Google's Guetzli Open Source JPEG Encoder Compresses Images by 35 Percent

Google Creates Guetzli, a New JPEG Encoder With 35 Percent Smaller Files Without Loss of Quality

Images pushed through the Guetzli encoder are still available in a regular JPEG format, making them compatible with just about every application and browser now on the market. A good encoder will produce a great looking compressed image while reducing file size as much as possible.

Smaller file sizes might seem an arcane technology concern, but they're crucial to fast-loading websites.

Google says its Guetzli compression produces JPEG images with fewer photo-degrading artifacts. Left: The uncompressed original. In addition to making image files that are smaller, Guetzli is also focused on creating images that look better than other compressed images, too.

Google has developed a new open-source JPEG compression algorithm, dubbed Guetzli (apparently that means "cookie" in Swiss German). Despite the increased time, Google's post assures that human raters preferred the images churned out by Guetzli.

"Butteraugli takes into account. properties of vision that most JPEG encoders do not make use of", Google researchers said in a Guetzli research paper (PDF).

The JPEG encoder is open-source and available for you to download and implement in your own projects from this GitHub repository. "The visual quality of JPEG images is directly correlated to its multi-stage compression process: colour space transform, discrete cosine transform, and quantization". Mozilla, maker of the Firefox web browser, began a project in 2014 called Mozjpeg created to improve on standard compression engines.

Furthermore, Google claims that Guetzli does this compression without compromising on image quality. Guetzli strikes a balance between minimal loss and file size by employing a search algorithm that tries to overcome the difference between the psychovisual modeling of JPEG's format, and Guetzli's psychovisual model. Presumably Google are considering optimisations and rational solutions to speed up Guetzli's compression speed performance.

The downside to this methodology is that compression takes significantly longer than now available methods. Guetzli (on the right) shows less ringing artefacts than libjpeg (middle) without requiring a larger file size.

However, it's worth noting that Guetzli is slower than other options out there.