Why did my PDF barely get smaller?

Most likely it is mostly text or vector graphics rather than images, so there is little to remove — text compresses to very little on its own. It can also happen when the images inside were already compressed heavily before being placed in the document.

What DPI should I compress to?

Use 150 DPI as a default: it looks clean on screen, prints acceptably, and typically removes 70-85% of a photo-heavy file. Choose 72-100 DPI for screen-only documents and 300 DPI when the file is going to a printer.

Why does the text in my compressed PDF look blurry?

That is JPEG compression applied to images containing hard edges. JPEG discards fine detail, which suits photographs but smears text and line art. Keep the quality setting high for scans and screenshots and save size through resolution instead.

Does compressing a PDF lose quality permanently?

Yes, if image quality or resolution was reduced — the discarded detail cannot be recovered from the compressed copy. Keep the original file, and treat the compressed version as an export rather than a replacement.

Can I make a PDF smaller without losing any quality?

Often, yes. Removing pages you do not need, flattening unused form fields, and dropping embedded page thumbnails all reduce size without altering how any remaining page looks.

How to Compress a PDF Without Wrecking It

Almost everyone meets PDF compression the same way: an email bounces back because the attachment is too big, or an upload form refuses a file over 5 MB. You search for a way to shrink it, run the file through something, and either it barely changes or it comes back looking like a photocopy of a photocopy.

Both outcomes have the same cause — compression settings applied without knowing what is actually taking up the space. This guide explains where the bytes go and how to remove them in the right order.

Find out what is big before you compress

PDF file size is almost never spread evenly across a document. In practice it concentrates in one of three places:

Images. In the overwhelming majority of oversized PDFs, this is the answer. A single phone photo dropped into a report can be 4 MB on its own. A scanned document is entirely images — one per page.
Embedded fonts. Each font family embedded in full can run to several hundred kilobytes, and a document that uses six weights of two families carries all of them. This matters in short documents, where fonts can be most of the file.
Everything else. Text content streams, vector graphics, bookmarks, form definitions. Usually small — a hundred pages of plain text is often under 500 KB.

A quick diagnostic: divide the file size by the page count. Under roughly 100 KB per page and the document is mostly text — there is little to win, and aggressive compression will cost you quality for almost no benefit. Over 500 KB per page and it is images, which is good news, because images are where the savings are.

The setting that does most of the work: resolution

Images inside a PDF have a resolution relative to the page, measured in dots per inch. A photo placed at 3000 pixels wide across a 6-inch column is sitting at 500 DPI. Your screen will show it at roughly 100 DPI. A good office printer will use around 300. The rest is data you are carrying and never seeing.

Downsampling — reducing the pixel dimensions to match a target DPI — is the single most effective step, because the saving scales with the square of the change. Halving the resolution removes about three-quarters of the pixels.

Sensible targets:

72–100 DPI — screen-only documents, web downloads, anything that will be read on a phone and never printed.
150 DPI — the usual sweet spot. Looks clean on screen, prints acceptably, typically cuts a photo-heavy file by 70–85%.
300 DPI — print quality. Use when the document is going to a printer or contains fine detail such as small text in a scan.

If you are unsure, start at 150. It is very hard to see the difference on screen and the saving is dramatic.

The second setting: image quality

After resolution comes the JPEG quality level, which controls how aggressively each image is re-encoded. This one behaves differently from resolution: the damage it causes is not uniform.

JPEG compression works by discarding fine detail that the eye is poor at noticing. That assumption holds beautifully for photographs and badly for anything with hard edges — text, line drawings, screenshots, diagrams, charts. Push the quality down on a photograph and it softens. Push it down on a screenshot of a spreadsheet and you get visible smearing around every character, the artefact people describe as "blurry text".

So the useful rule is not a number, it is a question: what is in the images?

Photographs — quality 60–75 is usually invisible.
Scanned text pages — keep quality high (85+) and save through resolution instead.
Screenshots and diagrams — reduce as little as possible; these are the files where compression shows first.

When a PDF will not shrink

Sometimes you compress a file and it barely moves, or grows. The common reasons:

It is already compressed. PDF applies Flate compression to content streams by default and images arrive as JPEGs. Compressing a JPEG again gives back very little, and asking a zip tool to compress a PDF usually saves nothing at all.
It is genuinely text. A 200-page text document is small because text is small. There is nothing to remove.
It is vector artwork. A dense map or CAD drawing is thousands of path instructions. Image settings do not touch them.
Re-encoding at a higher quality than the original. If an image arrived as a heavily compressed JPEG and you re-encode it at quality 95, the file can get bigger. Compression is not always subtraction.

Things that shrink a file without touching quality

Before reaching for quality settings, there are lossless wins:

Remove pages you do not need. The fastest compression is deletion. If you only need pages 4 to 9, extract them — a 30-page document becomes a 6-page one and the size follows.
Flatten form fields. An unused interactive form carries field definitions, appearance streams and often a JavaScript layer. Flattening turns it into a plain page.
Drop embedded thumbnails. Some producers store a preview image of every page inside the file. They are pure overhead in a modern viewer.

A practical order of operations

Remove pages you do not need.
Flatten forms and annotations if the document is final.
Compress at 150 DPI, quality around 75, and look at the result.
If it is still too big, drop to 100 DPI before dropping quality further — resolution costs less visible damage per byte saved than quality does.
Check the pages that matter most at 100% zoom, not zoomed out. Artefacts hide at small sizes.

Compress PDF exposes both controls and shows you the before and after size, so you can try a setting, look at the number, and adjust rather than guess.

If you are fighting an email limit

Mail systems typically cap attachments around 20–25 MB, and they apply the limit after encoding, which inflates the file by roughly a third. So a 20 MB limit means you want to be under about 15 MB to be safe. If you cannot get there without visible damage, split the document and send it in parts — two clean halves beat one mushy whole.

In short

Look at the file size per page first. If it is images, downsample to 150 DPI and you will usually be done. Save through resolution before quality, remove pages you do not need, and check the result at full zoom before you send it. If the document is text, accept that it is already as small as it is going to get.