Google has named the new encoder Guetzli, the Swiss German term for cookie. It aims to improve on existing JPEG encoders by introducing a highly optimised multi-stage compression process that can retain more detail than current algorithms. As compression increases, the file size reduces but more artifacts enter the image.
Images created by Guetzli are on average 35% smaller than those from the current industry-leading encoders. This leads to lower photo file sizes and shorter download times over the Internet. Alternatively, it’s now possible to significantly improve the image quality of a file without raising its size.
Google said the breakthrough is the result of an algorithm that balances quality reductions and file size by searching for differences between JPEG and Guetzli representations of image data. The technical details are available in Google’s blog post. In principle Google is able to make estimations of colour perception and visual masking in a more detailed way than current techniques allow.
A comparison test of images created by Guetzli and the popular libjpeg JPEG generation library found human testers prefer the Guetzli images. For the experiment, Google kept the file sizes constant, effectively giving Guetzli a 35% lead on quality. This was carried into the results as the users preferred the higher clarity of the new algorithm.
Higher-quality images packed into smaller file sizes will be particularly useful on the web. As the Internet expands into emerging countries, many people use slow and unreliable connections that struggle to download the high-resolution imagery used on most sites. Like Google’s other web-centric compression algorithms, Guetzli doesn’t change the base JPEG format so all current devices and browsers can view the generated images.
“At Google, we care about giving users the best possible online experience, both through our own services and products and by contributing new tools and industry standards for use by the online community,” said Google. “That’s why we’re excited to announce Guetzli, a new open source algorithm that creates high quality JPEG images with file sizes 35% smaller than currently available methods, enabling webmasters to create webpages that can load faster and use even less data.”
As will all encoding techniques, Guetzli does have a trade-off. Because it is more complex than existing encoders, it takes “significantly” longer to complete compression. While the results are worth it in the end, this could have repercussions for camera makers. The time between shots might be reduced if Guetzli compression is used. Cameras could queue up data for encoding but this may cause bottlenecks and still compromise performance.
Nevertheless, Guetzli is a significant step forward in the development of image encoding algorithms that looks set to benefit a wide audience of photo creators and consumers. Developers looking to try it out in their next project can find the code on GitHub.
