Thanks for getting in touch - the way it is supposed to work is like this
* SharpCap knows where the pole is (that's easy - the RA/Dec for a date come straight out of lookup tables)
* SharpCap knows the point that the sky appears to rotate around - that's where the RA axis is pointing
* It can also transform any point in the image to RA/Dec co-ordinates (because it has plate solved the image)
so... SharpCap can work out the RA/Dec that the mount axis is pointing at. Once it has that it can work out the Alt/Az of both the pole and the center of rotation point. Knowing both the the points in Alt/Az co-ordinates should make it easy to work out the up/down/left/right adjustments needed, but it isn't working out that way it seems
The directions are supposed to be interpreted in terms of movements to be made if you are standing on the equator side of the mount looking towards the pole. They aren't supposed to be in terms of the movements on the screen.
Just looking through the code it seems possible that I have left/right switched round, but I will have to wait for a clear night to check. Up/down seem to be correct.
The other problem is that SharpCap needs to know your latitude and longitude to be able to convert RA/Dec into Alt/Az. It estimates the latitude as 45 and the longitude from your time zome offset from GMT. This isn't precise but it should be good enough to get the directions generally OK.
If anyone does get a clear night sooner than I do and can confirm or deny that left/right are inverted (and whether up/down are the correct way round) it would be very helpful.
cheers,
Robin