(Auto)focusing & flats - Forums

Topic starter

Apr 6, 2023 · Edited: Apr 6, 2023

Hi all,

Right now I am always focusing only at the beginning of each session (my flow with KStars/Ekos is first 2-star alignment, plate solving, polar alignment, autofocus, start shooting, after shooting I take darks and in the morning I take flats/dark flats without changing optics of course). I forget about focus once I start the session (I know naive me) and just assume it remains the same.

I was thinking of enabling autofocus during the session (either triggered if focus decreases or just periodical focus) but I didn't integrated it yet to my process for two reasons:
- I was a little bit afraid it would go bad in the middle of session and I'd need to spend a lot of time to get focus back again

(I had quite some issues at the beginning but after fixing it now I feel much more comfortable as I got really good and stable results lately)
- I don't know if it's OK to refocus and do flats in the morning only with latest focus

So my question is - what is your experience with re-focusing during session and then taking flats at the end (in the morning)? I assume the re-focusing is so small that it's not noticeable on flats? Is there any unwritten rule what is the tolerance which one could use which flats are still useful when focus changes?

I started to think about this question after stupid and sad story involved de-focusing during session yesterday, full story available here (yes yes I know, stupid me not to triple check before spending 3 hours

): https://www.astrobin.com/m78e9r/C/

andrea tasselli

Frequent contributor

Apr 6, 2023

Short answer: don't bother, whatever shift in focus you may end up with isn't going to affect the flats.

Jim Raskett

Apr 6, 2023

andrea tasselli:
Short answer: don't bother, whatever shift in focus you may end up with isn't going to affect the flats.

Absolutely! Don’t bother!

As far as focusing during the session. My focus changes drastically some nights and very little on others.
Temperature variation is the reason and I autofocus (NINA Hocus Focus) at the beginning of the sequence and then every hour regardless of the temperature. There are other options besides time, such as HFD, or temperature change (if you use a temperature sensor.

Good luck!

Jim

Jure Menart

Topic starter

Apr 6, 2023

Hi Andrea, Jim,

Thanks for your answer. Up to know I didn't enable it just as I wasn't sure about this but it's good to read that I shouldn't overthink this

I will enable it and just re-focus regularly from now on, it seems like the best solution especially if HW & SW is there to support it.

Jure

AstroBertUK

Apr 6, 2023

Similar to Jim, I start the session by manually running the autofocus - this is to ensure that platesolving works OK. The NINA sequencer is then configured to autofocus after acquiring the specified target, and then automatically refocuses if the HFD changes by 10%. This, I believe ensures the best focus possible during the session.
Flats I will normally take either after the first manual autofocus, or at the end of the session. I'll also take them if I change filter.

NINA certainly helps with most of the 'hard' work

Well Written Helpful Insightful Concise

Jure Menart

Topic starter

Apr 6, 2023

Thanks Bert!

Maybe I should start to think of buying the panel to take flats. For now I need to wait until the morning to do the white t-shirt flats

It has been working quite good for me TBH.

AstroBertUK

Apr 6, 2023

A panel certainly makes things easier - I bought this 'cheap' tracing panel from Amazon and never had a problem with it

Amazon Tracing Panel

Jim Raskett

Apr 6, 2023

Similar to Jim, I start the session by manually running the autofocus - this is to ensure that platesolving works OK. The NINA sequencer is then configured to autofocus after acquiring the specified target, and then automatically refocuses if the HFD changes by 10%. This, I believe ensures the best focus possible during the session.
Flats I will normally take either after the first manual autofocus, or at the end of the session. I'll also take them if I change filter.

NINA certainly helps with most of the 'hard' work

Thanks Bert for reminding me!
I also run an initial manually triggered autofocus to insure plate solving works ok for both polar alignment (TPPA) and initial target centering (plate solving) in the sequence.

Georg N. Nyman

Apr 6, 2023

IMO, it is of absolutely no importance to keep a certain focus for aqcuiring flats…. important is only to have the same optical train for flats as for imaging…

CS

Georg

John Hayes

Frequent contributor

Apr 6, 2023 · Edited: Apr 6, 2023

Jure,
The important thing is to A) achieve precise focus when you start and B) maintain precise focus throughout the night. Then, when you take flats they will also be taken with the proper (precise) focus. Your C8 has a temperature sensitivity of about +/- 1.3C before it goes out of focus enough to matter under roughly 1" seeing conditions. That means that if you focus very precisely to the center of the critical focus zone, a temperature drift of only ~1.3C will pull the focus out of the CFZ. If the ambient termperature of your site during the entire night is VERY stable, you don't need to refocus; but, that is rarely the case. How often you need to refocus will depend on your site but in general, I would recommend refocusing every 30m - 40m with a C8. I think that you should see a noticeable improvement in sharpness if you periodically refocus during the night.

John

Well Written Helpful Insightful

Jure Menart

Topic starter

Apr 6, 2023

Hi John,

thanks for very good description and woth real numbers for C8 - I am not that far yet (just installed autofocuser last month). I will take your advice and implement re-focusing every 30 minutes. Let's hope for nice weather soon to see this improvement

Thanks again and clear skies,

Jure

Respectful Supportive

andrea tasselli

Frequent contributor

Apr 6, 2023 · Edited: Apr 6, 2023

The value of the CFZ is equal to 2*2.44*Lambda*F#^2, which with Lambda=0.55 um and F#=10 for a C8 (not reduced) yields a value of 0.268mm. This, if I am not mistaken, is the distance from best focus to cause a wavefront error of 1/4 Lambda. I strongly doubt that you need to be strict in a typical amateur seeing between 2" and 3" (and that is when it is good). Unless you want to spend a lot of time refocusing instead of imaging it would be a far better strategy to refocus only when your measure of best focus (typically HDF) changes by a fixed amount (most of us would settle at around 10%).

Helpful Insightful

John Hayes

Frequent contributor

Apr 6, 2023 · Edited: Apr 6, 2023

andrea tasselli:
The value of the CFZ is equal to 2*2.44*Lambda*F#^2, which with Lambda=0.55 um and F#=10 for a C8 (not reduced) yields a value of 0.268mm. This, if I am not mistaken, is the distance from best focus to cause a wavefront error of 1/4 Lambda. I strongly doubt that you need to be strict in a typical amateur seeing between 2" and 3" (and that is when it is good). Unless you want to spend a lot of time refocusing instead of imaging it would be a far better strategy to refocus only when your measure of best focus (typically HDF) changes by a fixed amount (most of us would settle at around 10%).

There are different ways to define the CFZ. The one that you've given is a geometric definition that uses the size of the Airy disk, which can be simplified to be ~ 2.5(F/#)^2 (in microns) in visible light. If you use the Rayleigh quarter-wave criteria, the CFZ is 4*lambda*(F/#)^2 or about 2(F/#)^2 (in microns) in the visible. The Rayleigh criteria is great for a space telescope but as you pointed out, it's overly restrictive in the presence of seeing. One thing to remember is that smaller telescopes with an aperture "about" the size of the Fried parameter are less sensitive to seeing than larger telescopes. Even though most locations don't experience spectacular seeing, I favor using an optimistic, "best-case" assumption of 1" seeing, which is appropriate for either small telescopes (<10") or for really good sites. In that case, the S-CFZ is determined geometrically from the seeing limited blur diameter. Designing to 1" also provides some "head-room" so that things will work just fine even when the conditions are worse.

The really important thing is to know the temperature sensitivity of the optical system. That's the change in temperature needed to shift the focus by half the full value of the CFZ. Here is a chart that I put together a few years ago showing the computed values along with some published values for various telescope designs. The disclaimer in yellow reflects the fact that these numbers are all computed from the design of the systems; not from measured data. There may be other subtle effects in any particular system (such as loose fasteners, or poor bonding) that could cause problems.

John

Well Written Helpful Insightful Engaging Supportive

andrea tasselli

Frequent contributor

Apr 7, 2023 · Edited: Apr 11, 2023

Personally, I never experienced 1" long exposure seeing, even with a 1 mt scope in Chile's Acatama. But that is digressing. I shall only note that I use some of the configurations listed above and I never have to refocus as often as that table suggests. In fact, the telescope I need to refocus most often isn't a telescope at all, it's my Nikkor AF-S 300m lens @ f/2.8. But then it is every hour or so at most.

Miguel G.

Apr 11, 2023

andrea tasselli:
Short answer: don't bother, whatever shift in focus you may end up with isn't going to affect the flats.