Can somebody explain how this works?
Every level is split into quadrants.
So when someone zooms in, it will fetch the quadrants that is visible to the browser viewport. By doing this way, the browser doesnt have to load the rest of the image and thus save alot of bandwidth!
https://max1.prodibicdn.com/9ydmr8zdmerork/00/10/0_0.jpg
https://max1.prodibicdn.com/9ydmr8zdmerork/00/11/0_0.jpg
https://max1.prodibicdn.com/9ydmr8zdmerork/00/12/0_0.jpg
https://max1.prodibicdn.com/9ydmr8zdmerork/00/13/0_0.jpg
https://max1.prodibicdn.com/9ydmr8zdmerork/00/14/0_0.jpg
https://max1.prodibicdn.com/9ydmr8zdmerork/00/15/0_0.jpg
should give you an idea of what's going on.