How long does it take to auto-focus?

I use Hocus-Focus and my autofocus sessions take between 2 and 3 minutes.

This depends on the image acquisition time you have defined.

I first made a Filter Offset calculation and always use the L-filter for the focus session. My defined time is 10 seconds.

This is IMHO a question of trial & error looking at the result you like.

andrea tasselli · Mar 16, 2026, 07:43 PM

Never did this and I wouldn't but I may have to. So this is the question:

Suppose to have to refocus during an acquisition run using NINA, how long would it take with an autofocuser (any EAF really) to achieve focus starting close to focus. For extra points, which one gives the best/faster results: Hocus-Focus or the built in default procedure?

Thanks

It varies, but generally with a ZWO EAF, it takes a little over a minute with Hocus Focus (1 sec exposure for LRGB, inner crop ratio 0.4, # steps 4) and a full frame camera. The inner crop ratio really speeds things up. I use several different focusers and the time are all very similar speed-wise. For HSO, I use anywhere between 4 and 8 second exposures depending on the scope speed. 4s for my f2.7 Ocal, 8s for my f7.7 TOA-130 . That lengthens the time by about 30-60 seconds.

With Hocus Focus, if you want to save some time, you can turn off “Validate HFR improvement” which takes an extra image before starting and after AF to compare to see if there’s an improvement, but I like having it on.

Rick

Typically 3-4 minutes but I use two exposures per point for better statistical reliability. I also use the green filter for focus and use offsets. Many people use Lum because it is faster but I have found green more reliable. Less light, smaller stars, and middle of the visual spectrum.

As an aside, never depend on one or two runs of the NINA offset calculator. It will often give highly variable results that are well outside the DOF so one needs a multiple run statistical approach to that as well.

between 1:10 and 1:30 every time. Longer is typically because the on scope miniPC is running a bit slow. Bin2 with offsets, and I’ve never had a single issue with autofocus on any of my setups.

📷 AFGraph
All my scopes are dialed in for autofocus in around 1 min. All scopes have offsets dialed in and only focus in luminance. Most everything is Bin2. Also, the native autofocus in NINA is far more reliable than Hocus Focus has ever been for me. I firmly believe if you aren’t shooting for 1 minute autofocus runs you are losing so incredibly much time every night. Even more for those who don’t use offsets. Add it up folks, you’d be surprised. I had an entire 5 autofocus runs on my last imaging night, lost a whole 8-9 minutes of time (to include guider settle). How much do some of you lose in a night? This is mostly directed at the folks who are claiming 4 minute autofocus runs and 8 second exposures. I’m not sure how you can live with that, especially when you add it all up.


andrea: When you first set up an EAF, it’ll be a couple minute runs. Depending on the focal length used, you may or may not need to bin. The idea is to keep the exposures fairly fast to shorten that time. Bin2 increases signal but also increases the ‘width’ of your autofocus run so that you’ll never be so far out of focus that the system can’t find itself again. If you want help, hit me up.

Brian Puhl · Mar 17, 2026, 04:27 AM

Also, the native autofocus in NINA is far more reliable than Hocus Focus has ever been for me.

Same here, the internet praises Hocus Focus for autofocusing and I really wanted to make it work but NINAs default autofocus routine is so much more reliable and faster on my setups.

Just find your ideal step size and exp. time and forget about it.

NE-FL-Astro · Mar 17, 2026, 03:50 AM

Rick Krejci · Mar 16, 2026, 08:09 PM

andrea tasselli · Mar 16, 2026, 07:43 PM

Never did this and I wouldn't but I may have to. So this is the question:

Suppose to have to refocus during an acquisition run using NINA, how long would it take with an autofocuser (any EAF really) to achieve focus starting close to focus. For extra points, which one gives the best/faster results: Hocus-Focus or the built in default procedure?

Thanks

It varies, but generally with a ZWO EAF, it takes a little over a minute with Hocus Focus (1 sec exposure for LRGB, inner crop ratio 0.4, # steps 4) and a full frame camera. The inner crop ratio really speeds things up. I use several different focusers and the time are all very similar speed-wise. For HSO, I use anywhere between 4 and 8 second exposures depending on the scope speed. 4s for my f2.7 Ocal, 8s for my f7.7 TOA-130 . That lengthens the time by about 30-60 seconds.

With Hocus Focus, if you want to save some time, you can turn off “Validate HFR improvement” which takes an extra image before starting and after AF to compare to see if there’s an improvement, but I like having it on.

Rick

Are you happy with your OCAL?

I must say I am. The optics are fantastic and, the mechanics are very solid. Collimation is very easy with the included camera. Tilt has not been an issue whatsoever, even at Full frame with a heavy 7 position filter wheel hanging.

Brian Puhl · Mar 17, 2026, 12:31 PM

Tobiasz · Mar 17, 2026, 08:31 AM

Brian Puhl · Mar 17, 2026, 04:27 AM

Also, the native autofocus in NINA is far more reliable than Hocus Focus has ever been for me.

Same here, the internet praises Hocus Focus for autofocusing and I really wanted to make it work but NINAs default autofocus routine is so much more reliable and faster on my setups.

Just find your ideal step size and exp. time and forget about it.

This, exactly. Hocus Focus needs lots of tweaking, and i’ve never had the time to waste getting it there. Folks come to me with autofocus issues, I ask them to swap back to NINA native, and they’re up and running in 10 mins.

I find Hocus Focus didn’t need tweaking at all, but it does allow tweaking if you have optics that sometimes cause focusing issues (like very large central obstructions or refractors with spherical aberration where the hyperbola is “tilted”). With my current fleet of scopes, I could probably go back to Nina focusing, but HF has been working great so I don’t have a reason to.

I bin 1×1 since my focal lengths aren’t large, but since I’m full frame and my scope all have fairly flat fields, I only use the 0.4 center of the frame to focus which speeds up times quite a bit without reducing focus accuracy.

Focusing was the bane of my existence for several years. Manual was always a hit and miss,miss,miss… So I built an Arduino based manual electronic focus driver. Much better but still required human intervention.

Finally, after doing a bunch of upgrading (or as a matter of the curse of upgrading…) I decided to man-up and buy a real auto-focuser. A comedy of errors with ZWO EAF, I finally have my Pegasus Astro Focus Cube2. Which has worked great out of the box.

My reason was it has a separate 12 VDC to power the motor so the USB is not trying to control as well and actually power the motor. Separates the control and the actual power to run the focuser. I got mine one week before they discontinued the Cube2, and went to the Cube3. The Cube3 is powered by the USB circuit.

It doesn’t really matter about how long autofocusing actually takes to do the fine adjustments. What matters is the quality you will realize from going to fully automatic focusing in NINA.

I’ve never bothered with how long it takes, although it does interrupt the actual imaging while the autofocus routine functions. And it can in my experience effect to total number of images I will gather that are usable.

The quality of the images in the Subs is sharper and better when stacked. So whatever time it takes, makes up in better overall images to work with. The amount of time Autofocusing takes out of your routine more than returns in higher quality subs.

Something I found very valuable was Cuiv’s tutorials about NINA. You might be able to mine out some clarity from his one about Auto-Focusing. When transitioning to full-auto operation (Well, simi auto) I went from front to back with his NINA stuff. My only problem was getting my Meridian Flip working right. But that was settings in my mount, not NINA.

Yep, fully automatic focusing is worth every second it might waste. Because the returns are great in better images. I use Hocus-Focus in my NINA routine.

Focus run duration is a function of settings and equipment.

For my refreactor/QHY600 combo, focus runs take about 90 seconds. For my ASA600/Moravain C5 combo, about 3 minutes. Much of the difference in time is due to the different cameras.

For HF in Nina, you can set exposure time for focusing and you can set the number of focus steps. The shorter the exposure time and the fewer the number of steps, the faster the focus run, obviously.

Image download times are also a factor. On my QHY600 system, these are de minimis. On my Moravain C5 system these are substantial and induce delay.

On my ASA600 RC system, I use 4 offset steps on either side of the focus point, and I do a check at the end, so a focus run generally involves 10 images.

I do not use filter offsets, so focus for each filter. I use 7 second exposures as my standard for focusing, because that works with LRGB and NB. 3 second exposures would work with BB, but 7 second are needed for reliable NB focusing. I did some testing, and switching filters from NB to Lum to focus and then switching back, did not save me any time, so this approach allows me to focus for every filter and has a minor time penalty for BB filters.

Neglected to mention that I have also selected a 1 second settle time after the focuser moves on my ASA600, but not on my refractor. Honestly, it probably is unnecessary, but for an extra 10 seconds per focus run, it buys me piece of mind.

Incidentally, In NINA focusing is dependent on parameters you can set and have work automatically.

So your equipment isn’t taking up precious nighttime. Mine does things like focusing after dithering (I like dithering every ten images), or temperature changes (10° IIRC), and other monitoring events, after a meridian flip. Honestly, they are so automatic they are more “set-it and forget-it”. Not really worth noting because they work for my telescope and imaging train but may be different for other imaging trains and telescopes.

Autofocusing can really raise the bar for your image results. Definitely with the effort and focuser expense. BOAH seems to apply in this sport. (Break Out Another Hundred) 🤑

Same. Generally speaking, my focus points for all filters are grouped closely together clearly all within the CFZ of each other, so no need for filter focus offsets. I’ve run a number of focusing sessions rotating through the filters at different temperatures to test that out, and the results are pretty consistent.

There also seems to be only modest sensitivity to temperature, though admittedly the temperature range at Obstech is generally moderate over the course of the night. Very different than my refractor which sees substantial focus drift with a 1C change in temperature.

So, it is clearly my paranoia that has me doing a focus run after each hour of integration/filter change, and I know it. But it costs me 15-20 minutes a night so I am willing to pay that price for the irrational peace of mind it buys me.