This sub-album contains images from album ProcessedOriginal and cleans up defects in each image. Processing notes explain what cleaning was done to each image. These cleaned images are assembled as layers into the final layer. I. Cleaning images overview Seven images from the prior processing step were edited in GIMP and cleaned for artifacts. Cleaned output files were: Green Channel: 201000115_origgreen_bit8.bmp -> 201000115_cleanedgreen_bit8.bmp Red Channel: 201000115_origred_bit8.bmp -> 201000115_cleanedred_bit8.bmp OIII Channel: 20100115_origOIII_bit8.bmp -> 20100115_cleanedOIII_bit8.bmp Ha Channel: 20100115_origHa_bit8.bmp -> 20100115_cleanedHa_bit8.bmp OIII-Ha Difference Channel: 20100115_origOIIIHaDiff_bit8.bmp -> 20100115_cleanedOIIIHaDiff_bit8.bmp Blue Channel: 201000115_origblue_bit8.bmp -> 201000115_cleanedblue_bit8.bmp Stars Only Channel: 201000115_origstarsonly_bit8.bmp -> 201000115_cleanedstarsonly_bit8.bmp All images were exported to native GIMP xcf format and to jpg. From the OIII-Ha Difference Channel, a small transparent layer containing on the highlight detail was created: 20100115_cleanedOIIIHaDiffHighlightLayer_bit8.bmp II. Specific cleaning on each image was: Green Channel: 201000115_origgreen_bit8.bmp Heal tool used to remove visually intrusive bloated stars and dust particles. Round stars will be added back in using 201000115_origstarsonly_bit8.bmp. For image purists, note the 10-25-2009 APOD of a Hubble M1 where the foreground stars have mysteriously disappeared. http://antwrp.gsfc.nasa.gov/apod/ap091025.html Clone and smudge tools used on five stars with large amounts of bloat that were beyond the Heal Tools capabilities. Three of these stars are at the edges and might be cropped off in future versions. Red Channel: 201000115_origred_bit8.bmp Heal tool used to remove visually intrusive bloated stars and dust particles. Clone and smudge tools used on five stars with large amounts of bloat that were beyond the Heal Tools capabilities. Three of these stars are at the edges and might be cropped off in future versions. Blue Channel: 201000115_origblue_bit8.bmp Only very minor edits. Most of the stars are slightly oval - probably from a registration error. Only four stars were removed - the two central stars of the nebula and on the edge. Star bloat in the blue channel is probably too minor for an unsophisticated eye, but an experienced amateur will notice it. Two minor dust particles were removed with the Heal tool. OIII Channel: 20100115_origOIII_bit8.bmp Heal tool used to remove visually intrusive bloated stars, dust particles and apparently an NEO or satellite trace right at the visual center of the image. Many small CCD hot pixels were removed. Clone and smudge tools used on a few stars with large amounts of bloat that were beyond the Heal Tools capabilities. This channel had to be heavily worked on - mostly due to hot pixels. Part of the central nebulosity ring is overexposed. May try to use the "OIII-Ha Difference Channel" to fix this later. Ha Channel: 20100115_origHa_bit8.bmp Serious "L"-shaped star bloat in image such that image may not be useable in final. Virtually every size - even faint magnitudes - have "L" shape bloat. Heal tool used to remove heavy star bloat where visually instrusive. Dust particles, NEOs and hot pixels were present. Part of the central nebulosity ring is overexposed. OIII-Ha Difference Channel: Only a small masked area of this image will be used to provide luminuous details to part of the central nebulous ring. First, the Heal and Clone tools were sued to remove many small hot pixels and NEO streaks around the nebulousity region of interest. Created "20100115_cleanedOIIIHaDiffHighlightLayer_bit8.bmp" - a small layer box with a transparent background. Stars Only Channel: 201000115_origstarsonly_bit8.bmp Image was gamma stretched to make the stars brighter. Anything not a star was removed. Then the stars where cut to a new alpha channel (transparent) layer, where they can be floated over the top of the final image. The stars alpha layer created "201000115_cleanedstarsonlyLayerAlpha_bit8.bmp". ========================================== Re: [Utah-astronomy] Wow!!! (SLAS Maui imaging session) Hola, como te va, Group? Below are my imaging notes on doing a first pass stack and deconvole on the Faulkes M1 session. The results of these processing steps are in Utah Astro Gallery directory: http://www.slas.us/gallery2/main.php?g2_itemId=2868 By posting these notes, I am working towards another version of the Crab from the Faulkes session and am simply thinking outloud about the image processing steps for the benefit of anyone how wants to follow along. Kibbutizing is welcome. The end product objective is an asethetic, not natural observation version. Most of the images are underexposed and have tracking problems. So, don't expect the result to beat or meet a professional telescope equipped with adaptive optics, e.g. - http://antwrp.gsfc.nasa.gov/apod/ap040128.html The idea here is get what can be teased out of the image and work with it. The end product of the original pass processing step are the 8-bit images, posted in lossless jpg but available from me as 8-bit fits: Green Channel http://www.slas.us/gallery2/main.php?g2_itemId=2869 201000115_origgreen_bit8.jpg Red Channel http://www.slas.us/gallery2/main.php?g2_itemId=2873 201000115_origred_bit8.jpg OIII Channel 20100115_origOIII_bit8.jpg http://www.slas.us/gallery2/main.php?g2_itemId=2875 Ha Channel 20100115_origHa_bit8.jpg http://www.slas.us/gallery2/main.php?g2_itemId=2877 OIII-Ha Difference image 20100115_origOIIIHaDiff_bit8.jpg http://www.slas.us/gallery2/main.php?g2_itemId=2879 Blue Channel 201000115_origblue_bit8.jpg http://www.slas.us/gallery2/main.php?g2_itemId=2883 Stars Only Channel 201000115_origstarsonly_bit8.jpg http://www.slas.us/gallery2/main.php?g2_itemId=2881 The next imaging step will be to work with these 8-bit images in GIMP, a freeware analogue to Photoshop, and clean up some of the artifacts in each image. Clean up will be done before the individual images are assembled into a layer stack. Since the nebulosity in the raw images was confined to less than 500 pixel values, there probably won't be any improvement by working with the images in an 8-bit (255 levels) as opposed to 16-bit levels. By posting my processing notes, I hope I'm not boring the group. Clear Skies - Kurt ---------------------------- I. Green channels A. Processing steps Individually clipped to 15-525 adus. Wavelet 5 step filter, no. 2 step high. Linear brightness stretch to 255 8bit levels Saved to fits with suffix clip_wave5bit8 B. Processing steps applied to l_e_20100115_20_1_1_1.fits Green 360s to l_e_20100115_20_1_1_1clipwave5bit8.fits l_e_20100115_23_1_1_1.fits Green 360s to l_e_20100115_23_1_1_1clipwave5bit8.fits l_e_20100115_26_1_1_1.fits to l_e_20100115_26_1_1_1clipwave5bit8.fits C. Combined processed green channels Registered and k-sigma combined - l_e_20100115_20_1_1_1clipwave5bit8.fits l_e_20100115_23_1_1_1clipwave5bit8.fits l_e_20100115_26_1_1_1clipwave5bit8.fits to 201000115_origgreen_bit8.fits D. Overall result Marginal or no improvement Comparing l_e_20100115_20_1_1_1.fits to 201000115_origgreen_bit8.fits II. Red channels A. Processing steps Individually clipped to 115-1100 adus. Wavelet 5 step filter, no. 2 step high. Linear brightness stretch to 255 8bit levels Saved to fits with suffix clip_wave5bit8 B. Processing steps applied to l_e_20100115_21_1_1_1.fits Red 360s to l_e_20100115_21_1_1_1clipwave5bit8.fits l_e_20100115_24_1_1_1.fits Red 360s to l_e_20100115_24_1_1_1clipwave5bit8.fits l_e_20100115_27_1_1_1.fits Red 360s to l_e_20100115_27_1_1_1clipwave5bit8.fits C. Combined processed red channels Registered and k-sigma combined - l_e_20100115_21_1_1_1clipwave5bit8.fits l_e_20100115_24_1_1_1clipwave5bit8.fits l_e_20100115_27_1_1_1clipwave5bit8.fits to 201000115_origred_bit8.fits D. Overall result Discernable improvement in sharpness, gradient detail and background contrast Comparing l_e_20100115_21_1_1_1.fits to 201000115_origred_bit8.fits III. OIII processing A. Processing steps Individually clipped to 15-250 adus. Wavelet 5 step filter, no. 2 step high. Linear brightness stretch to 255 8bit levels Saved to fits with suffix clip_wave5bit8 B. Processing steps applied to l_e_20100115_30_1_1_1.fits OIII 180s to l_e_20100115_30_1_1_1clipwave5bit8.fits l_e_20100115_31_1_1_1.fits OIII 180s to l_e_20100115_31_1_1_1clipwave5bit8.fits C. Combined processed red channels Registered and averaged combined (k-sigma and median require 3 images) - l_e_20100115_30_1_1_1.fits l_e_20100115_31_1_1_1.fits to 20100115_origOIII_bit8.fits D. Overall result Discernable improvement in sharpness. Highlights still overexposed. No improvement in detail gradient. Some artifacts added to image. Comparing l_e_20100115_30_1_1_1.fits to 20100115_origOIII_bit8.fits IV. Ha processing A. Processing Have only one image. This image was clipped to -15 to 500 adus and restretched to an 8 bit image. Processed l_e_20100115_29_1_1_1.fits HA 180s to 20100115_origHa_bit8.fits B. Overall result No discernable improvement. Image stretched to common bit size for further processing. V. Stars only image process A red image with relatively round stars was taken and clipped to remove the background nebula. The image was gamma log stretched to brighten the stars. This star only will be used as a mask to paste white stars over the top of the combined color images. Processed - l_e_20100115_21_1_1_1.fits to 201000115_origstarsonly_bit8.fits VII. Conversion of Fits images to bmp and jpg for processing in GIMP 8-bit processing in GIMP - a Photoshop freeware analogue - is used here. The processed images are exported from FITS to an 8-bit format useable as individual layers in GIMP. GIMP is then used to stack and assign colors and brightnesses to each layer. Processed (lossless) - 201000115_origgreen_bit8.fits -> bmp and jpg of the same prefix 201000115_origred_bit8.fits -> bmp and jpg of the same prefix 20100115_origOIII_bit8.fits -> bmp and jpg of the same prefix 20100115_origHa_bit8.fits -> bmp and jpg of the same prefix 201000115_origstarsonly_bit8.fits -> bmp and jpg of the same prefix VII. Searching for detail with special techniques Difference imaging sometimes can coax out hidden detail. Several combinations on the processed images were tried. The OIII minus Ha image yielded good detail of part of the central ring cloud. This image was set aside for possible use. Processing - 20100115_origOIII_bit8.fits minus 20100115_origHa_bit8.fits with gamma contrast increase to - 20100115_origOIIIHaDiff_bit8.fits This was also exported to bmp and jpg. VI. Blue image processing Initially, I decided not to process the blue channels because they do not increase any information contained in the green channel. On reflecting on final image composition, I preliminarily decided to try for a "cave like" appearance with a bright inner core along the lines of - http://antwrp.gsfc.nasa.gov/apod/ap061026.html or http://antwrp.gsfc.nasa.gov/apod/ap091025.html The foreground/background contrast is also better in the blue channel as opposed to the green channel. So, I decided to do a simple stack of the blue channel for possible future use. A. Processing steps Individually clipped to 15-525 adus. Wavelet 5 step filter, no. 2 step high. Linear brightness stretch to 255 8bit levels Saved to fits with suffix clip_wave5bit8 Processed - l_e_20100115_19_1_1_1.fits Blue 360s to l_e_20100115_19_1_1_1clip_wave5bit8.fits l_e_20100115_22_1_1_1.fits Blue 360s to l_e_20100115_22_1_1_1clip_wave5bit8.fits l_e_20100115_25_1_1_1.fits Blue 360s to l_e_20100115_25_1_1_1clip_wave5bit8.fits C. Combined processed red channels Registered and k-sigma combined - l_e_20100115_19_1_1_1clipwave5bit8.fits l_e_20100115_22_1_1_1clipwave5bit8.fits l_e_20100115_25_1_1_1clipwave5bit8.fits to 201000115_origblue_bit8.fits D. Overall result Barely noticeable improvement in sharpness and detail gradients. 201000115_origblue_bit8.fits compared to l_e_20100115_19_1_1_1.fits Exported - 201000115_origblue_bit8.fits -> bmp and jpg of the same prefix