Author Topic: Hubble's Hidden Treasures Processing contest  (Read 3207 times)

Offline Josh Lake

  • PixInsight Old Hand
  • ****
  • Posts: 424
    • View Profile
Hubble's Hidden Treasures Processing contest
« on: 2012 March 28 12:09:45 »
I'm sure many of you already know about this, but the ESA announced this year's Hidden Treasures processing contest:

http://www.spacetelescope.org/projects/hiddentreasures/processingcontest/

It's pretty tricky to find data that hasn't been covered, but it would be great to have a PixInsight user win this contest! Top prize is an iPad (not to mention glory and bragging rights...).

Offline Josh Lake

  • PixInsight Old Hand
  • ****
  • Posts: 424
    • View Profile
Re: Hubble's Hidden Treasures Processing contest
« Reply #1 on: 2012 May 01 13:47:15 »
Not a lot of love for this one, I guess everyone just wants to play with their own data...  >:D

I tried my hand at a few images during NEAIC and am pleased with the results, though they're fairly interpretive and non-scientific.

I did NGC 1073 entirely in PixInsight, obviously pushing hard on the dust detail. I don't normally try the more extreme processing methods on my own data, but I was just kind of playing with this spectacular Hubble data. I processed this one entirely within PixInsight.

http://www.flickr.com/photos/79510497@N07/7116750741/sizes/l/in/photostream/

After the fact, I found that the Hubble team had already processed this one, so it's probably not even eligible for the contest.


The second and third come from a small section of NGC 1763, a rich nebula region. I wasn't thrilled with the first, but the second has some interesting color dynamics. I used PixelMath to create a custom look to really define the dark dust and the warm star activity region.

http://www.flickr.com/photos/79510497@N07/7121819611/sizes/l/in/photostream/

I'll be doing more of these and will probably screencast one of my processing runs to share with Reddit. If it turns out well, I'll share it here, too.

The contest ends in a month! I recommend that everyone who has the time to give it a shot, I've learned a lot just by working with this data.
« Last Edit: 2012 May 01 17:46:56 by jlake »

Online pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4103
    • View Profile
Re: Hubble's Hidden Treasures Processing contest
« Reply #2 on: 2012 May 01 17:11:02 »
i missed your post the first time somehow. i'm not sure how i noticed this contest (i think someone on flickr was posting a bunch of stuff and i followed links back to the contest). i had previously done their insane M51 mosaic and also NGC1376 but both had been previously done by the hubble team. i think my 1376 is better than theirs :)

at any rate it's a huge hassle sifting thru all their data and since i'm not a professional astronomer i get lost pretty easily in all of their different filters...

Offline Josh Lake

  • PixInsight Old Hand
  • ****
  • Posts: 424
    • View Profile
Re: Hubble's Hidden Treasures Processing contest
« Reply #3 on: 2012 May 01 17:51:45 »
Yeah, working with the Hubble Legacy Archive (http://hla.stsci.edu/hlaview.html) daunted me completely at first and I gave up for a while. I only tried again recently and found that a few searching tips helped immensely, cutting down on the not-so-good data and offering up much more potential.

I always run Advanced Search, choose the ACS camera (and maybe WFPC2), and only choose Level 2 products (already combined). Then I use the Images and Footprints tabs to see what's available before downloading any data. Within a half hour, I had found 8 good bi-color pairs to download and process.

The whole experience is a bit odd for me, dealing with this intense, oddly angled, and surprising data -- but it's been a fun challenge and I'm glad I'm finally getting results of a sort. Compared to a lot of the stuff people have been uploading to that Flickr group, I'd bet our PI users could do much better.

Online pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4103
    • View Profile
Re: Hubble's Hidden Treasures Processing contest
« Reply #4 on: 2012 May 01 19:11:25 »
i guess the HST is of course primarily for science... the ACS seems to have 2 sensors which are separated by a huge dead band. unless they took care to do multiple images to span the gaps, there are often huge swaths of missing data. the NGC1376 i did is a perfect example of that. the top panel has zillions of faint galaxies (and the top of NGC1376) but it's aesthetically unpleasing to have a big gap in the middle, so i just omitted the top panel.

maybe i will give this a shot again.