It’s taken several months to get this image. Long, painful months, but I’m glad I persevered.
It all started after NEAF in April. NEAF is the Northeast Astronomy Forum, the largest astronomy related trade show in the US. I bought a Stellarvue SV80-3SV there and I was looking forward to using it for imaging.
Alas, the clouds that have dominated 2018 continued unabated.
Eventually, I got a few clear nights scattered over a few weeks and ran into problems with my Losmandy GM811 mount. At first I thought the problems were user error, I couldn’t get the PHD2 guiding program to calibrate in declination properly. It kept complaining about backlash. After much trial and error and without any success I reached out to someone in the Northern Virginia Astronomy Club for help and he helped me set the worm mesh correctly. I think some of my attempts may have been correct but as this was new to me it was hard to be certain. This time I was sure it was correct as he had years of experience with Losmandy mounts.
On the next clear night I tried again only to have the same problem. I ultimately let it run using the planetary nebula M57 as my target. M57 is much too small a target for my focal length but I was really just trying to see if I could get anything at all and not really concerned about getting a decent image. I ultimately let it image without guiding because the guiding was making things worse. Here is what I got:
This wasn’t bad but I couldn’t get that when guiding. My stars turned into eggs.
I spent the next few, rare clear nights trying to troubleshoot this and then my mount electronics developed an odd problem and Losmandy asked me to send it in so they could look at it. I also sent the the declination axis for the mount since it was still failing to calibrate. Two weeks later I had them back and the drive electronics were working but I was still seeing issues with the PHD calibration.
I continued to try to troubleshoot this, picking up a copy of PemPro to characterize my mount’s periodic error and generate a PEC (periodic error correction) curve for the mount. To my surprise I was seeing large periodic error in RA and about every two minutes sharp spikes in RA that lined up perfectly. That told me that the RA errors I was seeing were in the mount, at least in part.
I was beginning to think my mount might be a lemon but I wasn’t 100% convinced that it wasn’t all the mount’s fault. Specifically I was worried about cable management, particularly for the DSLR since the USB cable was so short and the power cable fairly heavy. In a last ditch effort to see if that might be it, I bought a ZWO ASI294MC Pro one-shot color cooled astronomy camera. I had ultimately intended to get a mono camera and filter wheel but my budget wasn’t ready for that and it also added complexity on the mount and right now I needed to reduce complexity. This required me to give up some field of view since the ASI294’s sensor is a 4/3 format chip compared to my full frame DSLR but it had slightly smaller pixels and more importantly set point cooling.
Amazingly, the night after the camera arrived was clear and I set up in the back yard to see what happened. I first ran PemPro again and while the periodic error was large, around 17 arcseconds, peak-to-peak, it was much smoother than it had been. I let PemPro collect date for 7 work cycles and then generate a correction curve and uploaded it to the mount. Checking the results, periodic error was reduced though still highish at around 6 arc seconds peak-to-peak. I’m definitely not a PemPro expert but it clearly seemed to help. The periodic error with the PEC curve applied was fairly smooth and the only question is would PHD be able to cope with this.
And the answer was that it could. My declination calibration problem vanished and my RA guiding error while on the high side was under control at my relatively coarse pixel scale (1.96 arc seconds per pixel). I tried exposures ranging from 1 second up to 3 minutes and the stars came out round on all of them.
I decided to image M27, another planetary nebula, and set planned to let it run all night to see what I would get. I took 60 second exposures at gain 250. I set it up to run and then went to bed. At about 2:30AM I woke up to check on it via remote desktop and saw that clouds (that weren’t in the forecast!) were moving through and had caused PHD to lose the guide star. But, I had taken 80 images at that point. I went back to sleep and around 3:30AM woke up and checked again. Clouds were still moving through. I watched it for about 30 minutes and decided to go outside and look at the sky directly. The clouds had covered most of the sky and seemed unlikely to go away any time soon so I decided to tear down. I took my flat frames and then shut everything down.
I quickly inspected the individual frames and to my delight, all were usable! Not one had to be tossed because of egg shaped stars. Amazingly I didn’t even have any airplanes go through the frames.
The image at the top was the result. Even that took a few iterations of processing to get to that point. In the first attempt I mistook a gradient for real nebulosity and didn’t remove it and then ended up making a real mess of the image. The second attempt was better but the stars were bright and hard and overwhelmed the image. The third attempt finally brought them under control. Image processing for astronomy is definitely harder than normal image processing.
There are still some things to work out:
Is the periodic error I’m seeing normal for a GM811?
Looking at the night’s guiding in the PHD Log Viewer it appears my RMS error was lower than what I was seeing before I went to bed. I was using PHD’s PPEC algorithm so perhaps it learned the error and was able to correct it. In that case I probably don’t need to worry but it could be that I’m not interpreting the data in the Log Viewer correctly. This bears more investigation.
My declination calibration actually seemed to overcorrect (a huge change from the lack of movement I had been seeing). It’s possible I need to still tune my dec axis a bit but overall the declination guiding was pretty good so I’m not inclined to touch it unless I have to. The goal isn’t to chase a guiding number but to get a good image.
That completes the saga of the last three to four months. I think I’m finally operational and I’m looking forward to the Almost Heaven Star Party in less than a week were I hope to spend four nights imaging under dark skies. Of course, the forecast is not good…let’s hope the weather gods cooperate!