It's as much fun as seeing white dog hairs on the black couch. This is to advance its efforts in getting webpages to load faster while using less data. Google describes the slower compression process as a "worthy tradeoff".
Google has announced a new open source JPEG encoder called Guetzli that creates images with file sizes 35 percent smaller than now available methods. The idea isn't to replace JPEG, but tweak its settings to minimize the likelihood we'll notice problems when files are squeezed.
In a blog post, Google said that Guetzli, which is "cookie" in Swiss German, is a JPEG encoder for web sites that it claims will create smaller overall files while still remaining compatible with current web sites, any image processing applications and the JPEG standard.
According to Google, the psychovisual modeling of Guetzli "approximates color perception and visual masking in a more thorough and detailed way than what is achievable" in existing methods. However, while Guetzli creates smaller image file sizes, the tradeoff is that these search algorithms take significantly longer to create compressed images than now available methods.
That automation is key. It would be interesting to see if Guetzli scores a wider acceptance.
The JPEG encoder is open-source and available for you to download and implement in your own projects from this GitHub repository.
Google isn't the only one trying this approach.
In plain English, as ArsTechnica writes, Guetzli, like all JPEG encoders, tries to "reduce a large amount of disordered data, which is hard to compress, into ordered data, which is very easy to compress". Google's tests showed that Guetzli outdoes Mozilla's tool by 29 to 45 percent. Presumably Google are considering optimisations and rational solutions to speed up Guetzli's compression speed performance. Uncompressed original on the left.
There's no free lunch here, though. However, Google said that images that use Guetzli are preferred visually over images that use libjpeg-based compression.
The only downside to Guetzli is that it's slower to encode JPEGs than with libjpeg. For example, a multi-megabyte image stored in BMP or PNG format can look nearly exactly the same converted to a JPEG that's only a few hundred kilobytes in size.