I thought it would be interesting to re-process one of my Plato videos and look at the results of each processing step.
Step 0: before I start I already have lots of images - each raw frame from the video. These images vary in quality depending on the viewing conditions at that moment but here is a typical frame from the video I'm about to process:

Step 1: the bulk of the Registax processing involves aligning the individual frames and selecting the best ones to stack together to create a new picture. I've used 400 frames out of the 3200 available. This creates an image that is less noisy (i.e. smoother) but also quite blurry:

Step 2: the step that makes all the difference is tweaking the image properties to bring out the detail. This is something of a black art and can range from simple changes to the contrast/brightness to sharpening individual image layers or manipulation in PhotoShop. Increasing the brightness and contrast gives us this:

Then sharpening some of the image layers:

And finally changing the saturation and lightness:

So this is the final image using a new processing technique! Here's the image using the old technique:

I think there's a clear improvement (especially if you click on each image to enlarge them) with the same detail but much less noise.