Loading ...
Sorry, an error occurred while loading the content.

Re: How do you fix bloated stars? NEED HELP!

Expand Messages
  • Leonard E. Mercer
    Hi Ron, Thanks for your very deetailed reply. You gave me much food for thought. I have placed your reply on my desktop so as to refer to it often. Thanks once
    Message 1 of 3 , Aug 4, 2008
    • 0 Attachment
      Hi Ron,

      Thanks for your very deetailed reply. You gave me much food for
      thought. I have placed your reply on my desktop so as to refer to it
      often.

      Thanks once again.

      Leonard.
      --- In ccd-newastro@yahoogroups.com, "Wodaski - Yahoo" <yahoo@...>
      wrote:
      >
      > Let's look at two themes here:
      >
      > * Is star bloat really a processing problem?
      >
      > * Are your stars really bloated?
      >
      > Is Star Bloat a Processing Problem?
      > ================================================
      >
      > I could make a case that the best place to attach truly bloated
      stars is in
      > data acquisition, although there are some things you can do in
      processing to
      > make things worse or better (there are some things you can do that
      will
      > increase star size, and there are some tricks you can use to reduce
      star
      > size; more shortly).
      >
      > Among the things that make stars fatter:
      >
      > * Dirty optics will scatter light. Careful processing to bring out
      faint
      > details will also reveal with "dirty toe" around the base of the
      star
      > profiles.
      >
      > * Poor optics will both scatter light and give you a fatter FWHM.
      Same issue
      > as above: careful processing reveals the problem.
      >
      > * Poor focus - even just a tiny bit off - creates problems with
      bloated
      > stars as well.
      >
      > * Detector characteristics. Many detectors have other methods than
      blooming
      > that allow electrons to flow from bright pixels to dim pixels. Some
      are
      > artifacts of design decisions (e.g., anti-blooming circuitry) or
      simply
      > substrate characteristics. In addition, many detectors have some
      level of
      > significant reflection off of the substrate or circuitry that will
      spread
      > the light out from bright stars.
      >
      > Here are processing steps that can affect how bloated stars get:
      >
      > * When you align images, if you optimize for smoothness (e.g.,
      using a
      > bicubic method), you are breaking down the existing pixel
      structure. This
      > spreads each aligned image's star out just a bit. When you combine,
      you have
      > noticeably fatter stars. It is better to get not-quite-perfect
      alignment,
      > even though that will still fatten up your stars a bit due to non-
      integer
      > pixel alignment in the original images.
      >
      > * You can use an erosion type of filter in Photoshop to literally
      eat away
      > the outer pixels of the star. This tends to create non-round star
      shapes,
      > and an unnatural star profile results if you get too aggressive.
      You also
      > lose the natural contrast between stars - only the brighter stars
      can be
      > eroded; the little ones will disappear!
      >
      > * You can select just the stars in the image (all but the dimmer
      ones, which
      > usually have no detectable bloating due to poor S/N of the "dirty
      toe"), and
      > then use Curves to modify the star profile smoothly. This requires
      a deft
      > touch and practice. The technique is explained in complete detail
      in my Zone
      > System book (www.newastro.com/zone).
      >
      >
      > Are Your Stars Really Bloated?
      > ================================================
      >
      > Your justification for having bloated stars is that your RGB images
      have
      > fatter stars than your Ha images. I would propose that using your
      Ha images
      > as a basis for what you expect stars to look like is not a
      practical idea.
      >
      > Why? Well, the Ha filter is blocking nearly all of the starlight.
      So any
      > "dirty toe" has poor S/N and is much less likely to show up. The
      limited
      > amount of light will also limit blooming and other pixel-to-pixel
      electron
      > shifts that tend to fatten up stars. Additionally, the Ha filter is
      allowing
      > your optical system to focus just a single, narror band of light
      > wavelengths. Especially in refractors, even with APOs, this can
      > significantly improve your FWHM.
      >
      > In other words, when imaging full or even one-third full spectrum,
      you can
      > expect significantly higher FWHM than with Ha imaging.
      >
      > Now I'm not saying with certainty that your stars are fine; I'm
      just saying
      > that comparing your L or RGB images to your Ha images with respect
      to star
      > size is not a fair comparison; the rules are quite dramatically
      different
      > and the Ha wins every time on FWHM size. <G>
      >
      > I'm a bit of a pragmatist. If you look at pro images, you'll see
      immediately
      > that things like blooms, reflections, and star halos are not
      touched up at
      > all. (Well, maybe the teams that make the pretty pictures do some
      of this,
      > but usually much less than we amateurs.) They are simply artifacts
      of the
      > imaging process. Since the S/N is always poorer in such areas, I am
      inclined
      > to leave them in my images these days so that the viewer knows
      where the
      > good stuff is and isn't!
      >
      > If you do want to alter the image to clean up artifacts, the
      methods are
      > there (including the ones I document in various forums and
      publications),
      > and I have to admit that my opinion on what to do about artifacts
      changes
      > back and forth so I'm not an absolutist on that front. ;-)
      >
      > My own favorite star blot reduction method is using Curves on a
      stars-only
      > selection in Photoshop (or, in some cases, such as when there is a
      lot of
      > nebulosity or big differences in degree of bloat) and then use
      Curves to
      > carefully alter the star's brightness profile. This takes time, but
      a
      > careful application gives you a very, very high degree of control
      so that
      > you can balance the various issues in play.
      >
      > Ron Wodaski
      >
      > -----Original Message-----
      > From: ccd-newastro@yahoogroups.com [mailto:ccd-
      newastro@yahoogroups.com] On
      > Behalf Of Leonard E. Mercer
      > Sent: Sunday, August 03, 2008 10:20 AM
      > To: ccd-newastro@yahoogroups.com
      > Subject: [ccd-newastro] How do you fix bloated stars? NEED HELP!
      >
      > Hi,
      >
      > As I improved my processing with time, I have now encountered a
      > problem which is really trying me out. That of bloated stars. This
      > problem is spoiling good images of mine and it seems I cannot solve
      > it no matter what I try. My images are mainly Ha,RGB or Ha,HaR,GB.
      >
      > I realised that the problem is imerging as soon as I add the Ha to
      > the RGB image.
      >
      > I would be very grateful if someone could give me some hints as to
      > how I can solve this problem by reducing the bloating.
      >
      > I tried Ken Crawfords method of reducing star halos, but got no
      > improvement.
      >
      > I appreciate any help.
      >
      > Thanks & clear skies,
      >
      > Leonard.
      >
      >
      >
      > ------------------------------------
      >
    Your message has been successfully submitted and would be delivered to recipients shortly.