How PDF Compression Actually Works (And Why Some PDFs Won't Shrink)

You've probably hit this wall before: you try to email a PDF and get bounced back for exceeding an attachment size limit, or a form upload rejects your file for being "too large." The instinct is to reach for a PDF compressor and hope for the best — but understanding what compression actually does (and doesn't do) will save you a lot of trial and error.

What's Actually Inside a PDF

A PDF isn't a single format so much as a container that can hold several different kinds of content: text (stored as actual characters, not pictures of characters), vector graphics (lines and shapes described mathematically), embedded fonts, and raster images (scanned pages, photos, screenshots). Each of these compresses very differently, which is the root of most PDF-size confusion.

Text is already extremely compact — a page of pure text, even a long one, rarely accounts for meaningful file size.
Vector graphics (charts, logos, line art) are also lightweight, since they're stored as instructions ("draw a line from A to B") rather than pixels.
Embedded fonts add some overhead, especially if a document embeds full font files rather than referencing standard system fonts.
Raster images are almost always the real source of file bloat — a single high-resolution scanned page or an uncompressed screenshot can be many times larger than an entire chapter of plain text.

Where the Size Actually Comes From

If you've ever compressed a 40-page PDF and watched it shrink from 25 MB down to 2 MB, that reduction almost certainly came from images, not text. Scanned documents are the most common offender: scanning software frequently saves each page as a large, uncompressed (or lightly compressed) image, meaning a 10-page scanned contract can easily outweigh a 500-page text-only ebook.

This is also why compressing a pure-text PDF (a résumé, an invoice, a Word-to-PDF export with no images) usually does very little — there's simply not much fat to trim. If your compressor claims 0% size reduction, that's often not a bug; it's a sign the file was already lean.

How Compression Actually Shrinks Images

PDF compressors typically work on embedded images in one or more of these ways:

Re-encoding at a lower quality setting — JPEG-style lossy compression discards some visual detail in exchange for a smaller file. Below a certain quality threshold, this becomes visible as blurring or blocky artifacts, especially in text-heavy scanned pages.
Downsampling resolution — an image scanned at 600 DPI (dots per inch) contains far more pixel data than most screens or printers actually need. Downsampling to something like 150-200 DPI for screen viewing (or 300 DPI for print) can cut image size dramatically with minimal visible quality loss.
Removing redundant data — duplicate embedded fonts, unused metadata, and orphaned objects left behind by editing software can all be stripped without touching visible content.

Why "Compress" Sometimes Isn't the Right Tool

If your PDF is large because it genuinely contains dozens of full-resolution photos that need to stay sharp (a design portfolio, a printed catalog), aggressive compression will hurt the document's actual purpose. In those cases, the better fix is usually structural, not compressive:

Splitting a large multi-section document into smaller files that are each individually shareable.
Extracting only the pages you actually need to send, rather than the whole file.
Removing pages that don't need to travel with the rest (blank scan pages, duplicate cover sheets).

A Practical Workflow

Check whether the bulk of your file is images or text — if it opens instantly and looks crisp at every zoom level, it's probably text-heavy and compression won't help much.
If it's image-heavy (especially a scanned document), try compression first — it typically recovers the most size for the least effort.
If compression doesn't get you under the limit you need, or the content doesn't compress well, consider whether splitting or extracting specific pages solves the actual problem — you may not need to send the whole document at once.
Always check the result. Lossy image compression is one-way — once quality is discarded, you can't recover it, so keep your original file until you've confirmed the compressed version still meets your needs.

Tools for the Job

Compress PDF — reduce file size by re-encoding embedded images, entirely in your browser.
Split PDF — break a large document into smaller, individually shareable files.
Merge PDF — the reverse operation, for combining multiple documents into one.

PDF compression isn't magic — it's a targeted trade-off between image quality and file size, and it works best when the size problem actually comes from images in the first place. Knowing what's inside your file before you compress it is the fastest way to pick the right fix on the first try.