hey guys i have done few sample projects (just for fun) using silverlight deep zoom. its really cool and exciting stuff. but i was curious to know about what goes on behind the scenes to achieve this sort of thing. the deep zoom composer generates many images and few xml files within sub folders of its data source.
can anyone explain how exactly all this stuff is working. what concepts are being used to achieve this
thnx in advance
As I understand it, it is similar to the techniques they use in games to keep the frame rate up; while something is small, only fetch a low-res version of it. As they zoom in, fetch increasing levels of detail.
Run wireshark and you'll see the traffic happening as you pan/zoom.
The tooling's job is to generate data that lets the client get images at different resolutions without too much pain.
Have a look at the Open-Zoom project. It is an open source project, so you'll be able to answer your question yourself by looking at the implementation of Open-Zoom:
Here's a screencast showing how a deep zoom 'tile pyramid' looks (watch in flash, or you can interact if you have SL3 beta).
There is also a DeepZoom Viewer that lets you 'explore' a Deep Zoom image "kinda" like Windows Explorer - seeing each layer and what images it consists of.
These might help to visualise how DeepZoom works - but the accepted Open Zoom answer no doubt gives you the greatest level of detail/technical understanding.