Journey to JPEG XL: open-source experiments shaped the future of image coding
5 hours ago
- #image compression
- #digital imaging
- #JPEG XL
- Images are essential for the internet, balancing between visual fidelity and bandwidth.
- JPEG XL developed over a decade through experiments in psychovisual modeling and compression techniques.
- Early projects included WebP Lossless, Brotli, Butteraugli, Guetzli, and Brunsli, which improved upon JPEG standards.
- PIK (2017-2019) merged psychovisual optimizations and efficient compression, forming the proposal for JPEG XL.
- JPEG XL combines PIK and FUIF features, using distribution selection and context trees for efficient decoding.
- JPEG XL is now adopted in photography (DNG, ProRAW), medical imaging (DICOM), and publishing (PDF, EPUB).
- Support exists across operating systems (Apple, Linux, Windows) and browsers (Safari, Firefox, Chrome).
- Commercial encoder IP cores are available for real-time, low-power image capture in ASIC/FPGA designs.
- JPEG XL aims to provide a long-lasting foundation for digital imaging, built through iterative research and community feedback.