Fits Realtime Explorer of Low Latency in Every Dimension


This is my second significant (for me, anyway) piece of scientific coding - the first was an automatic HI source finder, GLADOS (Galaxy Line Analysis for Detection Of Sources). FRELLED is a set of Python scripts for Blender to let it view 3D FITS files in the realtime display. I've already described my Blender FITS-viewing efforts - the difference with this script is that it's designed to be robust and user-friendly (or at least non-hostile). It has a GUI and everything !

I've seen a few attempts at realtime medical imaging in Blender, but most efforts seem to have concentrated on rendering rather than the realtime display. Which is fine, but such viewers already exist. I wanted to be able to freely examine my data sets from any angle, mostly because it's cool. I was also vaguely hoping that it might reveal new faint extended features that would be difficult to spot in two dimensions.


While it turned out to be indeed very cool, I didn't find any new features in the data. But I did find that it makes visual source extraction twenty-seven times easier. Instead of having to carefully record the co-ordinates of detected objects for further analysis (this is fine the first time, but it wears very thin the three hundredth time), you can now just point-and-click and FRELLED automatically generates catalogues. It can also search NED and the SDSS.


I don't know if this is of wider interest or not, but if anyone has any requests then I'm all ears. To lean more about FRELLED before trying it out, a user guide is available here. FRELLED itself can be downloaded here.

I've used FRELLED to participate in the National Science Foundation's annual International Science and Engineering Visualisation Challenge (yes, I think they could have thought of a shorter title too). This video shows the Universe as seen in neutral hydrogen, starting from our own galaxy and finishing more than a bilion light years away.


rhysy.net | feedback@rhysy.net