Auto Focus with overshoot

Anything that doesn't fit into any of the other forums
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Auto Focus with overshoot

#21

Post by admin »

Hi,

to further update, I have made changes that add an option to completely ignore any frames captured while the focuser is moving, which will address your first issue I think - that will be in next week's update.

When you say you can't get HFD/FWHM to work right, can you explain in more detail? The recommended version is the multi-star FWHM (HFD) and I can usually get very good results with this.

The idea that the best focus would be at the lowest contrast level seems counterintuitive to me - do let me know if you find it repeatable! The reason the curve does not appear is that it 'knows' that it should find a curve with a maxima, so any curve with a minima is obviously wrong...

cheers,

Robin
WB91
Posts: 27
Joined: Wed Jun 17, 2020 3:13 pm

Re: Auto Focus with overshoot

#22

Post by WB91 »

Robin: The idea that the best focus would be at the lowest contrast level seems counterintuitive to me - do let me know if you find it repeatable! The reason the curve does not appear is that it 'knows' that it should find a curve with a maxima, so any curve with a minima is obviously wrong...

Bill: Not contrast, but brightness. When Jupiter is out of focus, it gets brighter. I will be shooting it tonight and will try and video it for you.
WB91
Posts: 27
Joined: Wed Jun 17, 2020 3:13 pm

Re: Auto Focus with overshoot

#23

Post by WB91 »

Robin,

Here is a video showing the overshoot method working while using your focus brightness method mode to focus on Jupiter (sorry if the resolution is low). I ran this focus method on my RGB filters to help me create filter offsets for each filter and got spot on for focus. Also, here is a timelapse I put together from this night using this focusing method. Thank you so much for adding overshoot into your software, helps big time!! If you happen to test this brightness focusing method on planets, let me know how well it works for you.

Video download link for brightness focusing using overshoot: https://drive.google.com/file/d/1zqXig9 ... sp=sharing

Video link to my YouTube account of Jupiter timelapse: https://www.youtube.com/watch?v=62tzs6-xHTA&t=8s
note: Seeing at the beginning of the video was below average but got better as time went on. Not to bad for Southeast Texas skies. 8-)


Regards,

Bill
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Auto Focus with overshoot

#24

Post by admin »

Hi Bill,

thanks for taking the time to record the video - I honestly don't understand *why* it should work that way, but it clearly does! That will require some thought :)

The Jupiter animation is excellent - the improvement in seeing is clear, but even at the start the quality is nice.

cheers,

Robin
umasscrew39
Posts: 408
Joined: Fri Apr 28, 2017 1:28 pm
Location: Central Florida
Contact:

Re: Auto Focus with overshoot

#25

Post by umasscrew39 »

Hi Robin

I gave this a try but I must be doing something wrong. I checked Overshoot to go in the positive direction and it moved as expected but it never reversed direction to get a best focus position. Doing the traditional method by entering my focuser's backlash worked fine. Am I missing a step?

Bruce
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Auto Focus with overshoot

#26

Post by admin »

Hi Bruce,

if you have turned on 'overshoot +ve' and set an overshoot amount of (lets say) 800, then this is how it should work...

Moving from 20000 to 21000 - the focuser should just move normally to 21000

Moving from 21000 to 20000 - the focuser should first move to 19200 (overshoot of 800) and then move back to 20000, so approaching in +ve direction.

This applies to *all* focuser movements initiated from within SharpCap. You should be able to see it in action if you set a big enough overshoot and just move the focuser using any of the SharpCap controls.

The focus scan will only run in the direction you have chosen for the approach (to avoid an overshoot at every step). The final movement to goto the best position should include the overshoot to approach in the right direciton with cleared backlash.

cheers,

Robin
umasscrew39
Posts: 408
Joined: Fri Apr 28, 2017 1:28 pm
Location: Central Florida
Contact:

Re: Auto Focus with overshoot

#27

Post by umasscrew39 »

Thanks but I clearly am not understanding what buttons to push in SC. I understand the theory and have seen several videos on it but all I get is the focuser moving a number of steps then it says completed like usual. Nothing else happens. So, I clearly do not understand how to initiate this process despite entering an overshoot of 800 in the positive direction. I click on Scan OUT and it goes out say for 10 steps, taking 3 readings at each step. It then is completed. Nothing else happens.
User avatar
admin
Site Admin
Posts: 13177
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Auto Focus with overshoot

#28

Post by admin »

Hi Bruce,

yep, scanning out will behave as normal in that situation - it's when you go back to the best location from the end of the scan that the overshoot should kick in. Suppose that your best focus position was at 18450 and the scan finished at 20000. For the sake of keeping things simple, type 18450 into the focuser position box on the right and press <TAB> to make it get applied. At that point the focuser should move first to 17650 then to 18450. Depending on how quickly the focuser moves it may be hard to spot the overshoot + move back as SharpCap only updates the value shown once or twice a second.

cheers,

Robin
Post Reply