Veröffentlicht
Why That Scanned PDF Is 400 MB (and How to Make It Smaller)
A five-page scan shouldn't weigh more than an audiobook, but it often does. Here's what's actually inside a scanned PDF, what compression your scanning app already applies, and two reliable ways to shrink one on your phone without wrecking it.
Why a Scan Is a Photo, Not Text
A five-page contract you scanned last week probably takes up more space than a 400-page novel. That's not a bug in the PDF format. It's what a scan actually is.
When you scan a document, your phone isn't reading the words on the page — it's taking a full-resolution photograph of each page and wrapping those photographs in a PDF container. PDF supports embedding raster images directly as data streams, and a scan without OCR has no embedded fonts or text characters at all, just images. A photo of a page, even a fairly plain one, holds far more data than the few dozen words printed on it. Stack ten of those together and you've got a file that dwarfs an entire ebook, which is mostly compact text and font references.
You can't select text in a fresh scan the way you can in a typed document, either — there's no text there, only pixels. If you want to scan a paper document into a clean PDF in the first place, that's a separate step. This one is about the file you already have, and how to reduce a PDF's file size on your phone once it exists.
What Your Scanning App Is Actually Doing
Most scanning apps don't just save raw photos — they already apply some compression before handing you the PDF. The format supports several ways to compress the image data inside it: JPEG-based compression for photos and shaded content, CCITT Group 4 for stark black-and-white pages, and JBIG2 for text-heavy black-and-white scans.
JBIG2 works by finding repeated shapes, mostly individual letters, and storing one reference copy of each shape plus the position of every place it appears, instead of treating every letter as its own image. On a page of typed text, where the letter "e" might show up two hundred times, that's a huge space saving — often three to five times smaller than older fax-style compression.
There's a real cost to it, though. JBIG2's most aggressive mode doesn't reuse a perfect copy of a shape; it looks for one that's "close enough." In 2013, that caused a well-documented problem with certain office photocopiers: the algorithm occasionally decided one digit was close enough to another and quietly swapped it, turning a 6 into an 8 in the scanned copy. Some archival standards now prohibit this mode outright for that reason. You'll probably never notice it on a grocery list, but it's the kind of thing you want to know about before scanning a contract.
Shrinking It After the Fact
If the PDF is already on your phone and just too big, you don't need a third-party account or an upload to a cloud service to fix it.
On iPhone, Preview can do this on-device. Open the file, tap next to the filename, tap Export, choose a format, and use the size slider before you save. No sign-up, no upload — the file never leaves your phone.
For more control, apps like Adobe Acrobat Reader offer tiered compression: low, medium, and high. Low is the safer choice for a scan with photos or fine detail. High gives you the smallest file and works well on text-heavy scans, where a little softening around letter edges won't matter once you're just reading the words. Try medium first, then zoom in on the part of the page that matters most — a signature, a stamp — before trusting the result.
The Fix Most People Skip: OCR
There's a step that solves the size problem more permanently, and people tend to overlook it: running OCR (optical character recognition) on the scan. OCR reads the shapes in the images and adds a real, searchable text layer underneath them. Here's what OCR actually does if you want the fuller picture.
OCR alone doesn't shrink a file — technically it adds data, since the invisible text layer sits on top of the same images. The size drop happens because most OCR tools recompress the underlying images at the same time, often converting a JPEG-heavy scan into a leaner text-optimized format. Net result: a smaller file that's also searchable, which a plain compressed scan never is.
Once a document like that is on your phone, Margina opens it directly — PDF, with or without a text layer — without asking you to create an account or upload it anywhere first.
Frequently Asked Questions
Does compressing a scanned PDF lose quality?
Yes, to some degree — most compression for scans is lossy. Low and medium settings mainly discard imperceptible image detail, but high compression on a photo-heavy scan can visibly soften text edges. If the document needs to stay legible at 100% zoom (a contract, a form), test one page before compressing the whole file.
Why is a 10-page scan bigger than a 200-page ebook?
The ebook is text and font references — a few kilobytes per page. The scan is 10 full-resolution photographs wrapped in a PDF container. A single photo can outweigh an entire novel's worth of text.
Will OCR make the file smaller on its own?
Not by itself. OCR adds an invisible text layer on top of the same images, so the file usually gets slightly bigger, not smaller. The size win comes from OCR software also recompressing the underlying images, which most scanning apps do at the same time.
The Short Version
A scanned PDF is large because it's a stack of full-resolution photographs, not because the PDF format is inefficient. Fixing it comes down to one of three things: compressing the images after the fact, choosing gentler compression at scan time, or replacing the images with real text through OCR. Pick based on whether the document needs to stay pixel-perfect or just readable.
One more thing before you free up that space: iOS can quietly delete downloaded PDFs to save storage on its own. Once your file is smaller, make sure it's actually kept where you expect to find it again.