Author Topic: HLA images  (Read 3057 times)

Offline Eddy Timmermans

  • PixInsight Addict
  • ***
  • Posts: 149
    • View Profile
HLA images
« on: 2013 September 23 01:37:39 »
Many HLA images have a blank bar running across it. Others arre very grainy or have colored dots running across it.
This makes it difficult, not to say impossible, to create mosaics with these images.
The guys from Hubble however manage to do it.

Is it possible to do this with PI ? Or can this only be done by the Hubble guys ?

Eddy

Offline marekc

  • PixInsight Addict
  • ***
  • Posts: 177
    • View Profile
Re: HLA images
« Reply #1 on: 2013 September 23 08:04:50 »
Hi Eddy,

I'm woefully unqualified to answer your question, but you've raised some points that I've wondered about, too.

I've seen a number of beautiful-looking images from the Hubble Legacy Archive, which have been processed by amateurs using a variety of tools. From time to time, I've thought `I should go get some of those marvelous HLA images, and process them in PI.' Then I actually look at some of the HLA images, and I'm always a bit shocked by what I see. Despite the large aperture, sensitive detectors, and position in space of the HST, the images are always noisier-looking than I would have suspected, and have a lot more cosmetic defects than I would have suspected.

I guess I had an unrealistic expectation: I envisioned the HST acquisition workflow as being like that of an amateur imager: Shoot lots and lots of subexposures, stack them, and get a resulting image that has an extremely smooth background. And we've all heard of the massive total integration times that went into things like the Deep Field images.

But I think that's not really how a lot of HST data are acquired. I get the impression that the STScI allocates just enough integration time to get the job done, and not to make a dataset that we would find easy to work with.

I might be way off the mark here, but my overall impression is that if we, as amateurs, want to work with data such as that from the HLA, we're faced with a job that's much more like `painting' than like the `documentary school of astrophotography' that the PTeam espouses. PI is very powerful, but as Juan and others have told us, it's not really designed for the more painterly-type touch-up tasks that PS is good at. And, sadly, I'm coming to believe (perhaps wrongly?) that HLA and other `professional' data need an awful lot of painting-type touch-up work, in order to look presentable. Again, I could be approaching this from a state of unfortunate misconception, but I've found my enthusiasm (for working with `pro' datasets) considerably dampened once I actually look at those data. My PI knowledge - modest though it is - just doesn't seem to apply well to them.

- Marek

Offline NGC7789

  • PixInsight Old Hand
  • ****
  • Posts: 391
    • View Profile
Re: HLA images
« Reply #2 on: 2013 September 23 11:04:48 »
After attending a lecture about using HLA images by Robert Gendler and trying it a bit myself, I learned two things. The HLA data set, like most research data sets, are geared toward specific research needs and not making pretty pictures. The pretty pictures that do come out of this work involve combining the data with other data that IS pretty.

Offline chris_todd

  • PixInsight Enthusiast
  • **
  • Posts: 89
    • View Profile
    • I don't have hobbies, I have obsessions...
Re: HLA images
« Reply #3 on: 2013 September 23 13:51:31 »
Check out these threads, because they demonstrate that HLA data *can* be processed entirely in PI to generate beautiful (and contest winning) pictures:

http://pixinsight.com/forum/index.php?topic=4052.msg29507#msg29507

http://pixinsight.com/forum/index.php?topic=4564.msg31745#msg31745
Uncooled, unmodified Canon T2i/550D, various lenses, and AstroTrac TT320X-AG
I don't have hobbies, I have obsessions...

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4103
    • View Profile
Re: HLA images
« Reply #4 on: 2013 September 23 16:55:14 »
my understanding of this is that a very large fraction of HST observing time is allocated to astronomers/astrophysicists doing science, and their goal is not necessarily to make pretty pictures. lots of times they use filters that don't necessarily correspond to the LRGB filters normally used by amateurs.

however, nasa/stsci knows that they need to make beautiful pictures in order to keep up public support for the project. to that end once in a while zolt & co get to schedule orbits dedicated to optical imaging, with enough integration time to make low-noise images.

finding those projects is a little difficult but i assume that if you search for z. levay or l. frattare as the PI for the project you might find the projects that were meant for "the rest of us".

http://hubblesite.org/gallery/behind_the_pictures/

has a little bit about the odd shapes of the detectors, etc.

rob

Offline chris_todd

  • PixInsight Enthusiast
  • **
  • Posts: 89
    • View Profile
    • I don't have hobbies, I have obsessions...
Re: HLA images
« Reply #5 on: 2013 September 24 07:01:33 »
my understanding of this is that a very large fraction of HST observing time is allocated to astronomers/astrophysicists doing science, and their goal is not necessarily to make pretty pictures. lots of times they use filters that don't necessarily correspond to the LRGB filters normally used by amateurs.

however, nasa/stsci knows that they need to make beautiful pictures in order to keep up public support for the project. to that end once in a while zolt & co get to schedule orbits dedicated to optical imaging, with enough integration time to make low-noise images.

finding those projects is a little difficult but i assume that if you search for z. levay or l. frattare as the PI for the project you might find the projects that were meant for "the rest of us".

http://hubblesite.org/gallery/behind_the_pictures/

has a little bit about the odd shapes of the detectors, etc.

rob

Rob,
You're correct, the Hubble Heritage team gets some observing time on Hubble specifically to produce images for public consumption.  My favorite example was the fairly recent infrared Horsehead:
http://hubblesite.org/newscenter/archive/releases/2013/12/image/a/

One of the advantages of living near Baltimore is that our astronomy club can get speakers from STScI from time to time, and last Thursday, Zolt Levay spoke to our club about how they produce the "pretty pictures" that go with most Hubble press releases, and he went into some depth about how they produced that infrared Horsehead.  He also discussed how amateurs like us can troll the HLA for images we can process.  He is an excellent speaker and a really nice guy.

Chris
Uncooled, unmodified Canon T2i/550D, various lenses, and AstroTrac TT320X-AG
I don't have hobbies, I have obsessions...

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4103
    • View Profile
Re: HLA images
« Reply #6 on: 2013 September 24 11:32:19 »
yeah, that horsehead was amazing. the image release seemed kind of "fragmented" though - there was a square-cropped version of just the horsehead itself that got a lot of publicity... but there was also a wider-field view of the surroundings which was amazing. you can see galaxies behind the Ha area since i guess much of the image was shot with IR filters.

in general, i guess because of all the different entities involved with the HST, it seems like there are too many different websites and archives of data out there which confuses things, at least for me.

although i admit that the audience for raw HST data is probably pretty small and so there's not much reason for the powers that be to simplify things.

rob