PDF Guides

How PDF Compression Works: Quality Setting, Savings, and Image-Heavy vs Text-Heavy PDF Behaviour

17 July 2026 Toolmatico Editör Ekibi 4 views
How PDF Compression Works: Quality Setting, Savings, and Image-Heavy vs Text-Heavy PDF Behaviour

PDF compression is not a single method — behind the scenes, Ghostscript, iText7 and a raster fallback race against each other and the smallest result wins. Image-heavy PDFs shrink 50-80%; text-heavy PDFs only save 5-25%. This guide covers what the 10-95 quality setting really means, how to pick a value, what to do with encrypted PDFs, and why repeated compression hurts your file.

PDF compression is not a single magic button. Give the same 10 MB file to two different tools and one may hand back 3 MB while the other gives 7 MB. That is not luck; it is about which compression method best matches what your PDF actually contains. And the gap between "I compressed it, saved 5%" and "I compressed it, saved 75%" is not a difference of tool quality — it is a difference of file content. In this guide we explain how PDF compression actually works, when to pick which quality value, why image-heavy and text-heavy PDFs behave so differently, and what to do with encrypted PDFs, all grounded in real tool behaviour.

What PDF Compression Actually Means

A PDF holds three main types of data: text (fonts and letter positions), images (photos, scanned pages, screenshots) and object streams (bookmarks, metadata, form fields). Compression affects each of these differently. Text is already very efficiently encoded — a page of plain text usually only takes a few kilobytes, and extra compression on top yields little. Images, on the other hand, account for the bulk of file size; a photo-bearing page can easily be 500 KB, a scanned page 1-2 MB. That is why compression savings mostly depend on how much of your PDF is imagery.

Compression uses two fundamental approaches: lossless repacking and lossy image re-encoding. Lossless repacking makes the PDF's internal structure more efficient — merges duplicate objects, discards unused font resources, zips streams tighter. Lossy re-encoding lowers image quality slightly to save far more space. Modern compression tools combine both: first images are downsampled to a target quality, then the whole file is repacked.

Toolmatico Approach: A Race Between Three Methods

The Toolmatico PDF Compress tool is not tied to a single compression engine. When your file arrives at the server, three different methods run in parallel: Ghostscript, iText7 repack and a PDFtoImage raster fallback. All three outputs are measured and the smallest file is automatically returned. The reason is simple: different content types respond better to different methods. Ghostscript usually wins on mixed content (text + images). iText7 repack gives small but reliable gains on pure text PDFs. PDFtoImage is most aggressive on image-heavy, especially scanned PDFs.

This three-way race removes the "which method is right for my file?" decision from the user. Instead of manual trial-and-error, the system tries for you. The trade-off: processing takes 5-15 seconds (slightly slower than single-engine tools) and uses significant server CPU. Because of this, a 5 requests per minute per IP limit applies; if you compress many files back-to-back, keep this in mind.

Quality Value: What Do 10, 75 and 95 Really Mean?

The tool offers a slider from 10 to 95; the default is 75. This value maps to the JPEG quality applied to images — text is not affected. A practical guide:

  • 10-30: Extreme compression. Images break into pixel blocks; scanned text becomes blurred at the edges. Only makes sense when file size is critical and the document will not be viewed on a large screen.
  • 40-55: Visible quality loss but very small file. Fits small-screen previews; do not use for A4 print.
  • 60-75: The balanced choice for web sharing and email attachments. Visual quality drops slightly but readability is preserved. Correct zone for most users.
  • 80-90: Visual loss is barely noticeable; savings are moderate. Recommended for presentations, formal documents or print-bound files.
  • 95: Near-lossless. Savings come mostly from the repack step; image-heavy PDFs will not shrink much at this level.

A practical tip: if you cannot decide, start at 75. Open the output and compare against the original. If you cannot see a difference, drop to 60 and try again. At some point your eye catches a difference — step back to the previous value. This "binary-search" approach lets you find the right quality for your file in 2-3 attempts.

How Much Should You Expect to Save? It Depends on Content

The most common surprise is that the same setting gives wildly different savings on two different PDFs. The reason is simple: how much of the PDF is imagery. A rough guide:

  • Scanned documents / photocopy PDFs: 60-80% savings. Every page is essentially a large image; lowering quality shrinks it dramatically.
  • Presentations, portfolio PDFs with photos: 40-60% savings. Images dominate the file.
  • Mixed content (text + moderate images): 20-40% savings. Typical for a normal report.
  • Text-heavy documents (Word-to-PDF, academic papers): 5-25% savings. Text is already compressed; only small repack gains remain.
  • Already-compressed PDFs: 0-5% savings. If you already compressed the file once, a second round barely helps.

These numbers are not about tool quality — they are about your file's character. If someone tells you their 5 MB loan-application PDF cannot possibly need to be that big and it is pure text, the answer is probably "it is already close to optimum". Check whether your PDF is really text or scanned pages: can you select the text with your cursor when you open the file? If yes, it is text-based; if you can only draw a rectangle around it as an image, it is scanned and you can expect large savings.

Encrypted PDFs and Compression

You cannot compress an encrypted PDF directly. Compression needs to read the PDF's internal structure; encryption locks that structure. This is not a dead end — a three-step workflow solves it:

  1. Use our PDF Unlock tool first (you must know the password — this is not password cracking, just removing a known password).
  2. Feed the unlocked PDF to the compressor.
  3. If needed, re-encrypt the output with our PDF Protect tool — same password or a new one.

These three steps take 30-60 seconds end-to-end. Since compression runs on the unlocked file, the savings are full; trying to compress while encrypted will fail or yield poor gains.

Compression Limits and Correct Usage

Knowing the concrete limits helps:

  • Maximum file size: 50 MB. For larger files, consider splitting with our PDF Split tool first.
  • Rate limit: 5 requests per minute per IP. For bulk work, wait about 60 seconds every 5 files.
  • Retention: Your file stays on our server for 60 minutes, then is auto-deleted. Use your download link within that window; do not keep it as long-term storage.
  • No external APIs: Your file is never sent to a third-party service; everything runs on Toolmatico's own servers.

Common Mistakes We Keep Seeing

Let us walk through the same mistakes we see repeatedly.

1. Compressing the same file over and over. Compression is lossy. Compressing an already-compressed file yields almost no size gain but continues to degrade image quality. Keep the original safe; regenerate the compressed copy on demand.

2. Picking the wrong quality value. "Smaller is always better" is an illusion. If you compress a slide deck at quality 30 for a boardroom projector, the images may be unreadable. Think about use case: email attachment 60-75, archive 85-95.

3. Expecting miracles from text-heavy PDFs. A 2 MB text-only PDF from Word will come back as 1.8 MB. That is normal and correct; text was already compressed. Big savings need imagery.

4. Using the compressed copy for cloud backup. Keep original quality in cloud backup. Generate the compressed copy for email, but store the original — do not regret three years later that you kept a lossy version.

5. Ignoring the rate limit. Trying to upload 20 files back-to-back will fail on the 6th. For high-volume workflows, split into groups of 5 and wait a minute between groups — it runs more stably.

Summary and Toolmatico Tools

PDF compression works on any file, but your savings depend on your content. Image-heavy PDFs shrink 50-80%; text-heavy PDFs shrink 5-25%. Toolmatico's three-engine race decides which method fits your specific file, so you do not have to. The 60-75 quality range is balanced for most cases; 85-95 for print and archive, 40-60 when you really need small files. For encrypted PDFs, unlock first, compress, re-encrypt if needed.

The Toolmatico PDF Compress tool covers all scenarios in this guide; for encrypted files, PDF Unlock and PDF Protect complete the workflow. All tools run on Toolmatico's own servers — no external API calls, files auto-delete after 60 minutes.