How to compress a PDF without wrecking it

A PDF is a container. It can hold text, vector graphics, images, fonts, form fields, annotations and embedded files. Its size depends on what is inside and how it was produced.

There is no single “compress PDF” button that is always safe. Shrinking a PDF usually means one of four things:

1. removing content;
2. downsizing images;
3. reducing redundant data;
4. changing the format or quality of the output.

Knowing which one you need prevents the classic mistake: turning a crisp text document into a blurry image.

First, find out what is making it big

Open the PDF and check what it contains.

• Scanned pages are usually images. A 20-page scan can be hundreds of megabytes.
• Photo-heavy documents carry large JPEG or PNG images.
• Text documents are often small unless they embed many fonts or high-resolution graphics.
• PDF/A archives can be larger because they embed fonts and metadata for long-term preservation.
• Forms and annotations add data, but rarely dominate the size.

If the file is mostly images, image compression helps. If it is mostly text, aggressive image compression may do nothing.

Method 1: remove what you do not need

Before compressing, check for:

• duplicate or blank pages;
• unused attachments;
• hidden layers;
• high-resolution images that are not needed for the final use;
• unnecessary bookmarks or metadata.

Removing pages is lossless: the remaining pages are unchanged. Splitting a large report into sections can also make files easier to share.

Method 2: resize and recompress images

This is the biggest lever for scanned and photo-heavy PDFs.

• Downsample images to the resolution the document actually needs. A screen-only document rarely needs 600 DPI. 150–200 DPI is often enough for reading; 300 DPI is a common print target.
• Recompress images as JPEG for photos. JPEG is lossy, but photos tolerate it well at moderate quality.
• Keep PNG or lossless compression for screenshots, line art and text-heavy images where sharp edges matter.
• Avoid converting text pages into images just to shrink them. That destroys selectable text and can make the file larger.

A PDF optimizer can do this automatically, but the result depends on its settings. Always compare the output visually, especially small print, signatures and tables.

Method 3: optimize the PDF structure

PDF optimizers can remove duplicate objects, compress streams and subset fonts. This is often called “linearization” or “fast web view” when the goal is progressive loading, but linearization is not the same as compression.

Structural optimization is usually safe, but the savings may be modest. It will not rescue a file full of 12-megapixel scans.

Method 4: convert or rebuild

Sometimes the best route is to rebuild the document:

• Export a text document from its source app with a smaller image setting.
• Print to PDF at a lower resolution — but only for image-heavy files, and check that text stays selectable.
• Convert scanned pages to searchable PDF with OCR if you need to find text. OCR adds a text layer and can increase size slightly.
• For long-term archiving, use PDF/A only when preservation matters more than size.

Be careful with “print to PDF” as a compression trick. It can rasterize text, break links, flatten form fields and change accessibility.

What not to do

• Do not email a compressed file without checking that the recipient can read it.
• Do not assume a smaller file is a better file. A 2 MB scan that is unreadable is worse than a 10 MB scan that is clear.
• Do not compress a signed document if the signature or timestamp must remain valid. Changing the file can invalidate the signature.
• Do not upload sensitive documents to an unknown converter. Use a local tool or a service with a clear retention policy.

A practical workflow

1. Keep the original file.
2. Check whether the bulk is images or text.
3. Remove unnecessary pages and attachments.
4. If images dominate, downsample and recompress them.
5. Compare file size and visual quality at 100%.
6. Check that text is still selectable and links still work.
7. Rename the compressed copy clearly.

Our PDF workshop handles merging, splitting, rotating and building PDFs locally. For image-heavy pages, PDF to JPG can render pages at a chosen resolution, though that creates images rather than a compressed PDF.

The short version

Compression is a trade-off between size and fidelity. Remove what you can, resize images to the real requirement, and verify the result. The best setting is the one that keeps the document readable for its actual purpose.

For scans, see OCR explained.