I'm going to try to cover off all the points above in one post - hopefully I won't miss anything...
@PeterC65 - 16 bit is very unusual for high speed captures, as @Borodog says, because you are using high gain, all you are doing is recording the noise at higher fidelity, there is pretty much no useful signal in the extra 8 bits of data. The higher frame rates that you can get at 8 bit are almost always preferable to the 16 bit option. The one exception is narrowband solar imaging where you want to capture the (faint) prominences as well as the bright disc in the same capture.
That being said, I hadn't realised until I checked that there was an issue with 16 bit. In fact the stacking was all happening as expected in the background, but the copy of the image back to the display wasn't happening because of a check to see that the display data required was the same size and format as the data from the stack. I have a fix that will at least make it work, but it's not perfect as the colour adjust/saturation is still happening in 8 bit (the sharpening is 16 bit) - I will try to fix it all properly later.
@JohnD - glad you like it
@zerolatitude - looking good
@Borodog - good call on the display stretch. I have the ringing effect at the edge noted down as something to look into later on to see if I can find a way to minimize/prevent it in the sharpening process.
I've also had a separate report that there is still potentially an issue with ROI being changed causing a crash - I will be looking into that.
cheers,
Robin