How to compress a PDF without losing quality

You hit "Attach" in an email, choose your PDF, and the upload bounces back with a 25 megabyte limit. Or you try to upload to a portal and it rejects anything over 10 MB. So you reach for a "PDF compressor" and start clicking around hoping the file gets small enough without becoming unreadable.

This guide explains why PDFs get big in the first place, what compression actually does inside the file, and how to bring the size down with the least quality damage.

Why your PDF is huge

A PDF is a container. Inside it can be:

• Text and vector graphics stored as instructions. These are tiny. A 50-page document of plain text might be under 200 kilobytes.
• Embedded fonts, usually a few hundred kilobytes total, often less.
• Embedded images, which can be anywhere from kilobytes to many megabytes per page.
• Form fields, attachments, JavaScript, usually small unless someone went out of their way.

If your PDF is 20 megabytes, it is almost certainly because of images. Specifically, one of three things is going on:

1. The PDF is a scan. Each page is a high-resolution image, often saved uncompressed or lightly compressed. A 300 DPI color scan of a letter-sized page can be 5 to 10 megabytes by itself.
2. Someone exported from a design tool with image downsampling turned off. The PDF embeds the full 4000-pixel photo from their source file, even though the visible image on the page is only 800 pixels wide.
3. The document has a lot of screenshots, charts, or product photos. Each one is a separate image, each takes its own slice of the file.

The two real strategies

PDF compression is not magic. There are basically two levers, and one or both is what your "compressor" tool is pulling.

1. Downsample the images

If the PDF contains a 4000-pixel photo that is being displayed at 800 pixels wide, you can resample the image to 1200 pixels and lose nothing the eye can see. The file drops dramatically. This is the single biggest win for most over-large PDFs.

The trade-off: if you push downsampling too far (say, to 72 DPI), images start to look fuzzy on a high-resolution screen and pixelated when printed. A safe ballpark is 150 DPI for screen-only PDFs and 200 to 300 DPI if anyone might print it.

2. Re-encode the images with JPG (or JPEG2000)

If the embedded images were stored uncompressed or with mild compression, swapping them out for proper JPG at quality 75 to 85 can cut the file size significantly. The text on the page does not change at all. The images get slightly softer in places no one will notice.

Scanned PDFs benefit the most. A scanner often produces TIFF-style image data inside the PDF that is many times larger than the equivalent JPG of the same scan would be.

What "compression" does NOT change

Good PDF compression leaves these alone:

• Text. It is stored as character codes, not as pixels. It cannot get blurry from compression. If your text suddenly looks blurry after compression, the tool flattened your PDF into a single big image — that is bad behavior from the tool, not a fact of PDF compression.
• Selectable, searchable content. Should remain selectable and searchable after a clean compression pass.
• Vector graphics. Logos and diagrams made of lines and shapes do not benefit from image compression, but they also do not get hurt by it.

Practical recipes

Email attachment under 10 MB

Aim for image downsampling at around 150 DPI and JPG quality 75 to 80. For a typical text-and-image business document this usually lands you between 1 and 5 megabytes. If the source was a scan, expect closer to the high end.

Web download or upload to a portal

Same range. Most online portals cap somewhere between 2 and 25 MB, and the compressed file will easily fit.

Archive or print-ready

Be conservative. Stay at 200 to 300 DPI for images, and avoid heavy JPG re-encoding. The file stays larger but does not lose detail if it needs to be reprinted or viewed at high zoom levels.

A single huge scan that needs to be small

The most leverage often comes from converting color scans to grayscale (if the original is black-and-white text on white paper) and downsampling to 200 DPI. A 50-page color scan can drop from 60 MB to under 5 MB with no real loss for a black-ink document.

Other ways to shrink a PDF

If image compression alone is not enough, there are a few other levers:

• Split into smaller PDFs. If the recipient only needs pages 1 to 5, send pages 1 to 5. Splitting tools do not change file content, just slice it.
• Remove pages you do not need. Cover pages, blank dividers, and "page intentionally left blank" pages add up.
• Strip embedded attachments. Some PDFs come with attached files (spreadsheets, source documents). If the recipient does not need them, drop them.
• Linearize and re-save. A "save as optimized" pass through a real PDF library often removes unused objects, duplicate fonts, and editing history.

Two things to watch out for

First, online PDF compressors that ask you to upload your file are reading the entire document. If the document is sensitive (a contract, a tax return, a medical record), think twice before sending it to a service you do not know. A local-first tool like the one on day2dayfile keeps the file on your computer.

Second, beware of "100x compression" promises. PDF data is not magic. If something claims to shrink a 50 MB file to 500 KB without quality loss, it is either lying about quality loss, flattening your text into an image, or both.