I'm Cheap, but the Hα Filter Was Worth the Money

As I mentioned in my last post, I finally succumbed and laid out $160 for a Baader Planetarium 7 nm Hα filter. Last night was my first time out using it. Conditions were marginal at best; it's monsoon season so there were clouds and wind gusts as there were thunderstorms around. Synced on λ Sag and easily found M8 with the PicGoto; use of the PicGoto is becoming routine. Focusing with the Hα filter was interesting; the stars are so small that it is hard to use the half flux radius as a focusing metric, so mostly I just eyeballed it. M8 did turn out to be a good target for this first try. Because it is so bright, 300 s subframes were sufficient (and actually just about right) for the Hα. While waiting for clouds to clear, I did take a 300 s shot of the Crescent Nebula through the filter, and that was not enough for this dimmer object. Clouds rolled in after only about an hour, so this session was cut off prematurely (though I did get a few more subs through "cheater holes" in the clouds), but even with these problems I captured a lot more contrast in M8 than I got unfiltered. Here's the unfiltered version for comparison:

Date: 30 Jul 2014
Subject: M8, Lagoon Nebula
Scope: AT8IN+Antares 0.5x focal reducer
Filter: Meade IR cut
Mount: CG-5 (Synta motors, PicGoto Simplificado)
Guiding: 9x50 Finder/Guider + DSI Ic + PHD 2.3.0 (Win 7 ASCOM)
Camera: DSI IIc chiller at 2.50 A, 15 °C
Acquisition: Nebulosity 3.2.0, no dither
Exposure: 55x120 s
Stacking: Neb 3, bad pixel map, bias included, normalize first, trans+rot align, 1.5 SD stack.
Processing: StarTools 1.3.5.279 Crop; Wipe:Color & brightness 81%; Develop 80.11%; HDR:Core reveal; save; undo; HDR:Optimize; Color:Scientific, 250%; Deconvolute auto mask 1.5 pix; Life:Moderate; Track RNC 1.47%; Magic:Shrink 1 pix; to the Core Reveal image: Color:Scientific, 250%; Deconvolute auto mask 2.5 pix; Life:Moderate; Track RNC 1.72%; Magic:Shrink 1 pix; CS6 Astronomy Tools layer the HDR:Optimize image on top of the HDR:Core Reveal with 85% opacity; Increase star color; Astro Frame

Processing Hα data involves a lot of new tricks. You begin by doing normal preprocessing of the images with darks (a bad pixel map in my case), flats (taken through the Hα filter), and bias frames. Next, you have to extract only the red channel from the CCD because the G and B channels are mostly just noise; Nebulosity has a batch tool for doing that with a whole set of files. Next, I normalized the intensities. Next, I needed to square the pixels (since my camera has rectangular pixels, and the RGB image I want to combine with has already been squared when it was de-Bayered); again Nebulosity has a batch tool for dealing with that. Then I stacked the Hα images as I normally would: I aligned them using two stars, saving each aligned image, then stacked the aligned images with 1.5-σ clipping. A quick visit to Photoshop let me double the number of pixels so I'd be at the same scale as my RGB image when I'm ready to combine. I next took the stacked Hα into StarTools and processed much as I would an RGB image, except I skipped the "wipe" step since I didn't need to do any color fixing. Finally, back to Photoshop where all I did was false color the image and put a frame on it. Here's the result. The increased contrast in the nebula, even on a first attempt under poor conditions, is amazing. The Hα image makes the nebula look almost 3-dimensional (which of course it really IS):

Date: 2 Aug 2014
Subject: M8, Lagoon Nebula
Scope: AT8IN+Antares 0.5x focal reducer
Filter: Baader H-α 7 nm
Mount: CG-5 (Synta motors, PicGoto Simplificado)
Guiding: 9x50 Finder/Guider + DSI Ic + PHD 2.3.0 (Win 7 ASCOM)
Camera: DSI IIc chiller at 2.50 A, 20 °C
Acquisition: Nebulosity 3.2.0, no dither
Exposure: 24x300 s
Stacking: Neb 3, bad pixel map, bias included, extract R, normalize, square B&W, trans+rot align, 1.5 SD stack.
Processing: StarTools 1.3.5.279 Crop; Auto Develop; HDR:Optimize; Deconvolute 2.5 pix; Life:Moderate; Track RNC 0.74%. CS6 Astronomy Tools double the pixels; H-α false color; Astro Frame.

The next task was to combine the Hα data with the RGB data. There's a nice tutorial on Starizona's web site about how to do this in Photoshop (google it). They describe two methods. The first preserves more of the stars from the RGB image, while the second emphasizes the detail in the nebula. I show both results below. You can judge for yourself, but for this image I like the "Method 1" HαRRGB better than the "Method 2" HαRGB combination. I think the rich star field adds to the image and Method 2 pretty much wipes that out, and looks a bit overprocessed to me. Of course, this is all in the hands of a first-timer, so maybe I'll get better with practice. Here are the two images. Both show a lot more nebula than the RGB alone did. So even though the Hα was a big investment for me, I think it is worth the money. I'm looking forward to using it some more.