×

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

Bi-monthly release with minor bug fixes and improvements

New Attempt a Focusing

  • Posts: 309
  • Thank you received: 40
Ok, after a few attempts at auto-focusing (AF), I'm going to try only AF on the Lum filter and using the offsets found on the initial Bahtinov mask focus.

What I'll do since I only have been imaging with my Ha/Oiii filters, I'll have one Lum image stuck in between the others at only 1 sec just to get the AF to activate and let the offsets take care of the others. I've found that using the Bahtinov on the 1# filter (Lum) that the others rarely need to be changed... maybe by a very small amount.

But the AF on the Ha/Oiii just doesn't nail it.
Last edit: 1 year 5 months ago by David Tate.
1 year 5 months ago #87550

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

  • Posts: 70
  • Thank you received: 10

Replied by Tracy Perry on topic New Attempt a Focusing

I'm finding using Linear 1 that ALL my auto-focusing is off by about 50 steps. Use the auto-focus and it says it's good. Put the Bhatinov mask on (WO ZenithStar 103mm) and it's readily apparent that the focus is off and needs to be manually adjusted.
This is using the latest code as of 10/27/2022. I did a manual update/upgrade just before starting tonights session.
Last edit: 1 year 5 months ago by Tracy Perry.
1 year 5 months ago #87556

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

  • Posts: 593
  • Thank you received: 274

Replied by John on topic New Attempt a Focusing

David Tate. When you say "But the AF on the Ha/Oiii just doesn't nail it." can you explain what you mean, what does AF do/not do? Probably be good to show some screenshots of whats happening and include some details of your focus setup.

Tracy Perry: Could you include some screenshots of the focus setup please, Are you on 3.6.1 stable or bleeding-edge as of yesterday and is this a new problem with this software version (i.e. was it working fine and is now broken or has it never worked)?
1 year 5 months ago #87559

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

  • Posts: 309
  • Thank you received: 40

Replied by David Tate on topic New Attempt a Focusing


I see the same. AF, then write it down (I keep a spreadsheet to calc the offsets from the Lum) and when I tested using my Bahtinov, I'm just not spot on. But I think I'm close enough... I think. Downloading the images now.
1 year 5 months ago #87563

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

  • Posts: 309
  • Thank you received: 40

Replied by David Tate on topic New Attempt a Focusing

The Ha/Oiii will fail many times and either just quit and the whole process stops (unless I wake up and check I can loose hours of imaging)... or when it did work, they were very out-of-focus. So much so Pixinsight threw them out.
Here are my current settings:
1 year 5 months ago #87565
Attachments:

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

  • Posts: 309
  • Thank you received: 40

Replied by David Tate on topic New Attempt a Focusing

Here's what I did with the Sequence Queue. I have a Lum for the first row (that is missing), I deleted it after a successful Meridian Flip and Reset the Jobs so I wouldn't run out of rows


And my FilterSettings (ignore the Red/Green... I haven't been using them lately) But you can see I only AF on the Lum.
1 year 5 months ago #87566
Attachments:

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

  • Posts: 593
  • Thank you received: 274

Replied by John on topic New Attempt a Focusing

Hi David,
I see what you're doing with the 1 sec Lum to force a refocus after each Ha/Oiii sequence. Another way to do this is to turn AutoFocus on for Ha and Oiii and put Lum in the LockedFilter. When you swap from say Ha to Oiii it will trigger an autofocus for Oiii but will use the Lum filter. When the autofocus run competes it will then adjust focus by the offset for Lum to Oiii and start capturing.

Thanks for autofocus screen shots. A couple of observations:
1. You are using a single star. Personally I prefer to use multiple stars. If you want to try this select Full Field (rather than Sub Frame).
2. The step size looks too small. I'm guessing you have it set to 20. I would up it to say 60 and set the number of points to 4 or 5. If this works better then you can try reducing the step size a bit.
3. I would also suggest trying the Linear 1 Pass algorithm if your focuser is predictable in that if you move to position X it always goes to the same position. Sorry, I don't have a deepskydad. If it is predictable then Linear 1 Pass will likely give a more reliable solution than Linear. Set the curve fit to Hyperbola and tick Use Weights.
4. Points 1 to 4 on the focus graph are flat or trending slightly upwards. This could be uncorrected backlash. If you get this consistently its likely that your focuser has more backlash than Linear is correcting for. There are a couple of ways around this but I'm not familiar with the DeepSkyDad so worth familiarising yourself with that (if you haven't already of course). One way is to measure the backlash and add at least this amount into the backlash field. But if you do this then you have to make sure the focuser itself is not attempting to compensate for backlash or there will be a clash.

Hope this is helpful!
The following user(s) said Thank You: Chris Madson
1 year 5 months ago #87570

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

  • Posts: 309
  • Thank you received: 40

Replied by David Tate on topic New Attempt a Focusing


There's a lot to go over here... thanks BTW!
-What I'm doing is keeping HA//Oiii from doing any AF... since it doesn't seem to work well.
-Single star with a Sub frame is much faster than Full Frame. But I'll give the multi-star full frame a shot.
-I'll up the Step size for a try, but I've found when doing the Bahtinov mask it goes too far and I end up reducing the step to get it right. True AF bypasses that when fine tuning. You can see in the picture I have it stepped down gear-wise.
-Yeah that Linear-1 Pass looks promising. But I thought that Linear goes back and tries the best values.
-Backlash, I really tried checking for it and I see the knob moving with even 5 steps. So I've left it at 0. I should test with a step of 1.
Last edit: 1 year 4 months ago by David Tate.
1 year 4 months ago #87592
Attachments:

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

  • Posts: 593
  • Thank you received: 274

Replied by John on topic New Attempt a Focusing

-What I'm doing is keeping HA//Oiii from doing any AF... since it doesn't seem to work well.

Yes, you can obviously continue with the approach you describe. The method I described does the same thing but without the need to enter dummy rows in the scheduler. But both achieve the same thing so it doesn't matter.

-Yeah that Linear-1 Pass looks promising. But I thought that Linear goes back and tries the best values.

At the top of the focus section of the Kstars Handbook is a brief description of the various algorithms and their pros and cons...
docs.kde.org/trunk5/en/kstars/kstars/tool-ekos.html#ekos-focus

-Single star with a Sub frame is much faster than Full Frame. But I'll give the multi-star full frame a shot.

FYI, you can limit the amount of full frame that is used by adjusting the annulus fields, so if you used half of the full frame that would speed things up. There are other, more complex, ways of limiting the processing but the annulus should help to start with.
1 year 4 months ago #87598

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

  • Posts: 309
  • Thank you received: 40

Replied by David Tate on topic New Attempt a Focusing

Is there any benefit to using 2x2 binning? or even more? 3x3?
1 year 4 months ago #87609

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

  • Posts: 593
  • Thank you received: 274

Replied by John on topic New Attempt a Focusing

If you struggle with faint stars then you can either increase the exposure or gain or up the bining. Something to consider is that if you go from say 1x1 to 2x2 you are doubling the image scale (arc-secs/pixel) which will impair resolution. Whether that is noticeable in the focus process depends on many factors.

Personally I would start at 1x1 and only increase if there is a reason to do so. For example, if you need a 30sec exposure to gather enough light for the star detection to work. In that case you could try 2x2 bining and significantly cut the exposure time. Assuming of course that your image scale on 2x2 bining is good enough for the HFR process to work.
The following user(s) said Thank You: David Tate
1 year 4 months ago #87610

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

  • Posts: 309
  • Thank you received: 40

Replied by David Tate on topic New Attempt a Focusing


Last night after doing some fine tuning with the Bahtinov mask, I forgot to move the step size to 50 (had it at 10) and AF would not work. I saw it and changed it to 50 and it worked right-away!

Thanks John!
The following user(s) said Thank You: John
1 year 4 months ago #87649

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

Time to create page: 0.570 seconds