Imaging Technique for the Amateur Astronomer

1
Imaging Technique for the Amateur Astronomer

• by Frank Barrett
• Jan 27, 2007
• 15th Regional Gathering of Amateur Astronomers

2
Objectives

• Overview of Astro Imaging Requirements
• Emphasis on Imaging Technique
• If time permits, A Short Processing Demo
• End with a slide show

3
The Question

• What is required to take really good astronomical
  images?
• ???

4
The Answer
Priority
Patience
Time
Persistence
Equipment
Tools
Software
Acquisition
Technique
Processing
5
Imaging Technique

• Technique a method of accomplishing a desired
  aim - Merriam-Webster
• so when we speak of technique we are speaking of
  a methodology or perhaps a framework of
  methodologies working together to accomplish our
  goal
• Really Good Images!
• --gt Tools are NOT Technique

6
Two Phases of Technique

• Acquisition
• How do we use our tools (equipment software)
  when preparing and recording images?
• Usually done under the stars.

• Image Processing
• What can we do with the image data acquired to
  maximize signal and reduce noise?
• Usually done on the computer.

7
Acquisition Techniques

• Polar Alignment
• Guiding, PEC
• Balance
• Focus, focus, focus!
• Location
• light pollution
• latitude
• horizon
• weather
• Seeing/Thermal effects
• pavement vs. grass
• cool down time

8
Acquisition Techniques

• Dew Prevention
• Target Location
• low on horizon?
• over a chimney?
• meridian flip?
• Optical configuration
• focal length/FOV
• focal reducer
• Barlow
• mosaic preparation

9
Acquisition Techniques

• Imaging
• sub exposure duration
• total number of subs
• darks, flats
• bin factor
• filters
• light pollution
• narrow band
• color
• dithering

10
Processing Technique

• Two main phases to image processing
• 1) Noise reduction (both random and non random)
• 2) Signal enhancement
• Noise - an unwanted signal or disturbance in an
  electronic device or instrument irrelevant or
  meaningless data or output occurring along with
  desired information
• Signal - a detectable physical quantity or
  impulse by which messages or information can be
  transmitted
• (ref Merriam-Webster, http//m-w.com)

11
Noise Reduction

• Combatting Non Random Noise with Image
  Calibration (a.k.a. Reduction)
• 1) Darks - every electronic imaging device
  produces noise due to dark current. An image
  taken in the dark for the time and temperature
  of your light frames can be subtracted to remove
  this form of noise.
• 2) Flats - another form of unwanted noise may
  come from the optics of our system. Dust, dirt,
  and uneven illumination can conspire to produce
  unwanted artifacts in our images. A flat is an
  evenly illuminated image which can be used as a
  baseline for removing these artifacts.

12
A Sample Dark Frame
13
A Sample Flat Frame
14
An Uncalibrated Light Frame
15
After Dark Subtraction
16
After Flat Frame
17
Noise Reduction

• Combatting Random Noise by Combining Frames
• The predominant noise in a calibrated frame is
  random in nature...it will be different from one
  subframe to the next.
• Note that the signal is constant, but the noise
  is random.
• We can leverage this knowledge by mathematically
  combining the pixels of a stack of such images.
• Averaging N subframes will decrease noise by
  SQRT(N).
• Note subframes have to be properly calibrated
  and aligned before combining.

18
Noise Reduction

• Combination Algorithms
• Simple Average - Add up all the pixel values from
  the stack and divide by N.
• Produces a very smooth image, but may allow
  effects due to satellite trails or cosmic ray
  hits to bleed through.
• Median - sort all the pixel values from the stack
  and select the one in the middle.
• A bit noisier than average, but does excellent
  job removing hot and cold pixels and other
  unwanted artifacts such as satellite trails or
  cosmic rays.
• Sigma Clip - remove out-liers (both high and low)
  and average the remaining pixels.
• Gives the best of both worlds. Needs at least 6
  subframes to be effective.

19
Noise Reduction

• Combatting Non Random Noise by Dithering Frames
• Dithering is an acquisition technique where the
  guide star is purposefully moved from one
  subframe to the next.
• This has the effect of shifting the signal around
  on the imaging chip.
• After alignment any non random noise from the
  camera will be located at different locations of
  the image stack and will be removed nicely by the
  combination algorithm.

20
Noise Reduction

• A Final Note
• All subframes contain noise. This includes darks,
  flats, and light frames.
• For this reason all Frame types should be taken
  with multiple subframes and combined!
• Furthermore, Flat frames should be calibrated
  with their own darks, and light frames should be
  calibrated with darks and flats.
• A typical frame inventory
• w Flats (one set per filter)
• x Darks for Flats
• y Lights
• z Darks for Lights

21
Signal Enhancement

• Once we have done a good job of removing noise we
  can focus our attention on bringing out the
  signal of our target.
• Our primary goal here is to bring out the detail,
  color, and character of the underlying data. We
  strive not to distort, but to enhance.
• Please Note It is impossible to remove all the
  noise and therefore we need to be very careful
  that we enhance the signal and not the remaining
  noise.

22
First Step Histogram Stretch

• Most of the signal in a typical deep sky image is
  in the low end of the histogram
• There are many software techniques to perform
  histogram stretch. I prefer to use Adobe
  PhotoShop's Levels and Curves because it gives
  more control over the result.

23
Histogram Stretch with Levels and Curves

• The idea here is to bring up the low end of the
  histogram while preventing burn out in the high
  end.
• After applying an iteration or two of Curves it
  is usually necessary to adjust the black point
  with Levels.

24
Next Step Cleanup

• Despite our best efforts to remove noise there
  will still be unwanted artifacts in the image
  that should be removed at this time.
• To remove hot and cold pixels, specks, splotches
  or other nasties I prefer to use PhotoShops
  clone tool.

25
Next Step Separating Zones

• Ron Wodaski has written an excellent image
  processing book entitled Zone System for Astro
  Imaging whereby the image is separated into 3 or
  more zones based on the signal to noise ratio of
  that zone.
• The darkest areas of the image contain almost no
  signal and therefore contain the lowest signal to
  noise ratio.
• Conversely, the brightest areas of the image will
  contain the highest signal to noise ratio.
• The idea is to separate these zone out so that we
  can apply different processin techniques to each
  of the zones. This is typically done in
  PhotoShop with the Color Range Selection Tool and
  Layers.

26
Zone Processing

• The stars. Creating a layer with just the stars
  allows us to be aggressive in processing other
  zones without destroying any background stars.
• Dark Zone. This zone typically contains almost
  all noise and can be defeated by raising the dark
  point in Levels.
• Dim Zone. These are the faintest areas of our
  image. We can best mitigate the noise here by
  applying a mild blur to soften the noise.
• Mid Zone. If the image has a wide dynamic range
  I will create this zone from tones between the
  Dim Zone and Bright Zone. This zone typically
  does not require any smoothing, but is not
  detailed enough to warrant sharpening.
• Bright Zone. This zone will contain the
  brightest and perhaps most detailed portions of
  the image. Here we will want to apply
  sharpening. I like to use PhotoShops Unsharp
  Mask.

27
So What?

• For most of us our time is limited and we do
  astronomy on a budget.
• Therefore we have to fit in with the time
  available and make do with the equipment we have.
• The variable with which we have the most control
  is therefore Technique.
• The best way to go about developing your
  Technique is to focus on problem areas.
  Continuously ask yourself,
• What could I have done to make this image even
  better.
• If you decide to stay in it for the long haul you
  will never stop asking that question and your
  images will continue to improve!
• When you hit that brick wall (and you probably
  will) you may be faced with the fact that you
  have reached the limits of what your time and/or
  tools will allow. Either accept that or go about
  making improvements.

28
See more athttp//celestialwonders.comcontact
me atfrankb_at_celestialwonders.com
29
Bibliography

• The New CCD Astronomy by Ron Wodaski
• The Zone System for Astro Imaging by Ron
  Wodaski
• Photoshop Astronomy by R. Scott Ireland
• Photoshop for Astrophotographers by Jerry
  Lodriguss
• CCD Topics by Stan Moore http//www.stanmooreastr
  o.com/CCD_topics.html
• Technical Notes by John Smith http//www.hiddenlo
  ft.com/notes.htm
• SBIG Yahoo Group http//tech.groups.yahoo.com/gro
  up/SBIG
• Great CCD Software http//www.ccdware.com/
• CCDSoft, TheSky http//www.bisque.com/Products/
• FocusMax, PoleAlignMax http//users.bsdwebsolutio
  ns.com/larryweber/
• Cartes du Ciel http//www.stargazing.net/astropc/