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

74376Re: New to the Group!

Expand Messages
  • stan_ccd
    Aug 1, 2014
    • 0 Attachment
      Hey Joe,

      I like your attitude!  (intellectual curiosity and impulse to experiment).

      That's a nice stable of scopes and though they are all smallish they do not substantially overlap each other (though you will get aperture fever after struggling with DSO for awhile).  Of course the DOB is not really suitable for DSO imaging but the other 3 will produce good results. Aperture is extremely important for DSO object imaging, the exception being wide field starscapes / nightscapes, which is done with small refractors or camera lenses.  

      Most DSO imagers begin with short FL (usually means small aperture) so as to learn the basics without having to deal with the substantial issues of tracking/guiding/seeing/etc. that arise when attempting to do high resolution work.  So start with the M603 or AP; usually refractors are the least troublesome.
      You don't say what camera(s) you use.  A "real CCD" (cooled, mono)?  DSLR?  Hybrid (OSC CCD)? Each of those classes have substantially different capabilities and require different acquisition and processing, as well as optimal target types (e.g. DSLR/OSC is awful for narrow band)

      As for FFT, it has very limited usefulness in regards to DSO.  Wavelet processing is virtually useless (beyond the equivalent of unsharp masking) because it demands much higher S/N than is feasible for dim DSO (lunar/planetary has very significantly higher S/N because the targets are dramatically brighter).  Similarly, FFT blurring ("de-noising") is of questionable value and is usually no better than a Gaussian blur (and often worse).  I have experimented with FFT methods for DSO and have come to the (non-final) conclusion that FFT is primarily useful for frame alignment (FFT auto-registration is a clever and fascinating application of multiple level FFT applications - it utilizes the FFT of an FFT) and resampling (e.g. Lanczos/sinc), though resampling need not be done in the frequency domain.

      Something to keep in mind when processing DSO data is that there are real and severe physical limits on what can be done.  Only so much information (with unavoidable uncertainty) is collected and trying to push beyond the real information results in phony images (maybe "pretty" to some people, but phony nonetheless).  IMHO too many practitioners waste far too much time trying to polish the image (usually via Photoshop but also more esoteric software).  You really cannot "make a silk purse from a pig's ear" so beware of spending more time trying to salvage insufficient or lousy data than spending time and learning to acquire data of high quality and quantity.

      As a programmer you might like to try your hand at processing. FITS files are easy to read and write (there are several web resources that document the file standards).

      As for software, I invite you to tryout CCDStack (30 days full use free then unlimited evaluation time), for which I am the developer.

    • Show all 11 messages in this topic