I posted my guiding results from a 5 hour imaging session last night on Astrobin:
I wanted to show how an Absolute Encoder Mount can enhance the guide accuracy and thereby improve the image resolution when using medium to long focal length telescopes.
There are always discussions on the various newsgroups that question whether an encoder mount is really necessary if you're going to guide anyway. The idea doesn't make sense until you realize that the encoders provide the mount axes with a very precise position according to the input commands of the servo controller. It means that if you can detect sub-arc sec errors in the position of the guide star, you can correct for them at a very precise level. You can also take 90% of the guide burden away for really long exposures by modeling, and thus driving both axes at a custom rate according to the model.
The encoder loop is not only accurate but very linear down to the sub-arc sec level. For example, a non-encoder mount can have non-linear response and discontinuity in Dec during direction reversals, even if the gears are belt driven. That can cause delays and even oscillations around the zero point. All that goes away with encoder control.
Last night was not especially great seeing, Clear Sky Clock rated it as a 3 out of 5, but it was good enough for the nudge guiding plus model tracking to track down to the 0.2 arc sec Pk level. Without the model and just normal full guiding (2 - 3 sec guide exposure) the system would be guiding at around 0.5 arc sec Pk (which isn't shabby).