A 40MB PDF that needs to be under 10MB for an email attachment or an upload form is one of the most common small frustrations in office life. The instinct is to compress it and hope for the best — but "hope for the best" is exactly how you end up with blurry scanned signatures and unreadable fine print.

Compression doesn't have to mean visible quality loss. It just requires understanding what's actually taking up the space in the first place.

What's Actually Making the File Large

PDFs bloat for a small number of predictable reasons, and knowing which one applies to your file changes what you should actually do about it:

  • Embedded images at unnecessarily high resolution. A scanned page saved at 600 DPI when it only needs to be readable on a screen is the single most common cause of an oversized PDF.
  • Uncompressed or lightly-compressed image formats. Some tools embed images with minimal compression applied at all, leaving obvious room to shrink without any visible change.
  • Embedded fonts. A PDF with several custom fonts fully embedded can carry hundreds of kilobytes of font data your reader doesn't need for text you could display with a standard font just as well.
  • Leftover data from editing. Some editors retain earlier revisions or unused elements inside the file itself, invisible on the page but still counted in file size.

Why "Quality Loss" Actually Happens

Image compression works by discarding information that's judged unlikely to be visually noticeable — the more aggressively you compress, the more gets discarded, and past a certain point that discarding becomes visible as blurring, blocky artifacts, or washed-out color. The failure mode isn't compression itself; it's compressing harder than the content can tolerate.

A photo-heavy page can absorb more compression before it looks noticeably different than a page of dense small text or a fine-lined technical diagram — text edges and thin lines show compression artifacts much sooner than a photograph does.

The practical rule: match your compression level to what the document actually needs to do next. A PDF being read on a screen needs far less resolution than one being sent to a professional printer. Compressing for the screen use case, when that's genuinely all it's for, isn't a compromise — it's just not wasting file size on detail nobody will use.

A Practical Approach

  1. Start with a moderate compression level, not the most aggressive one available. Most tools offer a range — jumping straight to "maximum compression" is where visible quality loss usually starts, when a moderate setting would have been more than enough.
  2. Check the result at actual size, not zoomed in. Quality loss that's obvious at 200% zoom is often invisible at the size the document will actually be viewed or printed.
  3. If it's mostly text, compression will barely matter. Text-heavy PDFs with few or no images are usually already small; if one is still large, the cause is more likely an embedded font or leftover editing data than anything compression settings will fix.
  4. If it's mostly scanned images, that's where the real gains are. Scanned documents are the case where compression makes the biggest visible difference in file size, and where testing a couple of different levels is actually worth the extra minute.

Compressing Several Files at Once

If you're regularly compressing more than one or two files — a batch of scanned forms, or several reports before archiving — doing this one file at a time through a single-file tool becomes the actual bottleneck, not the compression itself. That's a genuinely different problem from occasional one-off compression, and worth solving with a tool built for it rather than repeating the same steps manually 20 times.

A Concrete Example: A Scanned Contract for Email

Say you've scanned a 12-page signed contract and it's come out at 35MB — common with a phone camera scan or a flatbed scanner set to a high default resolution — and your email provider caps attachments at 25MB. Walking through the decision:

First, check whether it's actually a scan (a photograph of a physical page) or a digitally-created PDF that happens to be large for another reason. A scan is almost always the image-resolution case above — the fix is compression, and it'll work well. A digitally-created contract that's still large despite having no images is more likely carrying embedded fonts or leftover data, where compression settings won't help much and the real fix is re-exporting it from whatever created it, with fonts not embedded.

For the genuine scan case: a moderate compression setting will typically take a 35MB scanned document down to a few megabytes with no visible change to a signature or printed text — the resolution reduction happens well below the threshold where a human eye would notice on a normal screen or printed page. Only if you're still over budget after a moderate pass is it worth trying a more aggressive one, checking the result carefully rather than assuming it's still fine.

The Font Embedding Case, Briefly

Worth a specific mention because it's easy to miss: a PDF that's unexpectedly large despite having little to no images has often embedded one or more custom fonts in full, including every character and style variant, even if the document only uses a handful of words in that font. This is common in PDFs exported from design tools rather than word processors. Standard compression tools that target images won't touch this — the fix, when it matters, is at the export stage rather than after the fact.