The Pixel Sweet Spot

Earlier this week the forecast was for an evening of average to perhaps better than average seeing with cold (but not biting) temps. I rolled out the scope a little before sunset to begin cooling and got things ready - with Jupiter just past quadrature it is always going to be highest in the sky as soon as it becomes visible. 

I did the routine alignment, collimation check, and finder alignment before finally popping in the ZWO camera. Activating the camera I was greeted by a strange sight - an emerald green Jupiter. At first I thought maybe a Debayer setting was off in the capture interface but soon noticed that the histogram was not registering in blue or red, only green. I rebooted the laptop hoping maybe that would restore things, but no luck. I brought up a different capture application, but it, too, sported a green globe.

Rather than admit defeat I located my retired Imaging Source camera and popped it into the Barlow. The view and histogram confirmed that we were back to getting a color image, but I immediately was struck by how much smaller the image appeared to be. Hmm - what was that about?

It turns out that my older camera, a DFK21AU042, has a pixel size of 5.6µ whereas my ASI178MC has a size of less than half that, checking in at 2.4µ. The formula for calculating how much sky each pixel registers for your setup is as follows:

  (Pixel Size/Telescope Focal Length) * 206.265  

For my setup using a 2.5x Barlow that becomes:

DFK21AU042 = 0.31"

ASI178MC = 0.13"

The theoretical ideal for planetary imaging for under average seeing conditions is around 0.15" per pixel (Note that this is different than DSO imaging, where the average is about 1-2" per pixel). Clearly, my ZWO camera is a lot closer to the mark, and the better thing to have done would have been to stop and swap out my 2.5x Barlow for my 4x one to get a little lower arc-second/pixel value. But the window of calm seeing that we often get shortly after sundown wouldn't allow that, so I forged ahead.

Below are comparative images taken about a week apart of roughly the same Jovian longitude. It is pretty obvious from it that we lose resolution in the image acquired using the DFK21AU042 camera.

Is the image from the older camera terrible? No, hardly. We can still make out details like Oval BA and anti-cyclone storm A1 - something that was unheard of using film a few decades ago. But in astronomy, and in planetary imaging in particular, it is all about getting all the parameters as ideal as possible so that you can capture all the details available given the seeing conditions. Hopefully I get my ZWO camera fixed, but in the meantime I know from experience now to at least break out the 4x Barlow to try to get closer to that desired arc-second/pixel value.