Trouble with focussing with NB filters

I am struggling to focus consistently with my Ha and OIII filters. I use 7 points in the run but often they just bounce up and down giving no real curve at all or direction on the optimum focus point. My LRGB are usually pretty good.

I am currently using 30 second focus times for the NB filters. I have Baader LRGB and Astrodon 5nm Ha and OIII so they are not parfocal at all between the two groups.

Should I be trying to determine focus offsets from L for these NB filters? Does SGP have an automated way of arriving at doing these? Any tips for improving my NB focussing?


In my opinion trying to focus through narrowband filters is a losing battle. I would highly recommend setting up filter offsets, which would allow you to focus using the LUM filter and then the focus point gets adjust for the NB filters relative to the LUM.

Have a look at the documentation here:

The gist of that is to make autofocus runs with each of your filters when the temperature is fairly stable. From the control panel, Autofocus with LUM and record the focuser position using “Set Focus for Lum”. This records the focuser position in the filter settings. Repeat this for each of your filters. You may need to use longer exposures for narrowband, like 1min. Once you do this choose “Auto Adjust Focus Per Filter” and use your LUM filter for autofocus, then the other filters will be offset relative to the LUM filter’s focus point.

1 Like

Hi Brett,
I had similar complaints when I began with SGPro trying to autofocus with the NB filters. I was encouraged by the gurus to create the offsets and use Luminance for the focusing … and my complaints went away.
My filters are close to being parfocal so your results may vary, but I would recommend you try Joel’s recommendation.


1 Like

I agree. Offsets are the way to go. I have Astrodon filters that are themselves parfocal but if there is even one refractive element in your optics (objective, reducer, flattener, etc), your optics are very likely not perfectly apochromatic so you will probably see some differences between filters. Even high end scopes are not perfect. Both my Taks and an AP had this issue. Minor, to be sure, but enough to want to deal with. This also means you will very likely need different offsets for any change in optical elements (but that is what profiles are for…).

You could slew to a bright cluster or something, I suppose, and that is what single star focus programs do (such as FocusMax) but with SGP offsets are the best way to go.


I couldn’t agree more.

I started using off-sets 3 or 4 months ago instead of focussing with 9 points and 40 seconds per NB filter. Once you have ran a few repetitions between Lum and your NB filter to gauge an accurate off-set, AF is a breeze and much faster. Importantly it is as accurate at achieving critical focus. The time that used to be absorbed by AF with a NB filter means I can now capture an extra 900s or 1200s sub (or two).



Thanks all for the feedback. So do I just use offsets from Lum for the NB filters or use Lum to focus for RGB as well. Previously I haven’t had any problems focusing with RGB.


RGB should be OK either way. You are still getting about 33% of the light through the colour filters so the exposure time doesn’t need much increase in exposure.

Try it and see what works best for you.



I’m using my luminance for both narrow and broadband focusing. No problems.
Good luck on what you decide.


The only time it makes sense to do that is when you have an imaging run that uses both narrow and RGB. It takes no extra time to focus with the RGB filters and focusing thru the Luminance when imaging with RGB can not improve anything, only potentially make them worse so I, personally, would not do that.

Maybe if you had a very star-poor area to focus on but I doubt even that would be sufficient reason not to focus thru the actual imaging filter when it is practical to do so.

Ok, I am doing some runs with Lum, Ha and OIII. Getting a fair bit of variability. Some of the NB charts are basically meaningless, they seem to move randomly going flat, up and down in random patterns. I am writing down the results and comparing the differences. I was just going to average 5 runs and then type in the values in the filter tab but I might be wasting my time. I never seem to have so much trouble with NB filters in the past, I wonder why just recently SGP seems to be unable to focus the NB filters.

Once I am setup with figures for Lum Ha and OIII I see there is a focus with filter option in setup that I have selected Lum for.

I am assuming that if I don’t put a value in for the RGB filters in the filter tab it will just by default focus each time using the actual filters for RGB rather than defaulting to the selected autofocus filter which is Lum?


What I wound up doing for the narrowband filters was to revert to FocusMax/MaxIm to derive the offsets, then put those numbers into SGP. It would have been a lot of work if I had not already had FM V-curves for the systems but it did work nicely since FM does produce focus much more quickly for narrowband filters. Even then I had to make multiple runs to get a statistically valid set of samples. Of course one would have to have not just FocusMax but also MaxIm or CCDSoft in order to do it that way, I already had those as they used to be my primary imaging software.

I’m not sure if this is the case or not. Personally, I would set up the offsets for all your filters, not just the narrowband ones. It takes much less time to do focus runs on the RGB filters to determine the offsets so in my opinion it is worth it to take the time.

The refraction of different optics might mean that the focus point for various filters is different. One could argue that this would mean that you should focus through each filter, but practically speaking I have not found this to be essential. I have been using offsets for all my filters for several months and I haven’t looked back. I actually believe that my focus is better on all my filters (not just NB) using offsets - I think because the Lum gets the best focus overall and then the other filters are adjust accordingly. Your mileage may vary…

It does also concern me that you don’t seem to be getting repeatable focus results for your narrowband filters. Sometimes that means your parameters are not set correctly (number of steps, step size), but it could also point to a mechanical problem. Just a thought, make sure your focuser is tight and working properly.


I agree with Joel that there may be a mechanical problem with your focuser if you are not getting repeatable results.

I had a nightmare getting my focuser to behave for autofocus with any filter even though I had measured the backlash and tried to compensate for it. I resorted to a Bahtinov mask with Bahtinovgrabber software to focus before each session and at times during the session. Needless to say, I lost a lot of subs after it cooled down a bit more than desired. I traced the problem to a wobbly pinion block…if I tightened it too much, the focuser would be too hard to turn, too loose, too much play. A piece of scrap channel aluminum and 4 nylon screws from Lowe’s for $1 fixed the problem and now I can get repeatable textbook V-curves with 15 second exposures binned 2x2 with narrowbands. Now if I can only get the weather to cooperate (darn winds)!

Frank Z…

I doubt there is a mechanical problem with the focuser because my Lum results (and generally RGB) have been consistently good in terms of accurate looking V curves and my subs have been sharp. If there was a mechanical error it would have to effect the LRGB filters as well and I am not seeing that.

My challenge now is to get the NB offsets as accurate as possible. Last night using the figures I calculated from some focus runs I got some pretty reasonable Ha subs with the offsets, all seemed to work correctly and the correct focuser position I was expecting was selected during the sequence. I will test the OIII tonight if I can. I will have to work out the best way of tweaking the offset but atleast I am in the ballpark for Ha.

For those that only use offsets for NB filters can anyone confirm I just leave the RGB filters with no value in the filter setup tab and SGP will just continue to focus with them rather then going to Lum as my selected focus filter? If this is not the way to set this up then how do I achieve this?


I am not sure that you can have offsets enabled for one filter and not for the others in the same sequence. I may be wrong but I suspect the result of what you propose would just be that no offset is applied (instead of focusing thru the current filter). That is because what filter you focus thru is determined by the dialog below, and that would not change:

You can, of course, set up different profiles, one for your narrowband and one for your RGB. That works fine as long as you are not doing both narrowband and RGB in the same sequence. If you are, I think you will need to just calculate and use offsets for all the filters. That is what I do for combined sessions.

It might be nice if there was a feature that would allow one to specify on a per filter basis whether or not the offset or the focus thru filter was to be used. Maybe make this a feature request? What are your thoughts?

Like Joel, I moved to focusing with filter offsets a few months ago and have found my results much better. I decided to use the red filter for focusing, though. Any thoughts on that? My instinct was that an autofocus run in red is less likely to be thrown off by atmospheric conditions. I use 15 second AF exposures, so even with my narrow 28 x 22 arcmin field of view I’m getting plenty of stars. So far it seems fine.

Brett, I sympathize with your experience in determining the correct offsets for NB. The autofocus runs for my narrowband filters were much more variable than for my LRGB filters - which, of course, is exactly why I wanted to start using red or lum for focusing. So I had to average a lot of AF runs before I was confident that I had the offsets correct.


Surprised to read these comments because I have no issue focusing with NB filters, either Astrodon 3nM or Baader 7nM. I use 10 sec exposures for focus and about 15 samples (mainly because I have a set of filters that are not parfocal and I don’t use offsets, so these 15 cover the whole range of expected movement).

screen grab of an Ha focus run: