PDF compression works by shrinking the parts of a file that take up the most space — usually embedded images — through downsampling and re-encoding, while reorganizing the file’s internal data more efficiently and stripping out bloat. Some of these steps discard visual detail (lossy) and some don’t (lossless). This guide explains what actually makes a PDF large, how each compression technique works, the difference between lossy and lossless, why text-heavy PDFs barely shrink, and what DPI to choose.
What makes a PDF large?
Before you can shrink a PDF, it helps to know where its bytes go. A PDF is a container of objects, and a few of them dominate the size:
- Images. By far the most common cause of a big PDF. A single high-resolution scan or photo can be larger than hundreds of pages of text. Scanned documents — which are nothing but images — are the heaviest files of all.
- Fonts. Embedded fonts add weight, especially when the entire font is embedded rather than just the characters used.
- Metadata and overhead. Document metadata, page thumbnails, form data, annotations, and duplicated resources all add up.
- Inefficient encoding. Internal data streams that were saved without tight compression waste space.
The practical takeaway: if your PDF is large, it’s almost certainly the images. That’s why compression strategy revolves around them.
How PDF compression works, step by step
A good compressor like Compress PDF applies several techniques together. Here’s what each one does.
1. Image downsampling (the biggest lever)
Most of the savings come from images. Downsampling reduces an image’s resolution — its DPI (dots per inch) — by discarding pixels the output doesn’t need. A photo scanned at 600 DPI but only ever viewed on screen carries roughly four times the pixels a monitor can show; downsampling it to 150 DPI throws away that excess and can cut the image’s size dramatically. This is lossy: the discarded pixels don’t come back.
2. Image re-encoding
Beyond resolution, images can be re-encoded with a more efficient codec and quality setting — for example, storing photographic content as JPEG with a chosen quality level. Lowering the quality discards fine detail the eye is unlikely to notice, trading a little fidelity for a smaller file. This is also lossy, and it’s where “aggressive” compression settings do their work.
3. Object-stream re-encoding (lossless)
PDFs store their structure as streams of objects. A compressor can repack and re-compress these streams more tightly — and bundle many small objects into compact object streams — without changing a single visible thing. This is lossless: the data is reorganized, not discarded.
4. Font subsetting
If a document embeds a full font but uses only 40 characters of it, the rest is dead weight. Subsetting keeps only the glyphs actually used, shrinking the font data. Done correctly, this is lossless — the text looks identical.
5. Stripping bloat
Finally, the compressor removes what isn’t needed: stale metadata, page thumbnails, unused objects, duplicate images, and other overhead. Each item is small, but together they add up, and removing them is lossless.
Lossy vs lossless compression
These two ideas explain almost every compression trade-off:
| Lossless | Lossy | |
|---|---|---|
| What it does | Reorganizes data without discarding any | Permanently discards detail to save more |
| Applies mainly to | Streams, fonts, metadata, structure | Images (downsampling + re-encoding) |
| Visible change | None | Possible blur/artifacts if aggressive |
| Reversible | Yes (data is identical) | No (detail is gone for good) |
| Typical savings | Modest | Large, on image-heavy files |
Most real compressors combine both: lossless techniques are applied safely everywhere, and lossy image compression is dialed up or down depending on how small you need the file versus how much quality you want to keep. A “high quality” setting leans lossless and lightly lossy; a “maximum compression” setting downsamples and re-encodes images hard.
Why text-only PDFs barely shrink
A common surprise: you compress a text-heavy PDF and it hardly changes size. That’s expected. Text in a PDF is stored as compact font references and character codes that are already small and already compressed. There are no big images to downsample, so the only savings come from lossless cleanup — re-encoding streams, subsetting fonts, stripping metadata — which on an already-efficient file might be just a few percent.
In other words, compression shrinks images, not words. If a text PDF won’t shrink, it usually means the file is already lean. To get meaningful reductions you generally need image content and a willingness to downsample it. (If your goal is actually to extract the pictures, PDF to JPG pulls each page out as an image rather than shrinking the PDF.)
A practical DPI guide
DPI is the dial that controls image size, and the rule is simple: use the lowest DPI that still looks right for the document’s actual purpose. Detail beyond what the output device can show is wasted data.
- 72–96 DPI — screen-only viewing where size matters most (web, quick previews). Smallest files; not for printing.
- 150 DPI — the sensible middle ground for emailing and general documents that might occasionally be printed. Crisp on screen, much smaller file.
- 300 DPI — the standard for quality printing. Use this when the document must print sharply or fine image detail matters.
- Above 300 DPI — rarely useful for normal documents; it mostly just inflates the file.
So for an email attachment, downsample to ~150 DPI; for a document headed to a professional printer, keep images at 300 DPI.
How to compress a PDF — privately
PDFs you compress are often sensitive: contracts, financial statements, scanned IDs. Many online compressors upload your file to their servers — exactly what you don’t want for confidential documents.
imisspdf’s Compress PDF tool runs in your browser: it downsamples images, re-encodes streams, subsets fonts, and strips bloat locally, so the file never leaves your device. The workflow:
- Open the Compress PDF tool and select your PDF.
- Choose a compression level (or target DPI) based on how the file will be used — lower DPI for email, 300 DPI to keep print quality.
- Compress and download. The smaller file is built locally; nothing is uploaded.
If you’re assembling several documents first, Merge PDF combines them in your browser before you compress the result.
Common misconceptions
- “Compression always halves the file.” Only if there’s a lot to compress — image-heavy files shrink a lot; lean text PDFs barely move.
- “Compressing twice makes it half as big again.” No — an already-compressed file has little left to remove; the second pass yields almost nothing.
- “Compression ruins quality.” Only aggressive lossy settings do; lossless cleanup and moderate downsampling are usually invisible.
- “Higher DPI is always better.” Only up to what the output shows — DPI beyond the screen or printer’s need is wasted bytes.
Related guides
- Best Free PDF Compressor 2026 (Tested)
- 10 PDF Compression Tools Benchmarked (2026)
- Shrink a file with the Compress PDF tool — free, in your browser.
PDF compression isn’t magic — it’s mostly the art of taking images down to the resolution you actually need and tidying everything else losslessly. Match the DPI to the document’s purpose, run Compress PDF in your browser, and you’ll get the smallest file that still looks the way it should.
Use Compress PDF: Reduce file size while optimizing for maximal quality. No signup, nothing uploaded.
Frequently asked questions
PDF compression reduces a file's size by shrinking the parts that take up the most space, while keeping the document readable. In most real-world PDFs the biggest consumers are embedded images, so the largest savings come from downsampling those images to a lower resolution (DPI) and re-encoding them with more efficient compression, such as JPEG for photos. On top of that, a compressor re-encodes the file's internal data streams more tightly, removes redundant or unused objects, subsets embedded fonts so only the characters actually used are stored, and strips bloat like metadata, thumbnails, and duplicate resources. Some of these steps are lossless (they reorganize data without discarding any), and some are lossy (they discard visual detail to save space, mainly in images). The right balance depends on the document: a scan-heavy file can shrink dramatically with image downsampling, while a text-only PDF that has little to compress will barely change in size.
Lossless compression makes a file smaller without throwing away any information — when decompressed, the data is bit-for-bit identical to the original. In PDFs this includes re-encoding internal streams more efficiently, removing unused objects, subsetting fonts, and stripping metadata; the document looks exactly the same afterward. Lossy compression saves more space by permanently discarding detail the eye is unlikely to miss, almost always in images: it downsamples them to a lower resolution and re-encodes them with a quality setting that drops fine information. The trade-off is that aggressive lossy settings can introduce visible blur or blocky artifacts, and the discarded detail cannot be recovered. Most practical PDF compressors combine both: lossless techniques are applied safely everywhere, and lossy image compression is dialed up or down depending on how small you need the file versus how much image quality you want to preserve.
Almost always because there was little to compress. The dramatic reductions you see advertised come from shrinking large embedded images, so a file made mostly of those — scans, photo-heavy brochures — can shrink a lot. A text-based PDF is different: its text is stored as compact font and character data that is already small and already compressed, so there's simply not much fat to trim, and you might only save a few percent. Other reasons include a file that was already compressed (running it again yields almost nothing), images that are already at a low resolution so downsampling can't help, or a conservative quality setting that prioritizes preserving detail over shrinking. If a text-heavy PDF won't shrink, that's expected and usually a good sign — it means the file is already efficient. To get meaningful savings you generally need image content and a willingness to downsample it.
Match the DPI to how the document will be used. For files meant to be viewed on screen or emailed, around 96 to 150 DPI is plenty — text stays crisp and images look fine on a monitor while the file gets much smaller. For general-purpose documents that might occasionally be printed, 150 DPI is a sensible middle ground. For documents that must print at full quality — anything going to a professional printer or where fine image detail matters — keep images at 300 DPI, the standard for quality print. Going above 300 DPI rarely helps for normal documents and just inflates the file. The key idea is that DPI beyond what the output device can show is wasted data: a 600 DPI image emailed for on-screen reading carries four times the pixels a screen needs. So downsample to the lowest DPI that still looks right for the document's actual purpose, and you'll get the best size-to-quality balance.
It depends on whether the tool uploads your file. PDFs you compress are often sensitive — contracts, financial statements, scanned IDs, reports — and many online compressors upload your document to their servers, shrink it there, and send it back, holding a copy for a retention period. The safer approach is a compressor that runs in your browser, so the file never leaves your device. imisspdf's Compress PDF tool processes locally: it downsamples images, re-encodes streams, and strips bloat inside your browser tab, with no upload, no account, and no watermark. For confidential documents, prefer in-browser or fully offline compression over any upload-based service. You can verify the claim yourself by opening your browser's Network tab and confirming that no file upload request is made while the compression runs — if nothing uploads, the file stayed on your machine.
Related articles
Best Free PDF Compressor 2026 (Tested)
We tested 10 free PDF compressors in 2026 on file size, quality, privacy, and limits. See the rankings, the comparison table, and which one wins for you.
Best Online PDF Tools 2026
We compared 10 online PDF tool suites in 2026 on breadth, privacy, and free limits. See the rankings, the comparison table, and which free PDF toolkit fits you.
Best PDF Annotator 2026 (Tested & Ranked)
We tested 9 PDF annotators in 2026 on privacy, free limits, and markup tools. See the rankings, the comparison table, and which annotator actually fits you.