×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Autofocus Star selection with a sub frame not centring the star

  • Posts: 437
  • Thank you received: 31

I understand that full frame is an option but as it takes around 30-45 seconds to download a Frame it is a slow process to use it.

My preference is to find out what is going on with the subframe, which works well and quickly so long as I manually select the star. Obviously this impacts on my efforts to automate things.

Paul
3 years 5 months ago #61220

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31

I could not completely duplicate this last night but have some interesting results.

First image is a first attempt and worked. Notice that the selected area is square.
The next image shows a later attempt which also also worked but notice that the selected area has become a rectangle.
The next image is another attempt that shows a selection problem. The area is again a square and it has chosen an area around a valid star but has selected what I recall was actuall two stars away from the centre. It was actually able to focus using them, but is not ideal the central star was probably the best to focus on and must have been what was used to define the square.
The fourth image clearly shows two stars selected which are not the centred star.
The final three images show some of the settings.

As can be seen the star used to create the sub frame has not moved, it is just the star(s) selected to do the focus are not the same as used to create the subframe.
Last edit: 3 years 5 months ago by Paul.
3 years 5 months ago #61336

Please Log in or Create an account to join the conversation.

  • Posts: 76
  • Thank you received: 5
I wonder if this is the same issue I’m running into? I’m using the full frame and multi star. But it sounds similar.

indilib.org/forum/general/7947-sep-autof...-star-detection.html
3 years 5 months ago #61360

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31
It is possible.
First of all it appears to be prone to selecting multiple stars or globulars to focus on and I have also found that it sometimes selects an area devoid of stars.
In my experience it appears the initial selection that is used to create the subframe is too often fooled by small globulars and multiple stars but they can still be used to focus, even though it takes longer.
The major part of my issue, and one that does not happen all the time, is the star that is initially selected is not being used to perform the focus, it seems to decide to use another even though the original star is centred in the subframe. Another hint is that on occassion the square is a rectangle, which leads me to believe a variable is being corrupted.
This seems to have been introduced sometime in the last 6-12 months.

Paul
The following user(s) said Thank You: Paul Muller
Last edit: 3 years 5 months ago by Paul.
3 years 5 months ago #61384

Please Log in or Create an account to join the conversation.

  • Posts: 319
  • Thank you received: 25
IMO this could be one of three possibilities (as it happened to me before), (1) you are not physically tracking, i mean you set tracking to ON and see position change, but the mount is not moving as it was ordered, or (2) you are don't have good polar alignment and the tracking is way off course, or (3) tracking for stars is not sidereal.

Technically, if the polar alignment is accurate, and the mount is tracking good (either on its own or controlled by ekos), you should see your target standstill, regardless of what you are doing with your camera. Guiding is for minor correction for astrophotography, but here you are just checking for fucus for 2-sec shots. You should have the stars stationary for a some time.
3 years 5 months ago #61418

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31
This is not going to be a tracking issue.
The star used to select the subframe is exactly centred, it is just that it is not being used to do the focusing.
Polar alignment is accurate, and I have not adjusted it for nearly 12 months, I am able to do 20 minute exposures with no issues. It was drift aligned using PHD2.
It changes tracking speed if I choose the Sun or Moon but stars use siderial.
And finally, if I manually select the star it works fine - it is only a problem with autoselection.

Paul
The following user(s) said Thank You: Paul Muller
Last edit: 3 years 5 months ago by Paul. Reason: update correction
3 years 5 months ago #61430

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31
One further thing to higlight is that this is only an issue when using autoselect.
If I select it manually it works fine, so I cannot see that this has anything other than to do with the autoselect routines.

Paul
The following user(s) said Thank You: Paul Muller
Last edit: 3 years 5 months ago by Paul.
3 years 5 months ago #61442

Please Log in or Create an account to join the conversation.

  • Posts: 437
  • Thank you received: 31
After a period of cloud I have now been able to collect more evidence of what is happening.

Please note that whenever I manually select the star to use for focusing, it works correctly, it is only ever an issue when using autoselect.

There seems to be two main circumstances -
1) It selects a star and centres it but then uses another star in the subframe to focus on. This sometimes involves two stars close together where it jumps back and forward between them.
2) It selects a star but instead of being centred in the subframe it is near the edge. This sometimes work fine but sometimes slips outside the frame while focusing.

The four new examples are attached
1) Selected two stars close together - nothing apparent in the center
2) Right on the edge
3) Plenty of stars in full frame but nothing in subframe
4) Star in centre but using one near the edge
Last edit: 3 years 4 months ago by Paul.
3 years 4 months ago #62686
Attachments:

Please Log in or Create an account to join the conversation.

Time to create page: 2.486 seconds