For the past month, I have kept adjusting my setup to try to get the best star shapes that I could.  For the longest time, I thought my problem with eccentricity was coming from poor tracking and excessive field curvature from the EdgeHD 0.7x focal reducer.  In a "Eureka moment," I noticed that the center stars were also elongated even with perfect tracking.  Going back to basics, I looked at my collimation in Metaguide.  Everything appeared spot-on.  Then, I noticed my camera's fan was off during collimation.  As soon as I turned it on, I watched my perfectly round and collimated stars turn into a warped and elongated mess.  This is when I realized fan vibration caused my problems.

This happened about halfway through the month.  I went through several different fans, added silicone posts to the camera, and added gaskets to the mounting screws for the back of the camera housing.  As an added bonus, I learned how to solder thanks to my best friend, an electrician down the street, and learned how to make a western union to get a 2 pin connector onto a fan meant for 3 pins.  End-result: after turning Metaguide on again, stars remained perfect.

Having conquered fan vibration, I had renewed interest in attempting to use the focal reducer.  I first focused the scope as best I could at f/10 using a rear cell Crayford focuser.  Then, I used that length when attaching the focal reducer.  Keeping that length constant, I then adjusted the primary mirror to achieve focus with the focal reducer.  Lastly, I fine-tuned focus with the Crayford focuser.  The result was much less field curvature and virtual elimination of tilt, which in hindsight I believe was artificial due to the fan vibration distorting the stars.

Through this process, I also tweaked PhD2 to work with my 10 Micron mount.  In order to improve guiding, I tweaked my sky model down to 6" RMS.   I lowered the DEC and RA aggression to 0.4, and set pulse guiding to every 4 seconds.  These things yielded major improvements, and often times got things down to 0.38" RMS.

The last part of this month's experiment was to try combining data shot at f/10 and f/7.  This, by far, was the easiest part of the month.  Astro Pixel Processor made that task ridiculously simple.  Using APP 2.0 Beta 6, I unchecked "same camera / optics" in registration, enabled dynamic distortion correction, and bumped multi-band blending to 10%.  The program is modern wizardry and just works.

So, the end result of all this is my longest project yet, clocked at just over 40 hours, of NGC 1491, the Fossil Footprint Nebula.