I switched the observatory camera from the original CCD SBIG ST-10XME to my modified CMOS Canon 500D/T1i. Replacing the observatory camera was more work, of course, than envisioned. The Canon sensor is set back compared to the SBIG’s and there was not enough range for the TEC-140 telescope to get focused. Jeff Dickerman, the super helpful president at Optec, made me a new adapter to go with another receiver that shortened the camera train enough to get easily into focus. I then rediscovered (I had forgotten but found my own online forum entry!) that CCDAutoPIlot had to turn off the feature on TheSkyX that would record both RAW and FITS files and records only the FITS files. It turns out the DeepSkyStacker can work with FITS and convert the RAW files that create them in color. But I had over time come up with a dark library in RAW frames for the Canon so I could temperature match with the light frames. I don’t have such with FITS files. But if I want to use the observatory automation of CCDAutoPilot, which I do, I will need a new FITS library at a range of temperatures to do it right.

Then I had trouble with setting exposure with CCDAutoPilot on the Canon images. I’m telling a lot of technical info here, but it took a lot of wrestling directly with complex technology to get through all the unforeseen obstacles. The Canon CMOS sensor uses a 2X2 Bayer matrix to record color. Two cells of the matrix are green, the other two are red and blue each. CCD’s, like my SBIG camera,have individual sensors not organized in a matrix. In something to do with the CMOS Bayer matrix, CCDAutoPilot only sees one cell as exposed, and it was saturated, while the other three were only about one-fifth exposed. It averages them in a meaningless way and it seems hit or miss if CCDAutoPilot can get the exposure close enough to get flats. Sometimes it gives up, sometime it settles on something. Focusing is tough too, as is plate solving. CCDAutoPilot did not have a solution as you can see on their forum here, but John Smith, the creator of the software, encouraged me to investigate on my own further. You are getting in pretty deep when the software designer basically gives up and says good luck.

Weirdly, after all that, I did come up with the image below.

Click on the image for technical detail.

Along the way I also worked with Software Bisque and a technical wizard named Rodolphe (rhymes with Gandalf) Pineau to come up with a new interface with the driver for the Optec filter wheel that avoids the timeout issue I was having. See my discussion about it on the help forum here. Rodolphe worked for free ( I sent him some token beer money) and Thomas Bisque at Software Bisque and I were dutifully delighted.

To keep things confusing, let me tell you I also tried to get an image of an open cluster, using the usual native RAW files from the Canon, via BackyardEOS instead of the FITS files you end up with using CCDAutoPilot. I have used BackyardEOS to good success numerous times, but this image would not come out with any star color no matter how I tried. I can’t explain it. Same camera, same telescope and image train.

I ended up giving up on using the Canon with CCDAutoPilot and put the SBIG back on the telescope. It takes some work. Then I realized I had in fact acquired all this data on the image above while I was experimenting. I processed it to see what I could come up with, and shucks, it looks okay. Now I am thinking about putting the Canon back on. I have some good flats at position angle 90 degrees but would need some for my more common usage of 0 degrees. That seems hit and miss but is probably possible. I may have some already. Bias frames are easy and dark frames don’t seem to be making trouble at the wrong temperatures.

Hmmm. Round and round with technology wrestling. Hard to account for this hobby.