Search engine giant Google released their open source algorithm Guetzli to efficiently compress high-quality JPEG images uptil 35% more than other available options. Guetzli will also ensure that size is compressed with zero compromise over the image quality. This announcement will prove very beneficial for those who are into developing an app or website where storing, manipulating or displaying huge amount of images is frequently required. This will eventually allow web pages to not only load faster but also use less data.
It reduces hard to compress disordered data and delivers easy to compress ordered data. Google claims it’s encoder will target quantization phase of the compression process to reduce JPEGs size without hurting the quality of image. In quantization phase more visual quality loss takes place hence, a smaller size of reduced file. It will not create a new image format of compressed image. It will work seamlessly with all available browsers as well as image processing tools. This step will help maximize the monetization of google’s Ads which is a primary source of Google income.
img credit: Google Research Blog
In above images the left one(original and uncompressed) has been compressed using Guetzli hence very less artifacts on the right image (compressed with Guetzli) can be noticed. Middle image is libjpeg in the middle.
On the downside, it is slower than its alternatives. It basically strike a balance between minimal loss and file size by using a search algorithm which will overshadow difference between Guetzli’s psychovisual model and JPEG’s psychovisual modeling. Large amount of images shrink translates to faster downloads, less stress on networking infrastructure and eventually highly satisfied users.
You may be interested in following: