SaratogaSkies Jim Solomon's Astropics


Latest news

December: Test shots with new scopes/mounts

Dec 21: TMB 80/480 Arrives!

Dec 3: AP1200 Arrives!

Nov 30: TMB 152/1200 Arrives!


C/2004 Q2 [Machholz] (Comet)

Image Details:

Camera: Canon Digital Rebel (300D)
Mount: Celestron AS-GT
Scope: Orion ST80 (80mm f/5 Achromatic Refractor)
Configuration: Prime Focus
Additional Optics: n/a (Prime)
Filter: None
Effective Focal Length: 400mm
Effective Focal Ratio: f/5
Exposure: 7 x 4min @ ISO 400
Total Exposure: 0hrs, 28min
Date: 1/12/2005, 9:31pm
Location: Saratoga, CA, USA
Acquisition: DSLRfocus
Focus: DSLRFocus
Dithering: None
Guiding: GuideDog via Philips ToUcam Pro II (840k) through Celestron C8-N (8" f/5 Newtonian)


  • IRIS: Dark subtraction, flat field, registration, cropping
  • JimP: Minimum N of M stacking
  • Photoshop: Compositing, levels, image scale, JPG conversion

Image Description:

This is Comet C/2004 Q2 (Machholz), named after the San Jose Astronomical Association (SJAA) member Don Machholz who discovered it. See The Discovery of Comet Machholz in the SJAA newsletter for an interesting read. Note the ion and dust tails in this image. The faint ion tail (right-to-left) is caused by ionized particles glowing from the solar wind (the sun is thus to the right, off camera of course ), while the dust tail is caused by debris being left behind by the comet's motion (the comet is moving mostly up, and ever-so-slightly to the right). North is up.

Recall from the writeup in my image of C/2001 Q4 (Neat), a straight-forward "stack" of comet frames would result in a smeared comet, smeared background stars, or both, since the comet is moving fairly rapidly with respect to those background stars. The basic technique in preventing either from being smeared is to create two separate stacks: one of the comet (with no background stars), and one with the background stars (with little or no comet), then to combine the two using layers in Photoshop or the like. The rest of the text below describes how I processed this image, and discusses why an achromatic refractor is perhaps not the best choice for comet imaging.

I did the "stacking" in my own hand-written piece of software that I affectionately call JIMP (Jim's Image Manipulation Program, with all due reference to GIMP). In order to remove the (smeared) stars from the comet-registered frames, I stacked them using an algorithm that selects the "Smallest N of M values" at each pixel location, where N is the number in the resulting stack and M is the total number of frames. (I had M=10 useful frames before the clouds ruined the fun, and settled on N=7 after some experimentation.) This algorithm seemed to strike a nice balance between removing the stars (typically the maximum values at each location) while still providing plenty of remaining signal to stack. I found that it worked slightly better than any other stacking method I tried, including Median, Sum, Kappa-Sigma, and Min. Since IRIS doesn't support this "Smallest N of M" stack, I had to implement it myself. Enter JIMP. This algorithm also proved to be useful in producing the background star field, in which one seeks to eliminate the comet itself.

I selected the ST80 for this image due to its wide field of view (roughly 2.5 times that of my 8" Newtonian), figuring that it would allow me to catch as much of the comet's tail as possible. This choice turned out to be problematic for a couple of reasons. First, the ST80, being an achromat, is incapable of focusing all colors of light at the same point in space, causing those goofy-looking blue/purple-ish halos around the bright stars. Second, those very same halos make eliminating the star-blobs from the stacked images very difficult. Pinpoint stars formed by better optics can be eliminated much more easily through the "Smallest N of M" algorithm, or several other algorithms. Should have used the ED80. You live, you learn.