One thing that the readers of this blog might incorrectly be thinking is that I believe the Portable Network Graphics (.png) file format is the end-all, be-all related to image optimization! While it is true that .png is an EXCELLENT file format to use in a variety of situations, with some of the most diverse and absolute best software/scripts available in the marketplace for “de-fragging” imagery, it is not the optimal file format to use in all applications.
An example of this can be found on the Red Geographics web site. From that site: “Red Geographics delivers technical support, consultancy and training in the fields of GIS (Geographic Information Systems), geographical data, 3D visualisation and digital cartography.” Hans van der Maarel, the owner of Red Geographics, is a mapping friend of mine from The Netherlands. He does EXCELLENT work, and he and I share an entrepreneurial spirit. He and I have also spent a great deal of time talking about GraphicsOptimization, and are fortunate enough to see each other at least once per year at the annual gathering of the North American Cartographic Information Society.
Unoptimized 800×600 pixel .png image – 126.7 KB
Optimized 800×600 pixel .jpg image – 47.8 KB
Unoptimized 800×600 pixel .png image – 126.7 KB
Optimized 800×600 pixel .jpg image – 64.9 KB
Unoptimized 800×600 pixel .png image – 266.4 KB
Optimized 800×600 pixel .jpg image – 156 KB
For the purposes of this study, I decided to focus on 12 visual examples of the imagery that Hans produced for Oolaalaa, a company which took his map design and placed it on a spherical chair to have it appear like a globe (VERY cool idea). The existing .png images on that web page totaled 2.69 MB…HUGE from the standpoint of bandwidth required to correctly display an entire web page’s content! After a bit of optimization work in xat.com Image Optimizer, those same 12 images (three samples shown above) now total 1.52 MB. That is STILL a ton of bandwidth to require in order to display a web page, but it does reduce bandwidth (and download/display times) by over 43 percent…well-worth the effort.
I applied only six percent compression to the overall .jpg images, to make sure to avoid any blurry/chunky text. I then spent a majority of my optimization time working on custom zonal compression on an image-by-image basis. That effort resulted in roughly 20% of the overall reduction in image file sizes. The remainder of the savings came as the result of converting the file to .jpg format.
A bit “backwards” from the typical direction that we tend to head related to optimizing imagery! However, it is a perfect example as to why there isn’t one universal solution that individuals can apply to every graphic. Each graphic is unique, and different solutions will produce optimal results on an image-by-image basis.
