NavList:

Thanks to every one for the really helpful info and ideas. My daughter is a profession photographer and is home at the weekend so I thought I might enlist her help as she has one of the top Canon professional 35mm DSLRs (I cant remember which model) which has a full size sensor and a 50mm prime lens. My plan was to try this on a standard tripod at home.

Peter's suggestion of a long and short exposure is an interesting one I had not thought of that. I was assuming I would take several in quick succession on a setting that is a compramise for the moon and the star.

Then the fun starts to do the image processing. I had not thought much tabout the moon's features. I was asuming segmentation and least squares fit of a circle would be a good first stab.

What I was thinking in the long term is that as DSLR sensors improve in resoluution and low light capability at some point it might be possible to make an partially automated CN for use at sea as a fall back when all the satelites crash in to a crowd of space junk and all the GNNS systems fail. I wanted to see how close we are to this now with top of the range DSLRs. Of course luna distance for time correction makes me seem like a seriously crazy "disaster preper" as there are planty of HF time signals. Partly I wanted to have a try at this as I could do it more easily from home and it seems easier that an artificial horizon with a camera as the angle is smaller. It may be that the sensitivity of the sensors is more important than the pixel resolution so shorter exposures can be used on a moving deck. I I was thinking that "software" image staibilization might be possibe rather than a giroscopic mount.