Humongous PDF files

July 20, 2006 11:08 AM

Why are my PDF files so huge?

For the past couple of summers, I've created a PDF directory of my program participants by printing from my FileMaker Pro database via OS X's print command. This summer, the file is huge, like 25+MB! Nothing about the database has changed (other than an upgrade of FMP), and last summer's directory is <300KB.

I tried opening it in Acrobat 6.0.5 and choosing Reduce File Size... from the File menu, but it hung up on the "duplicate fonts" part.

Any clues? I need this done by tomorrow.

posted by ancientgower to computers & internet (7 comments total)

I don't know about the specifics on a Mac, but images can make a PDF grow fat rather quickly. Try reducing the resolution of the images to the mimumum required (72dpi for screen 150dpi for print should be enough) and see how that helps...
posted by baggers at 11:12 AM on July 20, 2006

You can also use dropload to send me the file, and I'll try to decrease the size and send it back. hundertwasser at yahoo sot com.
posted by luriete at 11:17 AM on July 20, 2006

Thanks! I'll send you one and see if you can figure it out--no point in transferring two huge files. I suspect font shenanigans but can't really figure out how to make sure.
posted by ancientgower at 11:24 AM on July 20, 2006

The PDF printing system in Mac OS X is not designed to make compact files, but to quickly and accurately describe a page. If you use an icon 750 times on one page and print 100 pages, then unless the printing application is really smart about it, your PDF file will probably include 75,000 copies of the icon.

IMHExperience, that's usually the primary bloat in big printing-system PDF files - common images or elements that are larger than they look and repeated more often than you think. It's a difficult thing for PDF optimization software to catch, too, because such optimizers basically have to examine each of the images and compare them to all the others to find duplicates.

(This is harder than it sounds. Imagine code to draw a simple rectangle. Not difficult to draw four lines but every place that same rectangle appears on the page, the coordinates of the four corner points are different, so analysis software can't easily know that it's one rectangle that could be redrawn in a bunch of places. You almost have to have a built-in PostScript interpreter to know what the routine does, if you want to figure out that several objects are really all the same object drawn in different locations.)

Anyway, the good news is that FileMaker Pro is smart about this - if you're using version 8 or later, and you make your PDF file via File > Save/Send Records As > PDF, instead of just printing to PDF. I made a quick test using 38 records that each printed as 3 pages long. Exporting to PDF from within FileMaker Pro (Advanced, but this makes no diff) 8 created a 3.1MB file. Printing the same records via "print to PDF" gave me a visually-identical 114-page file, but at a whopping 11.5MB, or nearly 4X as large. (Why? Because the printed version included a separate copy of every label, every layout widget, every rectangle; the PDF version that FMP built itself does not.)

Your fastest way to saving lots of space may be to upgrade to FMP 8 if you haven't, and then just export the PDF from the File menu. If that still makes the file way too large, then yeah, your images are too big, and you'll need to reduce the quality either in the DB or via PDF optimization software to get the file size down. (A group I work with had a 2-page brochure designed for printing, and put the PDF file of the print-ready job online. It contained a huge-resolution image for the printer; downsampling that one image from 300DPI to 150DPI cut the PDF file size from 1MB to 156KB. Acrobat or PDF Enhancer can do that for you when you have the finished file.)
posted by mdeatherage at 11:42 AM on July 20, 2006 [1 favorite]

Fixed. Either FMP or the mac print to pdf mechanism embedded the fonts (Times, bold & italic) for EVERY SINGLE LINE of type which used those fonts. Once I stripped everything except a single instance - which probably isn't even strictly necessary, as everyone reading it should have at least some version of Times - the resulting v6 PDF was under 1.5 megs.

The Acrobat "PDF Optimizer" is your friend!
posted by luriete at 12:32 PM on July 20, 2006

Try PDF Shrink too. The 30 day demo has no restrictions (apart from being, you know only useful for 30 days).
posted by schwa at 12:34 PM on July 20, 2006

Thanks, Luriete! I'm retrying it--although it seems to be hanging up on the "consolidating duplicate fonts" again.

Thanks for everyone's help!
posted by ancientgower at 12:49 PM on July 20, 2006