.

No original books were hurt at all. Calibre makes copies. It never damages originals. It's also unlikely that the Calibre copies are bad either. It's more likely that you are trying to view them through the Calibre viewer, which results in a conversion, first, solely for display purposes. That conversion isn't saved. Simply turn off the Calibre internal viewer and use your pdf viewer.

I have hundreds of pdfs in my library and none have ever been harmed in any way.