Manual, one star

[Serg1]

It works.

[han.k]

It will not work if you have more then one image for one colour. This is because colour stacking is done in two steps. First it combines the blue, red, green (and luminescence) Then it combines the results. The manual position is lost in the second step. You could fix it manually by stacking the interim results. The problem only occurs for separated colours. Not for OSC (DSLR images)

I will fix it tomorrow. Do you have more then the three images for testing?

[Serg1]

Hello Han.
If you add more than 1 frame of each color, the comet does not add up correctly.
The sample files look in a personal message.

[han.k]

It is fixed in version 0.9.159 just uploaded.

46P_Wirtanen_Blue_60s_20190107_152847@ 0xFD 0xF 0xD 4x60R 4x60G 4x60B 0x0RGB 0x0L _stacked.jpg (86.15 KiB) Viewed 3186 times

[han.k]

In version 0.9.160, I improved the handling of selecting objects in manual alignment mode. So the object selection stays at the same place when you swipe/zoom.

[Serg1]

Hello Han.
The comet is not always highlighted with a red square. In version 0.9.159 it was better.

[han.k]

If it fails to lock then try again pointing the mouse accurately. It should only happen with very faint "comets". If I go through all your images it works, only for one or two images you have to retry.

I have modified the code (0.9.160) such that only when it locks, the square is shown. This guarantees that the comet is stacked accurately. In version 0.9.159 if locking fails it takes the mouse position which could be a few pixels wrong. If it locks, it will take the "center of gravity"

Does this work for you?

[Serg1]

Stands out, but not with first times. Can make one mouse click-center of gravity selection, double click-mouse position selection.

[han.k]

I will change back to the 0.9.159 behavior, but show an extra cross if locked.

The lock will not work with oval comet heads. This is because the lock is designed to avoid double stars.

[han.k]

Locking behavior changed back in version 0.9.161. If locking is unreliable, the shape is a square with a hint "unreliable lock". If a good lock is found with a high SNR, the shape is a circle with no hint.

It locks on an oval or stretched object showing a square. It just tries to find the "center of gravity".

Han