I have a really big PDF. 1.2GB. I want to make it smaller, but it's a really pretty art book so I don't want to compromise the image quality.
How can I make it small enough to share without ruining the art? Adobe Acrobat's compression tools all give me artifacts out the ass and make the text nearly illegible.
Also if there's a way to crop the white stuff on the side of each image rather than manually removing it, that would be really great.
Other urls found in this thread:
xnview.com
commons.m.wikimedia.org
twitter.com
Just put it up on MEGA or dropbox as is.
use djvu
it's a dedicated raster format suited for images that can dramatically shrink the size without compromising quality
pdf is vector + raster
you can use ghostscript to convert between the two
So I can make a djvu file out of the originals, which are even larger than the OP image, then convert it to a PDF and it will still be a lower size?
SAM when./
Introduced in two weeks at fanfest, playable in Stormblood. Probably DPS.
>he thinks it won't be BLU
>he didn't see the datamined files
Does the PDF have text or interactivity or something you want to keep? I'd personally just export the image file of each page losslessly, then convert those to JPEG/WebP or something and distribute it that way.
>yoship has never put fake files in for dataminers
Let me rephrase.
I scanned this book at 700DPI, creating 2.20 GB of .jpg images. When I combined them into a PDF with unskew, OCR, and bookmarks, they created a 1.20GB PDF with a barely-acceptable level of artifacts.
I want to create a .pdf in the low-hundreds MB range without compromising the quality too much. Making the images smaller is fine, filling them with artifacts is not.
That's what I'm trying to do here. I'm sorry I didn't make it clear.
>use djvu
only works well on B&W text. looks like shit in color.
download a cracked copy of JpegMINI Pro (pro version because your images are too big for Lite).
that will shrink jpegs considerably. don't use the reduction algorithm found in most PDF packages because they;re shit.
then just combine jpegs into a PDF.
without
Alright. Where would I go to download this thing? I'm not exactly up on piracy, I'm sorry.
Ahh, right. I know fuck all about PDF creation, sorry. Does the program you use necessarily transcode the data from the JPG files, or can you just embed them directly into the PDF? Not sure if that's even possible, again, not familiar with the format.
I'd personally start by scanning the entire book in the highest resolution you can manage, and saving every page as PNGs in a folder. Then just use XnView or ImageMagick or something to batch convert all of them, and experiment with various settings. Figure out what the lowest acceptable resolution is where the text and the images are still legible (20-30% of the high resolution scan, maybe), then fuck around with JPEG options to lower the size further without sacrificing the quality, and make a PDF from those files.
Embeds them in the PDF, then "reads" any text on the page, creating metadata that allows PDF readers to select and copy the text.
I'm assuming 's program does all that, but automatically?
Oh, and I already did the first step, scanned at 700DPI, but the batch scanner I used saved them all as jpegs. D'oh.
Are you using Acrobat? There should be a way to embed the images without compressing or touching them in any way. If that's a possibility, all you have to do is get the ~2GB collection of scanned JPGs you have down to a reasonable size and use those to create a PDF.
Which sends me back to looking for that JpegMINI program. Anyone have a lead on it?
Just use XnView, it does the same thing and it's free. There's a batch convert option, something like pic related should be fine, plus a transformation that resizes every image to 20-30% of its original dimensions.
Oh, awesome, thank you!
Is there anything to be done about cropping them or do I have to do that manually?
What are your original scans like? There are a bunch of options including cropping which you can apply, but it would affect every image the same way. If you only need to crop/rotate some of them, you'll have to either do it manually for each page that stands out, or use the automatic shit built into your PDF exporter, which might also ruin the quality a bit.
Each page alternates the side the white space is on. I was hoping there was some kind of macro that could search for large blocks of white space in the margins and remove them.
Use .djvu, it cuts the filesize down dramatically and doesn't compromise quality. Send link for me to download pl0x
>only works well on B&W text. looks like shit in color.
Most people would proably prefer it if you didn't compress the images at all and just uploaded it as is.
Why didn't you scan to Tiff and then just turn that into a pdf? Or any archive containing jpeg (cbr?) or even webp / bpg or the other new better codecs.
Batch Processing > Transformations > Autocrop
Not sure how well it works though, you might want a Photoshop macro or something if it's too advanced.
Not entirely true, depends on the quality of the images. I once had a 100gb PDF file because I took high quality scans and turned it into a 1gb .djvu file. Then again, it was really high quality.
Alright, forget the cropping for now, the problem is the images are too high-resolution, like 5100x6600. So I need a batch resize.
Do I need any special options for this or can I just pick a percentage of that and it'll work?
Pretty much. Again, just experiment with the options to figure out what's the "worst" quality output you can get away with.
JUST TRY .DJVU INSTEAD OF PDF
Compress to webp or bpg for you own archive and not backward users. Should do far better.
Make djvu and adjust settings until small enough for everyone else?
Supporting pleb help:
commons.m.wikimedia.org