Improving RGB imaging quality under Bortle 8 light pollution

9 replies270 views
Tony Gondola avatar

On thing I’m really struggling with is getting usable R,G and B data under B8 skies. For bright, high surface brightness objects it’s not too bad but with slightly dimmer objects it’s a real struggle. Getting good data in the near Ir and anything narrowband is of course easy. The wider bandpass RGB frames tend to be a real mess an will often not give a good result in term of accurate color even if the S/N is there. I know this is all due to LP but I’m wondering if anyone has techniques for minimizing this problem. And yes, I know that going to a dark site is the cure but that’s not happening. Until I move to Sardinia, I’m stuck in B8.

Rick Veregin avatar

Hi Tony

I’m in B8, so I struggle as you are struggling. First and obviously, if you want the best images I would suggest focussing on what your conditions best allow, mostly NB and/or bright targets. It is the same as if you would love to image something at 10 degrees altitude, it will never that great, accept the limitations. So one images things higher in the sky, as that is what your site gives you.

I do understand wanting to do other RGB targets, so here are my suggestions based on 9 years of experience with OSC cameras in B8.

· Image when the target is overhead as the LP and LP gradients are the least there. The closer to the horizon the worse the LP will be.

· Image in the middle of the night, when it may be darkest in the city and when the sky itself is at its darkest.

· If you can set up a light blocking shelter, open at the top, or at least block out any nearby lights with some sort of shade so they don’t shine directly on your telescope.

· Use a long dew shield. I just ordered what looks like an amazing dew shield with baffles from Rouz Astro, which knocks out stray light. His demo with a laser pointer is amazing. I will report on this on my pages once I get a chance to test it. https://rouzastro.com/product-category/dew-shields/

· The smaller the field of view the less gradients—either in capture or in processing. I’m at 1x0.6 degrees for my FOV.

· Since sky background is higher, background noise is higher, you need to increase your total exposure to compensate, compared to those in lower Bortle. Not longer subs, with high LP in RGB with no filter, or even a standard LP filter, subs can be a minute or less, longer subs just risks saturation.

· I have been most successful using the L-Enhance as a LP filter, for galaxies, I get good results. I even add in the L-eXtreme if there are H2 regions. See for example: https://www.astrobin.com/wjmz2n/

· The Optolong UHC filter provides some benefits, but I think the L-Enhance is better in my conditions. The L-Pro doesn’t work for me, worse than nothing. https://www.astrobin.com/gijsiv/C/?nc=&nce=

· I do recommend to anyone who is choosing a filter to look up Thompson’s work here where he does a very good job of experimentally comparing LP filters, and showing the benefits or not of particular filters on NB and BB targets.

· Note any filter will require work to get the colour correct, so in your processing software learn how to do colour correction.

· Note as well, reflection nebula in B8 are extremely challenging, galaxies are a more favorable target for sure.

For processing:

· I use Startools, which has a background wipe feature with many options. With experience and a lot of trial and error I can usually get a reasonably flat background. In PI there is something similar, learn all its tricks. Or GradientXterminator in Photoshop, not sure what you use. This is the key, without good background correction—no luminance or colour gradients—you will not get reasonable colour across the image, there is no way to colour correct. I usually push the background extraction as far as possible, without eating into the main image. Spend lots of time on this step.

· In processing, you will likely have to give up on the very faintest parts of the image, it is difficult to keep them and wipe the background properly.

Hope this helps

Rick

Well Written Helpful Insightful Respectful Engaging
Tony Gondola avatar

Rick Veregin · Mar 4, 2026, 07:44 PM

Hi Tony

I’m in B8, so I struggle as you are struggling. First and obviously, if you want the best images I would suggest focussing on what your conditions best allow, mostly NB and/or bright targets. It is the same as if you would love to image something at 10 degrees altitude, it will never that great, accept the limitations. So one images things higher in the sky, as that is what your site gives you.

I do understand wanting to do other RGB targets, so here are my suggestions based on 9 years of experience with OSC cameras in B8.

· Image when the target is overhead as the LP and LP gradients are the least there. The closer to the horizon the worse the LP will be.

· Image in the middle of the night, when it may be darkest in the city and when the sky itself is at its darkest.

· If you can set up a light blocking shelter, open at the top, or at least block out any nearby lights with some sort of shade so they don’t shine directly on your telescope.

· Use a long dew shield. I just ordered what looks like an amazing dew shield with baffles from Rouz Astro, which knocks out stray light. His demo with a laser pointer is amazing. I will report on this on my pages once I get a chance to test it. https://rouzastro.com/product-category/dew-shields/

· The smaller the field of view the less gradients—either in capture or in processing. I’m at 1x0.6 degrees for my FOV.

· Since sky background is higher, background noise is higher, you need to increase your total exposure to compensate, compared to those in lower Bortle. Not longer subs, with high LP in RGB with no filter, or even a standard LP filter, subs can be a minute or less, longer subs just risks saturation.

· I have been most successful using the L-Enhance as a LP filter, for galaxies, I get good results. I even add in the L-eXtreme if there are H2 regions. See for example: https://www.astrobin.com/wjmz2n/

· The Optolong UHC filter provides some benefits, but I think the L-Enhance is better in my conditions. The L-Pro doesn’t work for me, worse than nothing. https://www.astrobin.com/gijsiv/C/?nc=&nce=

· I do recommend to anyone who is choosing a filter to look up Thompson’s work here where he does a very good job of experimentally comparing LP filters, and showing the benefits or not of particular filters on NB and BB targets.

· Note any filter will require work to get the colour correct, so in your processing software learn how to do colour correction.

· Note as well, reflection nebula in B8 are extremely challenging, galaxies are a more favorable target for sure.

For processing:

· I use Startools, which has a background wipe feature with many options. With experience and a lot of trial and error I can usually get a reasonably flat background. In PI there is something similar, learn all its tricks. Or GradientXterminator in Photoshop, not sure what you use. This is the key, without good background correction—no luminance or colour gradients—you will not get reasonable colour across the image, there is no way to colour correct. I usually push the background extraction as far as possible, without eating into the main image. Spend lots of time on this step.

· In processing, you will likely have to give up on the very faintest parts of the image, it is difficult to keep them and wipe the background properly.

Hope this helps

Rick

Thanks for the details Rick. I did forget to mention that I’m shooting mono and usually running at 900mm F/6 or 1580mm at F/10.4.

I have just recently enforced the rule of never shooting much below 60 degrees of altitude for just the reason you mentioned. It’s not actually that much of a restriction for me as lower altitudes are blocked by trees to the south and west anyway. To accommodate that I’ve set up my sequences so that I’ll do multiple objects in the night, switching from one to the other as they enter the “good zone”. This has helped a lot in general quality and sharpness of the data but has really increased the time needed to get a complete data set to process out.

Other than a dew shield I haven’t done anything about light shielding but I probably should. Not sure if the best solution are carefully placed small blocking panels far from the scope or a portable encloser as you’ve mentioned. Both are certainly doable.

On RGB subs, I’ve pretty much settled on 60 sec. for most subjects although I suspect I can go longer when running at F/10.4

I have found that running the F/10.4 setup really helps in all areas, especially with gradients. With the darked background they are not as much of a problem compared to what seems to be a general pollution of the colors that can sometimes seem imposable to balance out.

Anyway, thanks for the input. I’ll take a look at the links you’ve provided.

Well Written Respectful Engaging
Rick Veregin avatar

Hi Tony

Yes, the longer FL and smaller field certainly would make it easier on gradients. I think a full dark blind will work the best, unless you have just a few stray light sources around you. Unfortunately, I image from my driveway, so a temporary shelter is problematic. All I can do is put up a small blind for my neighbours house lights as needed.

The problem is it is surprising how wide a field allows stray light in.

Standard dew shield is around 1.5X the diameter usually, so light starts to impinge directly on the aperture at 90 - arctan(1.5) = 34 degrees! So when you point your telescope, there is a cone of 34 degrees in all directions where light starts to impinge directly onto your aperture. Any light source within this cone will start to create severe gradients.

All the best

Rick

Well Written Helpful Insightful Respectful Engaging
Read noise Astrophotography avatar

Hi Tony , my sky varies a lot here due to snow fall.

Try shorter exposures to avoid flooding the sensor with LP Run L Early then R try to shoot G and B after people go to bed.

May try a LP filter ?

How are you colour calibrating ?

Tony Gondola avatar

My exposure times for RGB will usually be between 15 and 60 sec. I’ll generally time R for lower attitudes, G and B higher.

On color calibration I do everything from SPCC to a simple background neutralization approach. SPCC works spectacularly well or fails just as spectacularly. It all depends on the data set. I can usually tell right after RGB recombination if it’s going to be a breeze or a bear.

Well Written Helpful Concise Engaging
Read noise Astrophotography avatar

That’s short subs

I am stumped must be the type of light you are surrounded by. It’s mostly old fashioned street lights here .

Tony Gondola avatar

I’m pretty sure of that too. I think it’s mostly LED lighting in my area but depending on conditions I think the color of the sky glow can vary.

Read noise Astrophotography avatar

Such a shame

They just built a load of house behind the forest here it’s LED.

I am moving 2028 to Asia B1 skies on a remote island and be done with this

Tony Gondola avatar

Funny you should mention that, I’ll be moving to Sardinia this year where conditions are much better.

Well Written Respectful
Related discussions
PixInsight 1.9.0 MultiscaleGradientCorrection for narrowband missing SII and OIII data in MARS?
Has anyone tried MGC for narrowband yet? I get stuck on choosing filters for RGB - it seems there is only data for Ha at the moment. Should we just choose R for SII and B for OIII for now?
Addresses RGB filter selection and data quality challenges in astrophotography.
Dec 22, 2024
Shooting "in the zone"
After a few years of experience shooting both OSC and MONO under B8 skies it’s become clear to me that for the most part, getting data, especially L frames below an altitude of 60 degrees, just isn’t worth it. Anything much below that and transparenc...
Directly addresses RGB data acquisition challenges under light pollution conditions.
Jan 31, 2026
One Click, Real Science
Early last summer, while I was working on a new image, it occurred to me that the SpectrophotometricColorCalibration tool in PixInsight was generating all of the information that is needed to plot star data from a globular cluster as a Hertzsprung-Ru...
SpectrophotometricColorCalibration tool may help achieve accurate color in RGB frames.
Oct 13, 2025
Serious Stray Light Issue - William Optics UltraCat 56
I’m posting this to see if anyone else has experienced or can verify this problem in their UC 56 telescope. I received the telescope February 2, 2026 from Agena Astro. The quality control certificate that was enclosed was dated December 5, 2025. The ...
Author testing in Bortle 9; RGB imaging under light pollution directly applicable.
Mar 1, 2026