Lately I've gotten interested in playing with some of the raw (or at least, more raw than the standard press image) data from the Hubble and other space telescopes... There is a massive archive of all the data Hubble has collected at http://archive.stsci.edu/hst and an introduction to the FITS image format they use at http://www.spacetelescope.org/projects/fits_liberator/
However, I haven't made any impressive images yet because I'm still trying to wrap my head around the entire data pipeline & come up with a workflow that helps automate the process. There is a ton of preprocessing that goes into creating high-quality colored images from the raw data - you have to figure out white balance, clean up hot pixels, noise and cosmic ray hits among other things, align the images, decide which frequencies will represent which colors, align them correcting for geometric distortion, and finally combine them algorithmically.
There is an introduction to some of this work at http://www.stsci.edu/ftp/science/hdf/pipeline/pipeline.html and http://www.spacetelescope.org/projects/fits_liberator/improc... but I've had a hard time figuring out the best way to turn this process into a workflow - I've basically been manipulating them manually so far. Has anyone else experimented with this? I'm starting to consider putting together a collection of scripts to automate some of the process, but it's a difficult thing to do entirely automatically!
I've watched this video several times and enjoyed it. But I'm skeptical about the claim that it was "questioned" or that they really thought the image might have come out completely black. Did any scientist think that between us and the edge of the observable universe there was absolutely nothing?
Another minor nitpick: between this video and a similar one by the same person, the image is said to have been captured in 1995, 1996, or 2003. Which is it? http://www.youtube.com/watch?v=BG4IEePixoU
However, I haven't made any impressive images yet because I'm still trying to wrap my head around the entire data pipeline & come up with a workflow that helps automate the process. There is a ton of preprocessing that goes into creating high-quality colored images from the raw data - you have to figure out white balance, clean up hot pixels, noise and cosmic ray hits among other things, align the images, decide which frequencies will represent which colors, align them correcting for geometric distortion, and finally combine them algorithmically.
There is an introduction to some of this work at http://www.stsci.edu/ftp/science/hdf/pipeline/pipeline.html and http://www.spacetelescope.org/projects/fits_liberator/improc... but I've had a hard time figuring out the best way to turn this process into a workflow - I've basically been manipulating them manually so far. Has anyone else experimented with this? I'm starting to consider putting together a collection of scripts to automate some of the process, but it's a difficult thing to do entirely automatically!