Autofocus seems suspect

Questions, bug reports, requests for enhancements, etc.
iMustBcrazy
Posts: 338
Joined: Sun Jun 02, 2019 3:52 pm

Re: Autofocus seems suspect

Post by iMustBcrazy »

Hi Rumen,

Last night's session was fairly successful. Since the sky was murky with smoke, fog and haze, I tried imaging M13 at the end and it turned out pretty well.

So below are my notes and a link to a log for the whole session - after the focus tests, I put logging back to error or debug perhaps

https://www.dropbox.com/s/bgpflpnnb5oxd ... g.zip?dl=0

NOTES:

focus 1 = AF (auto focus) AS (auto select), success on star 1
looks ok

focus 2 = AF MS, failed on star 1

calibrated guiding
started guiding

focus 3 = AF AS on star 2
(auto selected bright star near saturation)
HFD = 4.7
peak = 47547
position = 29513
looks ok

measure star 1 -> HFD = 6.2 peak=2337 (why was HFD worse than bright star?)

focus 4 = AF MS on star 1
failed

focus 5 = AF MS on star 2
success
HFD = 4.4
peak = 60553
position = 29500
looks ok

focus 6 = AF AS
selected star 1
failed

stop guiding

focus 7 = AF MS on star 3
HFD = 5.1
peak = 5972
position = 29449
looks ok


Guiding

1st = multistar

2nd = aborted

3rd = DONUTS at edge clipping = 8

4th = DONUTS at edge clipping = 16

5th = multistar with 5 stars

multistar set back to 3 stars

imaging m13 = multistar selected m13!

DONUTS with edge clipping = 8, selected a frame away from stars on edge

imaging m13

stopped - guiding not so great

guiding with multistar = 3
recalibrate - no , selected m13

back to DONUTS

calibrate

guide


6th guide log on m13 during imaging

M13_Hercules_Globular_Cluster-mod-cbg-csc-St-lowres.jpg
M13_Hercules_Globular_Cluster-mod-cbg-csc-St-lowres.jpg (432.81 KiB) Viewed 192 times
iMustBcrazy
Posts: 338
Joined: Sun Jun 02, 2019 3:52 pm

Re: Autofocus seems suspect

Post by iMustBcrazy »

Just to note specifically there was a problem with multistar auto select in that it selected M13 as a star, so I couldn't use multistar.
rumen
Posts: 133
Joined: Wed Feb 22, 2017 7:31 pm

Re: Autofocus seems suspect

Post by rumen »

Let is start with the things one by one:
1. focusing:
I would say it performed quite well and 100% as expected. But we really need to write some documentation for it :roll:
It failed because the initial stellar image was quite out of focus and bigger than the box size, in this case FWHM is 2x radius +1 in your case 17. and HFD would be 3/4 of the diameter ~12 in your case. This means that the star is clearly larger than the selection.
This is repeated from the start to the end of the failed focusing process:
01:54:12.445956 indigo_server: 'FWHM' = 17
01:54:12.446020 indigo_server: 'HFD' = 11.986

It failed because you need to have a measurable HFD.
This is not a limitation only for our algorithm, V-curve for example is even more sensitive to the initial conditions.

Will analyze the guiding later and will write here need to go now...
iMustBcrazy
Posts: 338
Joined: Sun Jun 02, 2019 3:52 pm

Re: Autofocus seems suspect

Post by iMustBcrazy »

rumen wrote: Thu Aug 05, 2021 5:51 pm Let is start with the things one by one:
1. focusing:
I would say it performed quite well and 100% as expected. But we really need to write some documentation for it :roll:
It failed because the initial stellar image was quite out of focus and bigger than the box size, in this case FWHM is 2x radius +1 in your case 17. and HFD would be 3/4 of the diameter ~12 in your case. This means that the star is clearly larger than the selection.
This is repeated from the start to the end of the failed focusing process:
01:54:12.445956 indigo_server: 'FWHM' = 17
01:54:12.446020 indigo_server: 'HFD' = 11.986

It failed because you need to have a measurable HFD.
This is not a limitation only for our algorithm, V-curve for example is even more sensitive to the initial conditions.

Will analyze the guiding later and will write here need to go now...
Thanks Rumen, I should note that I increased my gain by 2x after these tests and that seemed to help as well. So normally I shoot gain=120 but then gain=240 for the focus tests, then 300 for my imaging focus (at 5 sec - scope is F7)
rumen
Posts: 133
Joined: Wed Feb 22, 2017 7:31 pm

Re: Autofocus seems suspect

Post by rumen »

Just passing by... About M13 being a star - yes this may happen and I am not sure how to handle this. It looks like a noisy star to the algorithm. Furher more to detect stars and avoid hot pixels being detected as stars we do some image filtering which makes globular clusters even more star like...

And as i mentioned in one of my earlier posts overpopulated star fields may confuse Donuts. M13 is one of there cases. In theory Donuts will do better with OAG where there are several stars visible. Also it may do better if you de focus M13 I am not sure...
rumen
Posts: 133
Joined: Wed Feb 22, 2017 7:31 pm

Re: Autofocus seems suspect

Post by rumen »

This may be useful. I have spent 2 hours today preparing it:

https://github.com/indigo-astronomy/ind ... _TUNING.md
iMustBcrazy
Posts: 338
Joined: Sun Jun 02, 2019 3:52 pm

Re: Autofocus seems suspect

Post by iMustBcrazy »

rumen wrote: Fri Aug 06, 2021 3:08 pm This may be useful. I have spent 2 hours today preparing it:

https://github.com/indigo-astronomy/ind ... _TUNING.md
Awesome. I may be able to use it tonight - going camping with scope.
rumen
Posts: 133
Joined: Wed Feb 22, 2017 7:31 pm

Re: Autofocus seems suspect

Post by rumen »

Let me know if it was helpful. Any suggestions a re welcome.
iMustBcrazy
Posts: 338
Joined: Sun Jun 02, 2019 3:52 pm

Re: Autofocus seems suspect

Post by iMustBcrazy »

rumen wrote: Fri Aug 06, 2021 8:55 pm Let me know if it was helpful. Any suggestions a re welcome.
It was a bit more complex than I was expecting. I’ll have to review it at home. But focus worked consistently tonight so that’s good!
rumen
Posts: 133
Joined: Wed Feb 22, 2017 7:31 pm

Re: Autofocus seems suspect

Post by rumen »

This is great, but you have already figured out the parameters :)
Post Reply