Hi Mike,
OK, thanks for the report - I did change the behaviour there in between 12529 and 12551 to try to fix another issue, so it sounds like that 'fix' has broken your usage
The behaviour of how plate solving should work during live stacking turns out to be quite complex and depend on what you plan to do with the plate solving info - if you want to annotate the image using deep sky annotation then you should plate solve the stacked image (not the frames from the camera). If you want to recenter the telescope then you should plate solve a frame from the camera, not the stacked image - the two may be offset due to any drift during stacking.
Added to that, if you want to plate solve a frame from the camera while live stacking, most of the time, it's good enough just to take the last frame added to the stack (which is always available) rather than waiting for a new frame. However, if the stack is paused for some reason then the last frame stored by the stack is the last one before the pause, not the last one from the camera... That's what I was trying to deal with last week which forces the plate solving to go to the camera for another frame in some cases where it didn't before.
As a workaround, try launching annotation first, then using the buttons in the annotation window to launch the plate solve operation - they flag the operation as being targeted at annotating the image and therefore that should bypass the need to get a frame from the camera.
I would like to fix the issue that you are seeing, but am not quite sure how to trigger it. Please can you give me a step-by-step to trigger the problem (including whether the camera is in still or live mode, how you 'finish' live stacking, etc).
cheers,
Robin
