How to Compress a PDF Without Losing Quality (2026)
Last updated:
Quick Answer
To compress a PDF without visible quality loss: strip metadata and unused objects (saves 5–20%), subset embedded fonts (saves 10–30% on PDFs with many fonts), and optionally re-encode embedded images at quality 85 (visually lossless JPEG). For digital (non-scanned) PDFs, these three steps typically reduce size 30–60% with zero visible difference. FileCurve PDF Compressor's "Low" setting applies all three without downsampling images below 200dpi.
Step-by-step
- 1
Upload your PDF to FileCurve PDF Compressor.
- 2
Select "Low" compression — this preserves image quality while stripping overhead.
- 3
Low compression applies: metadata removal (author, software, creation date), unused object cleanup, embedded font subsetting, and image re-encoding at quality 85 without downsampling.
- 4
Review the output — text, diagrams, and photos should look identical at 100% zoom.
- 5
If you need further savings without visible loss, switch to "Medium" (images at 200dpi, quality 80) — still imperceptible at screen viewing size.
- 6
Download.
Expected output
Format
PDF, images at 200–300dpi (vs. original)
Quality setting
Low compression — visually identical to original at normal zoom
Estimated size
20–50% smaller than original for typical mixed PDFs
Why you might need this
- →Portfolio PDFs for architecture, design, or photography — must stay sharp
- →Print-ready documents where any image quality loss is unacceptable
- →Product documentation PDFs — diagrams and screenshots must stay crisp
- →Legal contracts with signed images — quality must be court-submissible
- →E-book PDFs where image quality affects reader experience
Troubleshooting
My PDF went from 10MB to 8MB — only 20% saving. I want more without quality loss
The remaining size is likely large vector graphics or 300+ dpi images. Switching to "Medium" (200dpi max) gives more savings with still-invisible difference on screen.
Why is my already-compressed PDF not getting smaller?
If the source was already compressed (e.g., exported from Word at High PDF quality), there are no metadata or overhead objects to strip. The file is already lean.
Font subsetting removed characters from my PDF
This should not happen — subsetting only removes unused characters from font tables, not characters in the document. If characters are missing, the source font had incomplete encoding. Re-export the PDF from source.
Scanned PDF at "Low" compression is still 25MB
Scanned PDFs are all image — there are no fonts, metadata, or vectors to strip. For scanned PDFs, any size reduction requires image compression, which means some quality loss is unavoidable.
Client needs the original and compressed version to look identical when printed
Use "Low" compression at 200dpi minimum. 200dpi prints identically to 300dpi at any size up to A3. Below 150dpi, large-format prints may show difference.
Frequently asked questions
Why do PDFs get large in the first place?
Four main causes: (1) high-resolution embedded images (biggest factor), (2) full embedded fonts instead of subsets, (3) metadata and XMP/EXIF data embedded in the file, (4) unused PDF objects from editing operations that were never cleaned up.
What is the difference between lossless and lossy PDF compression?
Lossless: metadata stripping, font subsetting, unused object removal — zero quality change. Lossy: image re-encoding at lower quality or lower DPI — some quality change, usually invisible.
Do scanned PDFs compress differently from digital PDFs?
Dramatically differently. Digital PDFs (from Word, InDesign) have vector text and graphics that compress losslessly. Scanned PDFs are pure images — any compression is lossy and degrades image quality.
What is font subsetting?
A PDF with custom fonts embeds the entire font file (often 200–500KB per font). Subsetting removes unused characters — if your document only uses 50 of 500 characters, the font data shrinks by 90%.
Does FileCurve remove hidden data from PDFs?
Yes — metadata stripping removes author name, software version, creation date, GPS (for embedded photos), and hidden comments. This is also a privacy benefit.