PDF Compression Settings Explained: DPI, Downsampling, and Image Quality

The three PDF compression settings that matter are DPI (how many dots per inch an image keeps), downsampling (the rule for throwing away extra dots above a target DPI), and image quality (how hard each image is squeezed). For most everyday PDFs, downsampling images to 150 DPI with medium-to-high quality cuts size dramatically while still looking clean on screen.

If you have ever opened a "compress" dialog and stared at words like bicubic downsampling, JPEG quality 80, and target resolution, you are not alone. These controls sound technical, but each one answers a simple question about your images. Once you know what each setting is actually deciding, picking the right numbers takes about ten seconds.

Key takeaways

• DPI controls how much image detail your PDF keeps; lower DPI means smaller files but softer images.
• Downsampling is the rule that says "if an image is above this resolution, shrink it down to a target DPI."
• Image quality (often a 0–100 scale) decides how aggressively each picture is re-compressed.
• For screen reading and email, 150 DPI is the sweet spot; for printing, stay at 200–300 DPI.
• Compression mostly works on images and scanned pages; plain text and vector graphics barely shrink because they are already tiny.
• Always keep your original file, because compression throws away detail you cannot get back.

Why images decide your PDF's size

Open almost any oversized PDF and the culprit is the same: images. A page of text is just instructions for drawing letters, which takes almost no space. A photo, a scanned page, or a screenshot, on the other hand, is a grid of colored dots, and that grid can be enormous.

To put numbers on it: a single page of plain text typically lives in a few kilobytes, while one full-color scanned page at 600 DPI can run several megabytes on its own. A twenty-page scanned contract that way easily clears 30 MB, even though the same document typed out would be a tenth of that. The images are not a small part of the file. On a scan-heavy PDF, they are very nearly the entire file.

This is why compression settings are really image settings. When a tool asks about DPI and downsampling, it is deciding how to handle the picture data, because that is where the megabytes live. If you want the full picture of what bloats a file, Why Is My PDF File So Large? 7 Hidden Causes and Fixes walks through the usual suspects, from duplicated fonts to oversized scans.

The takeaway for now: tuning compression is mostly about deciding how much image detail you are willing to trade for a smaller file.

DPI: how much detail an image keeps

DPI stands for dots per inch. It describes how many colored dots are packed into each inch of a printed or displayed image. More dots mean more detail and a sharper picture, but also more data to store. The relationship is not gentle, either: doubling the DPI roughly quadruples the number of dots, because you are adding them across both the width and the height. That is why dropping from 600 DPI to 150 DPI can shrink an image so much. You are not halving the data; you are cutting it to a fraction.

Here is the part most people miss: a screen and a printer have very different appetites for DPI.

• Screens show images at roughly 72–150 DPI. Beyond that, your eyes cannot tell the difference on a typical monitor or phone at normal viewing distance.
• Printers can render fine detail, so they benefit from 200–300 DPI, especially for photos and small text in scans.

So if a PDF holds images at 600 DPI but you only ever read it on a laptop, more than half that data is invisible weight. Compression lets you drop it.

A quick mental model

Imagine a scanned receipt photographed at very high resolution. On paper at full size it looks crisp. On your phone screen, the screen physically cannot display all those dots, so it averages them down anyway. Lowering the stored DPI to match how the file is actually used removes data your reader was never going to see in the first place. You are not making the image worse for the way it gets viewed; you are removing detail that view could never show.

Downsampling: the rule that shrinks oversized images

Downsampling is the action that reduces an image's DPI. Compression tools usually phrase it as a rule with two numbers:

"Downsample images above 225 DPI to 150 DPI."

The first number is a threshold. Images already at or below it are left alone. The second number is your target. Anything above the threshold gets rebuilt at the target resolution, with the extra dots intelligently merged away.

You will often see the method named, too:

• Bicubic downsampling looks at a neighborhood of surrounding dots and blends them smoothly. It is the slowest and the best-looking, and it is the right default for photos.
• Bilinear is a faster, slightly softer middle ground.
• Average (or subsampling) is the crudest and quickest, fine for line art but rougher on photos.

For everyday work, bicubic is almost always the choice, and most tools pick it for you. The threshold-and-target pair is where you actually have a decision to make.

Why the threshold matters

If you set a target of 150 DPI with a threshold of 151 DPI, you will downsample nearly everything, including images that were already close to the target, which can soften them needlessly. A small gap between threshold and target (say, downsample above 225 to 150) protects images that are already reasonable while still catching the genuinely oversized ones. The standard advice is to set the threshold about 1.5 times your target DPI, which gives you that protective buffer automatically. Many tools handle this gap for you, but it is worth understanding when you have manual control.

Image quality: how hard each picture is squeezed

Downsampling decides how many dots remain. Image quality decides how cleanly those remaining dots are stored. Most PDFs compress photos with JPEG, which uses a quality scale, commonly 0 to 100 or a Low/Medium/High/Maximum label.

Lower quality numbers squeeze harder and produce smaller files, but push too far and you get artifacts: blocky patches, smeared edges, and muddy color around text. Here is a rough guide for JPEG-style quality:

• 90–100 (Maximum): Nearly invisible loss. Use for photo-heavy documents you might print.
• 70–85 (High/Medium): The everyday sweet spot. Big size savings, hard to see any difference on screen.
• 50–65 (Medium/Low): Noticeably smaller, with mild softening. Fine for drafts and internal sharing.
• Below 50: Visible artifacts. Use only when size matters more than looks.

One detail worth knowing: JPEG is especially harsh on sharp edges, which is exactly where text and line drawings live. That is why a heavily compressed scan often shows faint halos or speckles around letters while the photo regions still look fine. If your document is mostly text-on-white scans rather than photographs, a higher quality setting earns its keep; the savings you lose are small, and the readability you keep is large.

A subtle but important point: quality and DPI multiply. Dropping both DPI and quality at once can crush a file far more than you intended. Change one at a time, check the result, then adjust.

What DPI should I use to compress a PDF?

This is the question everyone lands on, so here is a direct answer.

Use case	Target DPI	Image quality	Result
Email / web / on-screen reading	150 DPI	High (≈80)	Small file, looks clean on any screen
General office / sharing	150–200 DPI	High (≈80)	Balanced size and clarity
Printing at home or office	200–300 DPI	Maximum (≈90+)	Larger file, print-sharp
Archival / legal master copy	300 DPI	Maximum	Big file, preserves fine detail
Last-resort tiny file	96–120 DPI	Medium (≈65)	Smallest, visibly softer

When in doubt, 150 DPI at high quality is the setting that satisfies the largest number of situations. It is small enough to email comfortably and sharp enough that almost no one will notice the difference on a screen. Only push higher when you know the document will be printed or kept as a master copy.

A note on scanned text specifically: if a document holds small print, fine handwriting, or numbers that must stay legible, lean toward 200–300 DPI even for screen use. Optical character recognition and human eyes both struggle once tiny characters fall below roughly 200 DPI, and the few extra megabytes are usually worth it for anything legal or financial.

If you would rather not fiddle with sliders at all, you can compress a PDF with sensible defaults already chosen, then download the result and check it looks right.

What compression cannot fix

Compression is powerful, but it is not magic, and a few realistic limits are worth naming so you do not chase the wrong fix.

Text and vector graphics barely shrink. They are already stored as compact instructions, not pixels. If your PDF is huge but contains mostly text, the problem is usually embedded fonts, duplicated resources, or hidden layers rather than image resolution.

You cannot recover lost detail. Downsampling and quality reduction are one-way doors. Once dots are merged or detail is squeezed out, re-compressing or "upscaling" will not bring it back. This is why keeping the original matters.

Already-small images give little back. Running aggressive compression on a PDF whose images are already at 100 DPI mostly produces artifacts without meaningful savings. Check what you have before you crush it.

Compressing twice rarely helps and often hurts. JPEG is lossy, so each pass discards a little more and can stack new artifacts on top of old ones. If a compressed file is still too big, go back to the original and compress harder once, rather than compressing the compressed copy again.

Scanned documents are the big win. A scan is one giant image per page, so it responds dramatically to downsampling. If you are working specifically with scans, How to Compress the Images Inside a PDF Without Re-Scanning covers how to shrink them without redoing the scan job.

A simple workflow that protects quality

You do not need to memorize the theory to get good results. This sequence works for almost any document:

1. Keep a copy of the original. Compression is destructive, so never overwrite your only master.
2. Decide how the file will be used. Screen-only, shared in an office, or printed? That answer picks your DPI.
3. Start at 150 DPI and high quality. This handles the majority of everyday PDFs.
4. Check the result at full zoom. Look at photos, scanned text, and any logos for blockiness or smearing.
5. Adjust one setting if needed. Too soft? Raise DPI or quality. Still too big? Lower one notch and recheck.

The realistic failure mode is overdoing it in a single pass: cranking DPI down to 96 and quality to 40 at once, then being surprised the document looks rough. Make one change, look, then decide. That habit alone prevents most disappointing compressions. If you want to inspect and adjust the actual pages, you can also open the file in the PDF editor and review images page by page.

Putting the settings together

The whole system clicks once you see the order of operations. Downsampling decides how many dots survive by reducing DPI to your target. DPI is the number that target aims for, matched to whether the file lives on a screen or on paper. Image quality then decides how cleanly those surviving dots are stored. Each setting is one honest trade between file size and fidelity, and none of them is a trick.

For most people, most of the time, the best PDF compression settings come down to a short rule: downsample images to 150 DPI, keep quality high, and bump both up only when the document is headed for a printer. Start there, glance at the result, and you will rarely need to touch a slider again.

FAQ

What DPI should I use to compress a PDF?

For on-screen reading, email, and general sharing, 150 DPI is the best all-around choice; it is small and looks clean on any screen. Move up to 200–300 DPI only when the document will be printed, since printers can render detail that screens cannot. Below about 120 DPI, images start to look noticeably soft.

What is the difference between DPI and downsampling?

DPI is the measurement, how many dots per inch an image contains. Downsampling is the action that lowers DPI by merging excess dots down to a target resolution. Think of DPI as the number you want and downsampling as the process that gets you there.

Will compressing a PDF make the text blurry?

In most cases text stays crisp, because it is usually stored as scalable instructions, not pixels, and compression does not touch it. Text only blurs when it is part of a scanned image, in which case lowering DPI or quality too far can soften it. Keep scanned pages at 200 DPI or higher if the small text needs to stay readable.

Why didn't my PDF get much smaller after compression?

Compression mainly shrinks images, so a text-heavy or vector-based PDF has little to give. If the file is still large, the bloat is probably coming from embedded fonts, duplicated resources, or hidden content rather than image resolution. A document that is mostly scans, by contrast, almost always shrinks a lot.

Is it safe to compress an important document?

Yes, as long as you keep the original. Compression permanently discards image detail, so you cannot undo it on the compressed copy. Work from a duplicate, compress that, and verify it looks right before you share or store it.

What image quality setting should I choose?

A quality value around 80 (high) is the everyday sweet spot, giving big size savings with no visible loss on screen. Use 90 or above for photo-heavy files you plan to print, and reserve settings below 50 for drafts where size matters more than appearance. Change quality and DPI one at a time so you can see which one affected the result.