×

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

Bi-monthly release with minor bug fixes and improvements

Full field focusing not in focus

  • Posts: 643
  • Thank you received: 62
Hi!

I've been playing a bit with the full field focus option, partly to aviod when the focus module fails to find a star to focus on or rather misidentifies some dust for a star after a meridian flip or similar. However, I cannot make it work.

It looks like it is working and settles for a solution, after some 7-10 iterations. I use 3 sec exposures, with a Canon 1000D camera. However, it settles for a position that is quite different from what I get if I focus on a single star. And the setting it settles on is definitely not in focus, upon visual inspection.

I don't know what to supply in terms of logs or data here. I use polynomial, gradient, 1 % tolerance for both full field and one-star subframe focusing.

And on the side, then - how to avoid misdentification of star, or non-identification? I do not mean "failed to find stars", but at times it just focuses on what looks like empty sky....

Best,

Magnus
6 years 1 month ago #23004

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

You have a sample of a FITS image that it fails to detect stars in? I'm thinking to use something based on sextractor in the future so that star detection is more reliable. So that should hopefully make this operation more fault-tolerant than now.
6 years 1 month ago #23025

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

  • Posts: 643
  • Thank you received: 62
Hi!

I have a 300 sec sub of the area around M97 where it failed several times - but I guess you want the actual fits file from the focuser module? Let me save one of those next time I see it.

What I've tried to do, is increase the exposure time, up to 20 secs, and to switch between gradient and centroid settings. I have a feeling that it works better with centroid - but that might be totally wrong, no affecting star detection at all...?

Either way: why is the full field giving me a different focus setting then? Wasn't it supposed to be more accurate...?

Magnus
6 years 1 month ago #23033

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

Here is the deal. The "centeroid" algorithm is the first one I wrote back in 2012 or 2013, but it's inefficient and quite susceptible to noise. However, it tends to work quite well for _most_ case nonetheless. It was designed to detect multiple stars but if in the first iteration if find a suitable star, then it stops, so you end up with 1 full frame image with one-star so accuracy wise, it's not really better. If your image ends up detecting 4 or 5 stars, then all the HFRs for those will be used and in principle would be more accurate.

At any rate, I plan to utilize something based on sextractor for this purpose in the future so that the results are more reliable, so you'd have to wait until that's done.
6 years 1 month ago #23064

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

Time to create page: 0.153 seconds