Narrow Band Imaging in Bortle 1 Skies

12 replies252 views
Steven Zuhoski avatar

After more than a year of running a remote 80mm F6 (FL=480mm) refractor in Bortle 1 skies, I am not getting the best results with my narrow band filters (4.5nm). My camera is a cooled IMX571 mono that I usually run in HCG (low read noise) mode at minimum gain. Pixel scale is 1.55 arcseconds with sub exposure times 3-5 minutes on a harmonic drive mount. Per all the exposure models (e.g., SharpCap model), in order for sky glow to beat out read noise (down to ~5%) in very dark skies, the sub exposure times need to be ~50 minutes. With my remote rig, even well calibrated Ha subs (~3 minutes) tend to be fairly noisy and several hours of stacked data shows artifacts like horizonal banding and not the best S/N.

A lot of folks here image remotely in very dark locations and post beautiful results. I am wondering what is considered the best practices for narrow band acquisition in dark locations and processing when dealing with read noise dominated subs? Should just ignore the sub exposure models and move on (the CanonBandingReduction script works fairly well….)? Should I move my much more capable Astro-Physics mount from my B9 backyard (almost all narrow band here…) out to the remote site (can easily do >10 minute subs)?

Thanks,

Steve

Well written Respectful
andrea tasselli avatar

If you can’t do 5 min subs then definitely think about replacing the mount. I mean even my Meade LXD55 virtually recovered from a skip can do that. In B1 sky I cannot find many reasons if any at all to go tight NB, especially with small, slow refractors and, yes, IMX571 can show some noticeable banding after some time when the BG level is virtually nil and the calibration isn’t absolutely perfect. Same here after 2 year of using it. Point is the is very little difference, as NB bandpass get smaller and smaller, from a B8 site and a B1 site. That’s the whole point of having them. Unless you have a faulty unit you read-out noise should be in the range of 1.5e- but you best check with SharpCap the actual results from your preferred mode of operation.

Tony Gondola avatar

I think I would work on the reason why you can’t take subs longer than 5 min. At 1.55” per pixel you’ve got a lot of wiggle room. You’ve got a light load and the mount is more than capable. Is it PA, guiding or something else?

Well written Helpful Respectful Concise Engaging
AstroGadac avatar

It is strange that you can’t do 5min subs with an harmonic mount. Theoretically if you are guiding you could make exposure as long as you want and get pinpoint stars discounting transient events like clouds or gusts of wind. If you are not guiding then this could be the answer since an harmonic mount need to be guided, otherwise this technology’s large periodic error will mess up your subs even at relatively short exposure.

Now for the second part, while theoretically as you computed the optimal sub exposure length can be very long under dark sky, remember that this is only a computation of the time required so that the read noise contribution vs the sky glow noise contribution is minimized relatively to your LP level. By that I mean that a 5min sub in a Bortle 1 sky will have far, far, far less noise than a 5min subs under Bortle 9, even if the former is ten times less than the “optimum” for a B1 sky, while the latter is more than enough to minimize the relative contribution of the read noise vs LP noise in B9. Its just that any remaining noise in the B1 sub will be mostly read noise, but the total amount of noise itself won’t be much (especially using an IMX 571). I’m suprised that you mention very noisy stacks with artifacts under such conditions to be honest.

Helpful Engaging
Steven Zuhoski avatar

Yes - the measured read noise for the camera was about 1.4e- at the minimum HCG gain setting when I sent it out to the remote site in 2025. I did not say 5 minute subs were beyond the capability of the light weight mass produced harmonic drive mount (the loaded rig is at about 2/3 the unweighted capacity of the mount). Sub exposures between 3 and 5 minutes seem to be a sweet spot for the set-up but I am tossing out more subs beyond those exposure times. The guiding runs a little under 1 arcsecond for the mount. I much prefer RGB subs around 3 minutes so I can aggressively cull the data without worrying about losing too much.

My questions are more related to how folks do narrow band imaging in very dark skies. Are they really exposing long double times to get optimal results are per guidance to have sky glow swamp out the read noise? What is the best compromise? This is not an issue in my light polluted home sky - it does not take long for sky glow to swap out the read noise.

Thanks for confirming that you can have banding with the IMX571 with a nil BG.

Tony Gondola avatar

I quick look at IOTD winners on the bin will show that 300 to 600 sec. is the most common for narrowband under b1-2 skies.

andrea tasselli avatar

When I use remote setups (so B1-B2 skies) I never go beyond 600s integrations, even with the very best mounts the market has to offer. With modern sensors RON is so small that that is a negligible contribution to the overall stacked noise. Satellite tracks or even airplane trails are far more harmful, especially with limited stack time as well as imperfect calibration frames. Obviously the real limit is the shot noise and there aren’t two ways about it, you got to shoot more frames. I normally take 300s subs and I’m done with it, no regrets from that point. I can see little difference from taking frames in my B7 as opposed to B1 sites (and at altitude compared to my 20m above sea level!). The main limit for my backyard is the lousy weather and seeing as far as I am concerned, not LP, when using tight NB filters.

Well written Helpful Respectful Engaging
Brian Puhl avatar

I think showing us pictures of your problem would be beneficial to helping you here…. but one thing I keyed in on is using ‘HCG’ and minimum gain. Would this imply you are using gain zero?

I run my OGMA at gain 200. I chose this because the ADU counts most closely matched up with my QHY-268M that I was running in cooperation, dual imaging on separate mounts.

Could this be your issue?

Helpful Engaging Supportive
Steven Zuhoski avatar

Thanks folks. I think I got my answer: I should run narrow band sub exposures between 300 and 600 seconds and not worry about swamping out the read noise.

I run my OGMA AP26MC most of the time in HCG mode at the minimum gain for that mode: Gain 100. With these cameras I can run it in HCG mode (or LCG) without having to set the gain to a particular level. I have used higher gain settings with narrow band filters, but you pretty quickly start getting ADU counts at 65K with modest sub exposures (~5 minutes).

Here’s a calculation result using some of my recent Ha data. Note the sub exposure length is 180 seconds but the calculator suggests I should be exposing for around 2000 seconds.

📷 Calc_Example.jpgCalc_Example.jpg

Well written Helpful Respectful Concise Engaging Supportive
Steven Zuhoski avatar

P.S. Yes I am fretting too much about what the calculated sub exposure times should be versus reality.

Alex Nicholas avatar

If I were to have the ability to have a rig at home and a rig at a remote site, and one of the mounts I owned was an AP… I can tell you now, the AP would be under bortle 1 skies, and whatever the other mount was would be with me at home…

I guess, I don’t have a remote scope, however, I feel that if I was going to outlay that amount of money on a monthly basis, I would want the best possible ROI… This would mean my best mount, camera and optics would be under the pristine skies, and I would keep a little toy rig at home to tinker with…

To your actual question - there are a lot of literature online about these things, and a lot of people go really hard out…

On my rig, I shoot both LRGB and narrowband with the same settings, and I don’t have any issues like this. Granted, I’m in Bortle 4/5 skies, and I know that its a different game in bortle 1, but, I run HCG, Gain 100, Offset 180, -10°C.. With my f/5~f/7 refractors, I’ll shoot 120s L, and 180s RGB, 300~600s narrowband depending on the target… with my F/2.8 newt I’ll typically shorten all that to 60s L, 120s RGB, but still shoot 300~600s SHO.

As far as offset is concerned, each camera will be different as to what the ‘best’ offset for a particular gain setting is, HCG/LCG is really a decision you need to make… If you’re running lots of short subs, HCG will yield the benefit of less read noise… If you’re running super long exposures, LCG can give you a slight edge in well-depth and dynamic range…

Well written Helpful
Brian Puhl avatar

Steven Zuhoski · May 12, 2026, 09:15 PM

Thanks folks. I think I got my answer: I should run narrow band sub exposures between 300 and 600 seconds and not worry about swamping out the read noise.

I run my OGMA AP26MC most of the time in HCG mode at the minimum gain for that mode: Gain 100. With these cameras I can run it in HCG mode (or LCG) without having to set the gain to a particular level. I have used higher gain settings with narrow band filters, but you pretty quickly start getting ADU counts at 65K with modest sub exposures (~5 minutes).


I run all my IMX571’s at 20 minute narrowband exposures standard, save for certain targets which are really bright to start with. Those I’ll run 10 minute since it helps build rejection quicker as well. I’ve done a couple runs at 40 minute narrowband but the gain is not worth the risk in rejection.

As for broadband I typically run 5 minutes standard, with a fewer dimmer targets in the 10 minute range.

You speak about ADU’s reaching 65k, but what you’re probably seeing is just the stars saturating here and there. Depending on the mag of the star, this can happen on even the shortest exposures and you should not be too concerned. There’s alot of misinformation running around the internet, I urge you to experiment for yourself.

I have done alot of experimenting myself. Had a rough night of bad seeing and thin clouds, so I decided to attempt 1 hour subs, just to test the ability of the camera. Now, let me preface by saying I am in no way recommending you run 1 hour exposures, the cores of two of the galaxies are in fact blown out, however… this image is still very recognizable. Here is a single 1 hour luminance exposure taken with the same exact camera you talk about above, the AP26MC. Gain 200. The flaws in the background are from thin clouds, however you can easily see I’ve picked up the entire tidal tail in a single sub, from a bortle 4 sky.

https://cdn.discordapp.com/attachments/247347265226801153/1423035890799874148/image.png?ex=6a04d773&is=6a0385f3&hm=d03423f943bcbc9982bed0efc250b22a64cff87aeb54f112da9c2ec435d9f3aa

Well written Helpful Engaging
Steven Zuhoski avatar

Thank you for your guidance. I need to use longer narrow band sub exposure times to get better results.

Well written Respectful