I have massively bloated pdf files, commercial in nature. The smaller size are 150mb or so for a monthly magazine, which given the content and page count, shouldn’t be more than 20mb or so. Some similar issues (different month) are north of 800mb. With a magic ghostscript command found on StackOverflow:
gs -o downsampled.pdf
-sDEVICE=pdfwrite
-dDownsampleColorImages=true
-dDownsampleGrayImages=true
-dDownsampleMonoImages=true
-dColorImageResolution=150
-dGrayImageResolution=150
-dMonoImageResolution=150
-dColorImageDownsampleThreshold=1.0
-dGrayImageDownsampleThreshold=1.0
-dMonoImageDownsampleThreshold=1.0
-dFIXEDMEDIA -dPDFFitPage -dNOPAUSE -dQUIET -dBATCH
original_file.pdf
The 113.2mb file becomes 22.6mb, and the larger files have even more extraordinary savings. Bumping the dpi down to 90 or 75 starts to look crummy, but it becomes even more absurd still. Zooming in on the vector parts of the PDF show that these are preserved as vector… it’s only touching images.
The fatal flaw though is that transparency around words is broken, so the article headlines on the cover have ugly black boxes around them (some sort of mask, I suppose). Interior pages show the same problem. My understanding is that this is unavoidable with ghostscript, as it does its magic with postscript which does not offer these transparency features. Is there some alternative process that can give me these storage savings without messing up the transparency? Everything else is there… fonts are preserved, vectors, even apparently form fields (though that’s not important to the files I’m attempting to compress). Surely there’s some way to do this?
