Compress PDF

The Science Behind Data Compression

Data compression divides into two fundamental categories: lossless and lossy. Lossless compression reduces file size without discarding any information, guaranteeing that the original data can be perfectly reconstructed. Algorithms like Deflate, used in ZIP files and within PDF content streams, work by identifying repeated byte patterns and replacing them with shorter reference codes. The Deflate algorithm combines LZ77 (finding matching sequences in a sliding window of previously seen data) with Huffman coding (assigning shorter bit patterns to more frequent symbols). For text-heavy PDF content streams, Deflate typically achieves 60-80% compression ratios. Lossy compression, used primarily for embedded images within PDFs, achieves far greater size reduction by permanently discarding information that human perception is unlikely to miss. JPEG compression, the dominant lossy method in PDFs, applies the Discrete Cosine Transform (DCT) to convert 8x8 pixel blocks from spatial data into frequency coefficients. The quantization step then rounds these coefficients, discarding high-frequency detail that contributes to sharpness but that the human visual system is less sensitive to. The quality parameter controls how aggressively quantization rounds these values: higher quality preserves more detail at larger file sizes. PDF compression extends beyond raw image handling to include structural optimizations. Font subsetting removes unused glyphs from embedded fonts, potentially reducing a 500 KB font file to just the 50-100 characters actually referenced in the document. Object stream compression groups multiple small PDF objects into a single compressed stream, reducing the overhead of individual object headers. Duplicate resource elimination identifies identical images or fonts embedded multiple times and replaces redundant copies with shared references. The practical impact of compression on everyday workflows is substantial. Email providers typically limit attachments to 10-25 MB, and many web upload forms impose similar restrictions. A 45 MB presentation PDF compressed to 8 MB crosses below these thresholds, eliminating the need for file-sharing services. For organizations managing thousands of documents, even moderate compression across an archive translates to significant storage cost savings. The balance between quality and file size depends on the intended use: documents destined for professional printing should use minimal compression to preserve image fidelity, while those shared digitally for on-screen reading can tolerate more aggressive compression without noticeable quality loss.

Methodology

This tool applies several complementary techniques to minimize PDF file size. The primary method is image resampling: each embedded image is rendered onto a canvas element and re-encoded as JPEG at a configurable quality level. Light compression uses a quality factor of 0.8, preserving fine detail while reducing size by roughly 20-30%. Medium compression at 0.6 strikes a practical balance, achieving 40-60% reduction with minimal perceptible quality loss. High and maximum settings (0.4 and below) push compression further, often reaching 60-70%+ reduction, though photographic detail may show JPEG artifacts. Beyond image resampling, the compression pipeline benefits from stream-level optimizations within the PDF structure. Font data, metadata, and content streams are re-encoded using efficient binary representations. Redundant or duplicate objects are consolidated. Unnecessary metadata such as creation software tags and editing history can be stripped, contributing additional size savings. The quality presets are designed around common real-world scenarios. Low compression is ideal for documents heading to print or professional review where every detail matters. Medium suits the vast majority of everyday sharing, from email attachments to cloud uploads. Maximum compression targets situations where file size is the overriding concern, such as meeting strict upload limits or archiving large document libraries for long-term storage.

Understanding Your Results

Understanding compression ratios helps you choose the right settings. A compression ratio of 50% means the output file is half the original size—the higher the ratio, the more space you save. Typical results vary widely: scanned documents often achieve 50-70% compression, while photo-heavy presentations may reach 40-60%. Text-dominant PDFs rarely exceed 10-15% because text data is inherently compact. The quality-versus-size trade-off is the central decision in PDF compression. At low compression, images retain nearly all their original detail, making differences imperceptible even when zoomed to 200%. Medium compression introduces subtle changes that are invisible at normal viewing distances but may be noticeable in side-by-side comparisons of high-resolution photographs. High and maximum compression can produce visible artifacts—blocky patterns in gradients, softened edges, and color banding—particularly in images with smooth tonal transitions. Compression rarely affects text readability. Because text in PDFs is stored as vector outlines (font glyphs) rather than rasterized pixels, it remains perfectly sharp regardless of compression level. The same applies to charts, logos, and diagrams created as vector graphics. Only embedded raster images—photographs, scanned pages, screenshots—are affected by the compression quality setting. If your document is primarily text with occasional small images, even maximum compression will produce a highly readable result.

Practical Examples

A marketing team needs to email a 45 MB product catalog to distributors. Medium compression reduces it to 12 MB while keeping product photos sharp. They switch to maximum compression and achieve 8 MB—small enough for any email provider. A law firm archives thousands of scanned contract PDFs. Using high compression, they reduce their 500 GB document archive to under 200 GB, saving significant cloud storage costs while maintaining perfectly readable text. A university student compresses a 30 MB research paper with embedded figures before uploading to a submission portal with a 15 MB limit. Low compression preserves graph clarity while meeting the requirement.

Tips & Best Practices

Always start with medium compression when you are unsure—it offers the best balance between quality and file size for most documents. If the result is still too large, increase to high or maximum compression. For documents heading to print, stick with low compression to preserve image sharpness. For email attachments, maximum compression often gets you under the 10 MB limit most providers enforce. Check the before/after comparison to verify image quality meets your needs before downloading. Keep your original PDF as a backup. Compression is a one-way process—you cannot restore the original quality from a compressed file. When compressing scanned documents, medium compression typically yields excellent results since scanned images contain noise that compresses efficiently.

Sources

• PDF-lib – JavaScript PDF Library
• JPEG Compression Explained