PDF Compression Algorithms

Deflating the Data

PDFs support multiple internal compression algorithms for different data types. Text streams and vectors are typically compressed using FlateDecode, which is essentially zlib/DEFLATE—the same algorithm used in ZIP files and PNG images.

Images within a PDF can use DCTDecode (JPEG compression) or JPXDecode (JPEG2000). A PDF is effectively an uncompressed structural shell containing highly compressed inner streams.

Common Mistakes

• Trying to ZIP a PDF to make it smaller. Since the internal streams are already compressed (FlateDecode), zipping a PDF usually yields less than a 5% reduction.

Next Steps

Put this theory into practice. Use our client-side tools to inspect documents directly in your browser.

Open Developer Tools