savas parastatidis

An updated view of the current state of the 3D Star Browser in Avalon

Chris Sells referred to my early screenshots (post 1, post 2) and commented on the way the stars were lined up 🙂 There is a good reason the stars look like that in those screenshots. I was using the personal version of the SDSS DR1 data (MySkyServer) that only include few tens of thousands of stars from a particular part of the sky. When, after the calculations of their distances, those stars are put in 3D, they appear as if they are all aligned in a single cone with its top representing earth (the position of the SLOAN telescope).

For my most recent tests (the queries are done by a background worker), I downloaded more stars from all around the sky. So, here's how it looks in 3D with ~7,000 stars only. Again, note that the centre of the 3D space here is meant to be the SLOAN telescope (earth) while the camera is positioned to circle at a few thousands parsecs away (of course you don't see the camera movement from this still 🙂

I am currently working on performance enhancements and the necessary logic to allow for lazy loading of the stars as the user moves the camera in 3D.

2 responses to “An updated view of the current state of the 3D Star Browser in Avalon”

  1. senkwe
    This is cool. What is the performance like so far?
  2. Savas Parastatidis
    Speed? Don’t ask. WinFX is only in pre-beta though and I don’t think Avalon was meant to be used as a tool for visualising millions of stars 🙂 Another important factor is that I am developing on a 2-year old IBM Thinkpad T40p. I am about to buy a new laptop, so i am expecting huge perfomance gains there. Finally, I am working on some application-level optimisations that will mean that only the number of starts necessary to give an idea of what’s going on will be added to the 3D model while a bitmap will be mapped in a sphere to represent the background. I am working on something WS-related at the moment so expect something in this area in a couple of weeks or so.