How PDF Compression Algorithms Target Metadata and Graphics
PDF bloat is primarily caused by uncompressed high-resolution images, duplicate embedded font packages, and rich metadata streams. Our compression algorithm traverses the PDF catalog, identifies raster image objects, downsamples their resolutions (e.g. from 300 to 150 DPI), and recompiles them with optimized JPEG coefficients. It also removes unused metadata objects, strips duplicate font subsets, and compresses text streams using standard Flate compression—all executed in WebAssembly memory without sending your files to a server.