The Iris and the Ghost
I had a week off of work around my birthday, and with having just gotten my new Celestron RASA 8” telescope & ZWO ASI2600MC Camera, it was a perfect time to go to my friend’s farm and break things in!
The forecast was for a week of clear skies, and having set up at home and taken a test shot of the Elephant’s trunk without major issue, I was hoping to have a similar experience at the farm. If only I were so lucky!
The first night of imaging I took some time trying to get the spacing just right with my optics - I’d noticed that things were a little off with my test shot, so I worked on adjustments using a metal shim & clear window combination in addition to a UV/IR Filter. The entire first night was dealing with spacing, as well as some software issues with things not wanting to connect. I got a few test shots to diagnose the spacing, but that was about it before clouds rolled in.
I thought I’d improved the spacing, and with my 3d printed cable guide, it looked like I had good shots with small diffraction spikes on the stars. I spent the first night out trying to get the spacing just right before clouds rolled in and filled the sky. Diagnostics were showing I never quite had a good flat field.
Much of the second night was spent working on that spacing, having installed pixinsight onto my imaging laptop so I could better analyze it. I took my time the second night to get the spacing just right, and around midnight began gathering images, collecting a total of 260 x 30s shots of the Iris.
However, the next day, upon review I found that the diffraction spikes seemed to shift as they got closer to the bottom of the image. That’s a clear sign of trouble with the imaging train, so I set out to find ways to solve those issues.
To try to make improvements I designed and printed a new cable guide to eliminate the spikes altogether, and a centering ring for my camera adapter. For the third night I removed the UV/IR filter, and with the new 3d printed parts my stars looked good throughout the image. But then another issue popped up - my plate solve software, that is used to help find and center perfectly on my targets, would not recognize the sky at all around my target. It would get me close, then fail. I did not manage to get it centering until after 2am. I gathered only an hour of shots (140 x 30s)that night, on what should have been one of the best nights for imaging.
I went back out the next night, having installed a new plate solving software, and waited out lots of clouds. While waiting, I snapped a shot of the moon. The clouds finally moved out around 11:40, and I began imaging through the last of them. I was able to find my target quickly this time, and let my system gather images while I went in and slept. I checked on things around 12:30 and forced a refocus, the clouds having mostly cleared, and then slept more. I packed up around 3:30, having gathered a total of 380 x 30s shots.
It would turn out that was the last clear night of the week.
I began working on processing my images, and found that I had a strange brownish stripe on the right side of all of my processed shots. Digging in I found that the blue channel on all of my flat frames contains a bright bar along the right side. It’s in the same location regardless of my camera’s rotation. I tested the camera on my older scope, and the bar stayed in the same spot, and only in the blue channel. I realized I could diagnose this by putting my old filters in front of the camera (RGB filters). Effectively “double filtering”, I could use subtraction to figure out what was going on.
Blue Channel + Green Filter = Light bar on the right.
Blue Channel + Red Filter = Brighter bar on the right.
Blue Channel + Blue Filter = No bar on the right.
Effectively this tells me that the blue part of the CFA array on my camera, in a stripe 200 pixels wide, is not actually filtering out red & green light. It’s a CFA problem, and I’ll need to get the camera repaired or replaced.
Facing the reality of that, and the forecast with no more clear nights, I decided to dump ALL of the shots I got into pixinsight. The bad diffraction spikes, the light clouds, all of it. I just dumped it in, used SubFrameSelector to get rid of only the absolute worst ones, and stacked them. The result was surprisingly good. Small diffraction spikes, the shifting hardly noticeable. Good clear dust clouds. And a big brown bar on the right side. I cropped the image quite heavily, and then ran through the workflow (mostly following Adam Block’s guide for OSC processing, including his blend method).
The resulting image is quite good, better than it has a right to be given all the trouble the week gave me. I posted to Astrobin and shared it online, and it’s gotten very positive feedback. I made a few tweaks based on some suggestions, and honestly, at this point it’s probably the best image I’ve taken. Which really speaks to the potential of the RASA and 2600MC.
To see the full sized image plus details, visit my astrobin link.
