So, I finally got some time and went back to the Susanna occultation. AstroArt did a not bad job but I knocked up some quick code to have a look at what I could get out of it.

The image shows the result. The final accuracy is down to how wide you measure the PSF as  plus how accurately you can estimate the brightness of the uneclipsed star near the eclipse start. I'm not entirely happy with how I have automated that, but it will do until I get some more time. Its a start at least.

In the plot the blue dots are the background subtracted values for the star trail as it moves across the frame.

The red vertical lines are where I estimate the star trails started and ended (don't forget the PSF).

The yellow vertical lines show the start and end of the eclipse (not entirely happy with how this is currently being determined. Need to revisit.)

The cyan horizontal line is the threshold showing at what flux value it was decided that the eclipse was starting/ending. 

For the record it gives timings of:

Disappearance:00:10:06.28  

Reappearance:00:10:10.15

Duration:        3.93sec  +-0.053sec

The  duration agrees roughly with Scanayzer, but the individual timings do not. Its not as easy as it looks... More thought needed.