How to compress a PDF without losing quality

Most PDFs balloon for the same handful of reasons: high-resolution scans, embedded fonts, uncompressed images, and the editor that produced them not bothering to clean up. The result is a 30MB email attachment that bounces, a portfolio site that takes ten seconds to render a single page, and a CMS upload that refuses your file outright.

The good news: most of that weight is recoverable without anyone noticing. Here's how to compress a PDF without it looking compressed.

Why PDFs get big in the first place

A PDF is essentially a container. It can hold vector text, fonts, embedded raster images, forms, scripts, attached files, and revision history. Two PDFs that look identical on screen can be wildly different sizes depending on what their author left inside.

• Scanned PDFs are the worst offender — every page is a high-DPI image, often saved with no compression.
• Photo-heavy PDFs (portfolios, reports with charts) carry full-resolution images even when displayed at thumbnail size.
• Embedded fonts can add hundreds of kilobytes per typeface, especially CJK fonts with thousands of glyphs.
• Merged PDFs often inherit unused resources from each source file.

Step 1: try lossless compression first

Before you touch image quality, run a basic re-encode. Pixoate's PDF compressor does this in one click: it rewrites the file with proper object streams, removes duplicate resources, and applies modern image compression where the file was using older codecs. For most office-generated PDFs, this alone cuts size by 30–60% with no visible difference.

Step 2: target the image-quality preset that matches the use case

If your PDF is going to be:

• Emailed or attached — use the "email" preset (≈150 DPI). Text stays sharp, photos look fine on screen.
• Posted to a website — use the "web" preset (≈100 DPI). Smallest file, totally readable on a screen.
• Printed at home or office — use the "print" preset (≈300 DPI). Bigger file, but photo detail survives.

The trap is reflexively picking "high quality" for everything. If the file is being read on a screen, 150 DPI is plenty — humans can't tell the difference at normal viewing distance.

Step 3: pre-shrink the source images

If you control the source — say you're assembling a PDF from a folder of photos — compress the images before bundling them. Run them through Pixoate's image compressor at 70–80% quality, then build the PDF with image-to-PDF. You'll end up with a smaller, faster PDF than running compression on the final document.

Step 4: split or reorganize if you only need part of it

Sometimes the easiest size win is sending less. If page 47 is the only one anyone needs, use split or organize to extract just those pages. A one-page PDF that someone can find what they need in beats a 200-page PDF they have to scroll through.

What compression won't fix

Scanned PDFs without an OCR text layer stay huge no matter what. If you're seeing a 50MB scan, the right tool isn't a compressor — it's OCR. Run it through an OCR tool first to add a searchable text layer, then compress; you'll lose almost nothing because the image layer can be aggressively downsampled once the text is in a separate selectable layer.

The short version

1. Run PDF compress with the "email" or "web" preset.
2. If it's still too big, check whether you need every page — try splitting.
3. If you're building a PDF from scratch, compress your images first with image compress.
4. If it's a scan, OCR it before compressing.

Most files that "won't fit" do, with the right preset.