P1258768823riWab

1
MWAIC 2008 A Narrowband Processing
Workflow Saturday, June 21, 2008 Neil
Fleming (www.flemingastrophotography.com)
2
Agenda

• Light pollution? Why even try?
• Differences between regular RGB imaging and
  Narrowband imaging
• Capturing good data
• Initial processing workflow CCDStack
• Final processing workflow Photoshop
• My goal in this presentation is not strictly
  scientific, it is more the presentation of an
  enjoyable result

3
Imaging from LP Locales Why Even Try!?!?

• Typical raw and semi-processed RGB results from
  Boston
• I waserdisappointed!

4
Then I Found Narrowband Imaging
5
What is RGB Imaging?

• RGB stands for Red, Green, and Blue
• This mix of primary colors is what our eyes use
  to interpret color
• All imaging starts with a black and white CCD or
  a one-shot color (OSC) CMOS sensor
• Monochrome cameras are used in conjunction with
  filters

• Your DSLR or OSC cameras utilize tiny
  red/green/blue (RGB) filters over the individual
  pixels, placed in a Bayer matrix pattern

6
Whats the Difference?

• With either a black and white camera or a OSC
  camera, you use color filters to emulate the RGB
  results so you can obtain a color image
• The wavelengths captured are across the entire
  visible spectrum

7
What Does RGB From Boston Look Like?

• Light pollution typically causes horrible
  gradients
• This is a sample of a stack of 3x8 minute
  exposures

• A little processing in Photoshop helps, but
  still, not great results
• Bright objects are better

8
Narrowband Imaging

• Narrowband filters let through only the tiniest
  bandpass of light, light that is associated with
  emission lines in nebulae. Typically 3 to 13 nm
  in width
• Hydrogen-alpha (Ha) emission wavelength is 656.3
  nm
• Doubly oxidized oxygen (OIII) has its main
  emission line at 500.7 nm
• Sulfur (SII) emits at 672 nm

9
Image Capture

• Go deep
• Go for a greater subexposure (sub) duration and
  overall exposure time than for RGB imaging
• My RGB subs would average 1-4 minutes in exposure
  time
• For narrowband, I shoot at least 30 minute subs
• Get lotsa data
• I aim for 18-30 hours of usable data for an
  object
• I strive for at least 12 subs per channel, more
  if possible
• This allows for a reasonable signal-to-noise
  ratio (SNR), and includes enough subs in the
  stack to do efficient data rejection of outliers
  (satellites, airplanes, and cosmic ray hits)
• Gather ample calibration frames
• 30 bias, 25-30 darks, 1,000,000 ADU of flats
  (30 for me)
• I get automated dawn flats

10
Processing Workflow CCDStack

• Prepare calibration masters bias/darks/flats
• Load and calibrate your subs
• Apply DDP, and evaluate the quality of the subs
• Bloom rejection
• Registration
• Normalization
• Data rejection
• Channel master combines

11
CCDStack Calibration Masters

• Dark Frames
• These are used to subtract out the effects of
  hot pixels
• Optimally, these are taken at the same duration
  and camera temperature as your light frames
• Bias Frames
• Zero-duration dark frames used to time scale
  darks, and as a proxy dark frame for flats
• Flat Frames
• Used to accommodate light fall-off at the edges,
  as well as to eliminate dust motes
• Prepare calibration masters bias/darks/flats
• Use some sort of sigma rejection or clip min/max,
  don't use mean
• This eliminates the impact of outliers like
  cosmic ray hits (all three types) and stars (for
  the flats)
• Bias-subtract your flats when you create your
  flat master to accommodate differing temperatures
  later

12
CCDStack Load and Calibrate Your Subs

• Load all of your subs into CCDStack
• Under Process, select Calibrate
• Select your appropriate dark, bias, and flat
  masters
• Apply to all

13
Sample Uncalibrated Sub
14
The Same Sub Calibrated
15
CCDStack Evaluate Sub Quality

• Rotate all of the subs to the same orientation
• Carefully evaluate your data!
• CCDInspector for contrast, aspect ratio, and FWHM
  evaluation
• Mark I eyeball as a second step, especially if
  your data is undersampled, for gradients and star
  aspect ratio
• Good / Marginal / Bad
• I discard the Bad subs, and keep the Good
  along with a few of the Marginal
• The larger the stack of good subs, the more of
  the marginal I can include

16
CCDStack The Good, the Bad, and the Ugly
Bad
Good
Marginal
17
CCDStack Bloom Rejection

• Process, Data Reject, Procedures
• Select, Reject Blooms
• Set appropriate upper limit, e.g., 30000 ADU
• Apply to All
• Impute Rejected Pixels
• I use 0.2 pixels, with 3 iterations
• Apply to All

18
CCDStack Bloom Rejection
Before
Rejected Bloom
Pixels Imputed
19
CCDStack Image Registration

• I often image the same object over multiple
  nights
• This results in a little offset or a slight extra
  rotation between subs from each night
• Go under, Stack, Register

20
CCDStack Image Registration

• Under the Star Snap tab, click on Select
  Reference Stars
• I pick 3 to 4 widely spaced, medium sized stars
• I then click on, Align All
• Blink through the stack to ensure that all subs
  are well-registered
• Sometimes I will first try a pass with two
  closely spaced bright stars, then do a second
  pass with the 3 to 4 widely spaced, medium sized
  stars (Dual-pass method)
• When aligned, move to the Apply tab and select
  a method for registration, like Quadratic
  B-Spline, and click, Apply to All

21
CCDStack Image Registration
22
CCDStack Image Registration
23
CCDStack Image Registration
Unregistered
24
CCDStack Image Registration
Registered
25
CCDStack Image Normalization

• This is used to balance the individual subs
  contribution to the final combine
• The higher quality subs will contribute more,
  while the lower quality subs will contribute less
• Go to, Stack, Normalize, Auto, and click,
  OK

26
CCDStack Data Rejection

• Data Rejection
• CCDStack allows you to reject poor data like
  satellite trails, cosmic ray hits, and airplane
  trails independently of the combine method!
• You do not have to rely on mean, median, etc., to
  get rid of these pests!
• Options include
• STD Sigma
• Poisson Sigma
• Each of these methods will throw out the
  outliers and average the remaining pixel values
• I often use the Poisson Sigma reject, with 1.6
  to 2 standard deviations (sigma multiplier), or
  clip min/max for deep stacks
• Larger stacks can take tighter tolerances

• Linear Factor
• Clip Min/Max

27
CCDStack Rejected Data a Good Subexposure
28
CCDStack Rejected Data a Lower Quality Sub
29
CCDStack Subs Rejected Pixels
30
CCDStack Channel Master Combine

• Data combine
• You can do any of the following
• Sum
• Mean
• Median
• Minimum
• Maximum
• I almost always use, Mean for complete data
  sets
• Ill use Sum if utilizing data sets in further
  combine steps

31
CCDStack Mean Combine the Subs
Single Sub
Master Combine
32
CCDStack Deconvolution

• I prefer the Positive Constraint algorithm,
  with 25 iterations
• Choose a medium star on a dark background

Original
Deconvolved
33
The Ha Channel Master is All Done
34
CCDStack Do the Other Panel

• Now do the OIII panel
• After calibration and rotation, when I start
  registration, I re-load the Ha master and use
  that one to register the OIII subs
• This allows for only one destructive
  registration step to be applied, not two!
• You now have your Ha, and OIII master panels

35
CCDStack Our Final Channels
Ha
OIII
36
CCDStack Preparation for Photoshop

• For narrowband work, unlike RGB, I prefer to
  "optimize" each channel in Photoshop (PS) before
  the color combine
• This helps to maximize the contribution from each
  data set, especially the OIII and SII
• So, at this point, I will save the FITS as a
  "scaled TIF" from CCDStack for each of the
  channel masters
• This squeezes or translates the umpteen million
  data values into the 16-bit color space

37
CCDStack Preparation for Photoshop

• Save each channel master as a 16-bit scaled TIF
• I open each TIF in PS, and closely examine the
  histogram to make sure I have not clipped the
  data at either end
• I find the scaling process in CCDStack will
  usually clip just a bit on the dark end of the
  histogram
• So, I will go back into CCDStack, lower the dark
  value cutoff a bit, and re-save the scaled TIF
  until I am satisfied

38
Photoshop Channel Optimization

• To "optimize" each mosaic panel in Photoshop (PS)
  before the color combine, I generally do
• A contrast curve adjustment layer, although one
  is not needed this time for the Ha data
• Noise reduction layer with an inverted layer mask
• Local contrast enhancement layer, created with
  Noel Carbonis Photoshop actions
• Sometimes a Shadow/Highlight adjustment layer
  (in PS-CS2 or higher)

39
Photoshop Noise Control

• Ha data - we need some noise control here!

40
Photoshop Noise Control

• Duplicate your image layer, and use NeatImage

41
Photoshop Noise Control
42
Photoshop Noise Control

• Control application of your results by utilizing
  an Inverse Layer Mask
• This uses the characteristics of the image itself
  on the layer mask to allow the smoothing come
  through in dark areas, and to be lessened in the
  bright areas (areas of good SNR)
• Create a reveal all layer mask, for your
  de-noised layer, then uncheck the link icon

43
Photoshop Noise Control

• Click on your main image again, select all
  (CTLA), then copy it (CTLC)
• Go to edit your layer mask directly on the
  layer mask,(ALTClick)

44
Photoshop Noise Control

• Paste in (CTLV) the image right into the layer
  mask, then deselect (CTLD)

45
Photoshop Noise Control

• If you recall your mask principles, the dark
  areas in the mask do NOT come through, but the
  white areas do
• We want to reverse the effect we see make the
  mask for the background areas light, and the mask
  for the nebula areas dark
• This restricts the effect of the de-noising
  layer to the dark (noisy) areas of the image, and
  restores all detail to the bright areas
• So, while still editing the layer mask, invert
  the image by hitting CTLI

46
Photoshop Noise Control

• Remember, we are still editing the layer mask, so
  this inversion impacts the layer mask only

47
Photoshop Noise Control

• In order to accentuate the mask, go under Image
  and Adjustments and use a curves layer to make
  the whites whiter and the darks darker

48
Photoshop Noise Control

• Were donethe dark areas of the image include
  the de-noised results, while the good SNR areas
  retain the original details
• Click back on your image layer to see the
  results!

49
Photoshop Local Contrast Enhancement

• Local Contrast Enhancement with Noel Carbonis
  Photoshop Actions
• This works to increase the contrast in the
  mid-range of the image

50
Photoshop Shadow/Highlight Adjustment

• Duplicate your image
• Flatten the result
• Rename the layer as Shadow/Highlight
• Invoke the Image / Adjustments /
  Shadow/Highlight adjustment
• Adjust to bring the darks up to taste, and the
  highlights down to taste
• When done, SHIFTdrag the layer back into your
  original document

51
Photoshop Shadow/Highlight Adjustment
52
Photoshop Our Final Real Channels
Ha
OIII
53
Photoshop Create the Synthetic Blue Channel

• We have the Ha and OIII done
• I will use the Ha as Red, and the OIII as
  Green
• I need to create data for Blue
• For this, I will use 100 of OIII, plus 20 of Ha
  (as a proxy for the H-beta component of hydrogen)

54
Photoshop Create the Synthetic Blue Channel

• Duplicate the OIII document, save as Synthetic
  Blue
• SHIFTDrag the Ha data into the Synthetic Blue
  document
• Change the opacity of the Ha layer to 20

55
Photoshop Create the Color Image

• Open the Ha, OIII, and sBlue documents
• Flatten each document
• Change the Image / Mode to 16-bit Grayscale
  for each document
• On the Channel tab, go to Merge Channels

56
Photoshop Create the Color Image

• Do an RGB Channel Merge
• Select the appropriate channel for the right color

57
Photoshop Create the Color Image

• Here is the initial result, OK for a start

58
Photoshop Color Adjustments

• Id like to make the greens more teal
• So, I first select the nebula area

59
Photoshop Color Adjustments

• I use the Eyedropper tool to select a spot on
  the green color that I would like to adjust
• Now, Select by Color Range

60
Photoshop Color Adjustments

• Expand the selection (8px)
• Use the Lasso Tool in either or mode to
  adjust your selection
• Then feather (8px)

61
Photoshop Color Adjustments

• Try a Selective Color adjustment layer to
  adjust per taste
• Greens Neutrals
• Cyans Whites

62
Photoshop Color Adjustments
63
Photoshop Color Adjustments

• Lets reduce the saturation of the reds, shall we?

64
Photoshop Color Adjustments

• A Shadow/Highlight layer to do a high dynamic
  range adjustment

65
Photoshop Color Adjustments

• Sharpeningdeconvolution in Tria
• Plus Noel Carbonis Photoshop Action

66
Photoshop Color Adjustments

• Overlay RGB stars into your document
• As before, align your R, G, and B data to the
  original Ha master
• Combine the RGB via your preferred method
• In CCDStack
• In Photoshop
• Process as normal, like noise control, star color
  accentuation, gradient control
• I select by color range to select only the
  stars, expand the selection by 2 pixels, and
  feather by one

67
Photoshop Color Adjustments

• Bring your masked star layer into your main doc

68
(No Transcript)
69
Questions?
MWAIC 2008 A Narrowband Processing
Workflow Saturday, June 21, 2008 Neil
Fleming (www.flemingastrophotography.com)