Last updated
Compression Results by Preset
A 15MB scanned document compressed with different quality presets:
Original file: annual-report-scanned.pdf
Original size: 15.2 MB
Screen preset (72 DPI images):
Compressed size: 1.8 MB
Reduction: 88%
Best for: Email, web download, digital viewing
eBook preset (150 DPI images):
Compressed size: 3.4 MB
Reduction: 78%
Best for: Tablets, e-readers, digital distribution
Print preset (300 DPI images):
Compressed size: 6.1 MB
Reduction: 60%
Best for: Desktop printing, office use
Prepress preset (lossless):
Compressed size: 12.8 MB
Reduction: 16%
Best for: Professional printing, archival
File Size Breakdown Analysis
Understanding what's making a PDF large:
File: product-catalog.pdf (28.4 MB)
Size breakdown:
Images: 24.1 MB (85%) ← Main opportunity
Fonts: 2.8 MB (10%)
Text/vectors: 0.9 MB (3%)
Metadata: 0.6 MB (2%)
Optimization opportunities:
→ Resample 47 images from 600 DPI to 150 DPI: save ~20 MB
→ Subset 8 embedded fonts: save ~1.8 MB
→ Remove editing history metadata: save ~0.5 MB
→ Estimated compressed size: ~6.1 MB (78% reduction)
Image Compression Settings
Controlling image quality for different use cases:
JPEG quality settings for embedded images:
Quality 95 (near-lossless):
File size: ~8 MB
Visible quality loss: None
Use for: Professional printing
Quality 80 (high quality):
File size: ~4 MB
Visible quality loss: Minimal
Use for: Office printing, digital distribution
Quality 60 (medium quality):
File size: ~2 MB
Visible quality loss: Slight (visible on close inspection)
Use for: Email, web, screen viewing
Quality 40 (low quality):
File size: ~1 MB
Visible quality loss: Noticeable
Use for: Preview thumbnails, very size-constrained sharing
Font Optimization
Reducing file size by optimizing embedded fonts:
Before optimization:
Helvetica Neue (full embed): 892 KB
Times New Roman (full embed): 743 KB
Arial (full embed): 654 KB
Total fonts: 2.3 MB
After font subsetting:
Helvetica Neue (subset, 847 chars used): 124 KB
Times New Roman (subset, 1,203 chars used): 187 KB
Arial (subset, 412 chars used): 63 KB
Total fonts: 374 KB
Font savings: 1.9 MB (83% reduction)
Font subsetting keeps only the characters actually used in the document. A document using only English text doesn't need the full Unicode character set embedded.
Metadata Removal
Stripping unnecessary metadata from a heavily-edited document:
Metadata found in document:
Creation date: 2023-01-15
Modification date: 2024-03-17
Author: [name]
Software: Adobe InDesign 18.5
Edit history: 47 revisions (2.1 MB of revision data)
Thumbnail preview: 340 KB
Color profiles: 3 embedded ICC profiles (180 KB)
After metadata removal:
Removed: Edit history (-2.1 MB)
Removed: Thumbnail (-340 KB)
Removed: Unused color profiles (-120 KB)
Kept: Author, dates (minimal size)
Metadata savings: 2.56 MB
Scanned Document Optimization
A scanned PDF is essentially a collection of images:
Scanned document: 50-page contract
Original: 45 MB (50 pages × ~900 KB per page at 600 DPI)
Optimization steps:
1. Resample from 600 DPI to 200 DPI: 45 MB → 5 MB
2. Apply JPEG compression (quality 85): 5 MB → 2.8 MB
3. Remove metadata: 2.8 MB → 2.6 MB
Final size: 2.6 MB (94% reduction)
Text is still clearly readable at 200 DPI for a contract document.
For OCR-processed PDFs (text layer added):
Original scanned images: 2.6 MB
Added text layer: +0.1 MB
Total with searchable text: 2.7 MB
Batch Compression Results
Processing a folder of monthly reports:
Batch compression — 12 monthly reports (Screen preset):
File Original Compressed Reduction
report-jan-2024.pdf 8.2 MB 1.1 MB 87%
report-feb-2024.pdf 7.9 MB 1.0 MB 87%
report-mar-2024.pdf 9.1 MB 1.2 MB 87%
report-apr-2024.pdf 8.5 MB 1.1 MB 87%
...
report-dec-2024.pdf 8.8 MB 1.2 MB 86%
Total original: 101.4 MB
Total compressed: 13.4 MB
Total saved: 88 MB (87% reduction)
When Compression Has Limited Effect
Some PDFs don't compress much:
File: technical-drawing.pdf (2.1 MB)
Content: Vector graphics, no raster images
Compression attempt:
Images: None to compress
Fonts: Already subsetted
Vectors: Lossless only (minimal savings)
Compressed size: 1.9 MB (10% reduction)
Explanation: Vector PDFs (from CAD software, Illustrator, etc.)
are already efficient. Compression helps most with:
- Scanned documents (image-heavy)
- PDFs exported from Word/PowerPoint with embedded photos
- PDFs with full font embeddings
Privacy — Browser-Based Processing
Your documents never leave your browser:
- PDF processing happens entirely in JavaScript using PDF.js and pdf-lib
- No file upload to any server — check the Network tab in DevTools to verify
- Safe for confidential documents: contracts, medical records, financial statements
- The compressed PDF is generated locally and downloaded directly
- No account required, no file size limits imposed by server quotas