Good image acquisition is the secret to great astro images. It's as simple as that - if you don't have good data to start with, no amount of processing will help. What do I mean by good data? It's pretty simple:
Simple, yes, but getting there is another story. Read on for the details of my still-evolving process.
I used to PHD Guiding from Craig Stark (free from Stark Labs, donation highly recommended) for guiding. It's simple, Mac and PC, and just works. However, now that I am using ACP and Maxim for automation, I use the Maxim DL guiding routines. My main camera is a QSI583wsg, so I have the luxury of using off-axis guiding. The 583wsg has a guide star pick-off prism in front of the filter wheel, so there's never a problem funding a good guide star. The best Maxim settings are highly dependent on your specific setup - focal length, mount, camera, etc, but I think one of the most important settings is the exposure duration. Too short, and you will be chasing the seeing - I use a minimum of 2 seconds, and sometimes longer. I'm using a Lodestar guide camera which is fairly sensitive and can pick up faint magnitude guide stars if needed. However, for best results, as high as possible SNR for guide stars is always best.
With the Planewave CDK 12.5 and the Lodestar guide camera, guiding is a bit more of a challenge, given the focal length of 2541mm. By tuning the Maxim parameters, I've gotten down to an RMS of about .35 pixels on the Lodestar, which equals approximately 0.25 arc seconds. I've recently started binning the Lodestar 2x2 on the CDK, because given the very small FOV, I'm having to choose fairly faint guide stars at times.
At the 2010 OMSI Astrophotography Conference in Portland, OR, Tom Carrico (noted for his high quality data) had some additional great guiding tips:
Finally, if you have great mount performance, great polar alignment, etc, etc, do you even need to guide? Some people claim that they are able to take 30 minute unguided images with great results - I'm not one of those. Of course, it depends on the mount and the telescope and imaging scale, but even then there are effects of differential refraction as the mount tracks the objects, plus other random effects. Unless you are really good, and maybe using a sky modelling program like TPoint, I say guide.
For Polar Alignment I use Pempro, from Ray Gralak. Great program, easy to use, and fast to check or adjust Polar Alignment. Also, it comes with the AP1100 mount. My mount and telescope arrangement has a bit of non-orthogonality, which means that slews on different sides of the meridian don't always exactly center the target on the camera chip. Remember, Polar Alignment aligns the mount, not the telescope! Rather than mechanically adjust out this non-orthogonality, I'm waiting for Astro-Phyics to release APCC. This new software package will function like TPoint, and allow the user to create a sky model that takes into account small remaining polar alignment and mount alignment errors. In the meantime, ACP and plate solving always re-orient the mount after a flip with no trouble.
Obtaining optimum mount performance is pretty simple - spend a lot of money and buy a good mount! Seriously, if you have an AP or Paramount or similar, you've solved most of your problems right away. At that point, it's a matter of maintaining reasonable balance, checking for minor backlash adjustments every year or so, and keeping the mount properly polar aligned. My periodic error on the AP1100 is very low, but I leave PE correction turned on in the controller. As discussed earlier in the guiding section, the fewer guide corrections the better, so removing small Periodic Errors helps. The 1100 also has RA and Dec encoders, and I'm still experimenting with respect to their effect on tracking and guiding.