Upload & Sell: On
Coma, on the edges of astro images, is fairly common. Many scope manufacturers sell dedicated field flatteners designed for removing/reducing the coma. This problem is usually compounded with larger sensors. Andrew's 500, would probably be near perfect with a crop camera. This could be an ideal set-up, since you can use a full frame for nebula season (where you need a wider FOV) and a crop sensor (during galaxy season, when you need a little more reach).
Star images are really taxing on equipment, as every bit of field curvature or misalignment is evident. I suppose I'm a purist and want round stars to the corner of the frame, and have gone through several set-up's to achieve this. I also use a full frame sensor, but it's a dedicated astro CCD. A full frame sensor properly set-up and matched to a Takahashi FSQ-106ED will produce pinpoint stars all across the frame. Here's a star field that I did a few years ago.
The FSQ-106 is similar to the Canon 500, with a focal length of 530mm. At f5, it's a fairly fast scope. It also is a phenomenal scope/lens for terrestrial shots. Coupled to a DSLR it works well for bird and landscape shots (however not autofocus). Although similar in weight, it will end-up weighing a little more with mounting rings and mounting plate (it has no tripod mount).
There are a couple advantages of using a telescope with a dedicated sensor over a lens and DSLR. First, it's far easier to incorporate filters in a dedicated telescope system (and for the best results you'll want to use filters). They are readily available in the proper size and have housings (called filter wheels) that typically hold 5, 7 or 9 filters. Next is shooting with a dedicated CCD. These are available as color sensors (with installed Bayer matrix), but most prefer the use of a monochrome camera and filters. These dedicated CCD's have cooling to reduce noise, and large pixels to aid in efficiency. The big advantage, over a color sensor, is that each filtered image is full resolution. A typical color CMOS sensor uses a Bayer Matrix of RGGB. So in a 20MP sensor, ~5,000 pixels are red, ~5,000 are blue and ~10,000 are green. With a 10K Monochrome CCD, you shoot through filters and each image is 10,000 pixels. The difference between a combined monochrome CCD and a color CMOS is quite amazing. Finally, the cooling of dedicated CCD's helps to reduce the noise inherent to these types of sensors. CCD's tend to have less noise (vs CMOS), at the expense of requiring a lot more power. When cooled, the noise is substantially reduced. The noise level is cut in half for every ~5 degrees of temperature drop. I typically shoot with my sensor at -30C, which reduces the noise to a very low level.
The cost the FSQ-106ED itself, is less than than a EF500 lens but will produce spectacularly fine and round stars on a full-frame camera, provided the camera is mounted precisely perpendicular and on-axis. The down-side is that once you have it, you'll want the dedicated CCD, filters and filter wheel, EQ mount and other bits and pieces to get it all working together. Then you're about 2.5 to 3X the cost of the EF 500 I think I've posted it before, but here's what a typical short focal length (wide field) CCD imaging system look like.
I just thought I'd share these details, in case anyone was thinking of making the plunge into dedicated astro.
PS love this thread...