Approach and equipment recommendations for Galaxy Imaging in light polluted city

Hi Rob, 

I think the only way to "compensate" for light pollution in broadband imaging, is to gather 3-4 times more data than you would do under darker skies. 

I am "blessed" with Bortle 5 skies above my backyard and don't have much experience shooting under worse conditions. But I am following Cuiv The Lazy Geek and he does galaxy imaging from Tokyo (!), which is the most light-polluted spot on the planet. He uses  8" F/4 newtonian and Edge HD SCT and produced very nice images!

Hi Rob,

I do broadband imaging with OSC, galaxy or otherwise, in 6+/7 skies. The n.1 thing is to get the proper image scale and thus sampling for you local seeing. Mine is between 2" and 3" depending on the season (better in the summer months and worse in winter). Thus my image scale with the 294MC is ~0.8"/px thus sampling my typical average FWHM well enough (between 3 and 4 pixels). I keep doing 3 min subs but could go down to 1 min in most part of the sky without any loss, especially toward the southern horizon which is badly affected by LP. I just cannot be bothered with the extra faff of handling large sets of subs. As others have postulated, you compensate the LP (natural and man-made) by increasing exposure time. Think 8x of a truly dark site (Bortle 1 & 2). And you'll need a lot anyway for most of the galactic faint halos. So, no less than 10 hours in my case, and upward to 3 times as much for some faint targets. My preference would be pure OSC as I cannot for the life of me go down the rabbit's hole of tackling severe LP with 4 filters (LRGB). With an OSC you just need a good LP filter and you're done (mostly…).

Thanks Andrea, that very helpful. Do you use an LP filter on your OSC even for galaxy imaging? Cheers rob

Optolong L-PRO is the way to go, IMO. All broadband imaging I carry out (well, with the 294MC anyway) are done with it in place.

Lovely triplet Cesar. Great detail regardless but all the more impressive from Bortle 8 and using a DSLR. Thanks for sharing!. Rob

Thanks Falke, some lovely Galaxy shots there with your asi533mc. Shows a lot can be done without filtering just longer integrations. Love the star colours as well.

Hi Rob.  I faced the same problem.  Interested in galaxies but light polluted skies (Bortle 6) that puts the fainter stuff out of reach.

My solution  has been to take a slightly different tack and use lucky imaging to really dig into the detail of the cores of the brighter galaxies because that is something that you can do from a Bortle 6 site as well as from anywhere.  The seeing can be OK even while the transparency and light pollution may be poor.  Then on occasional trips to dark sky sites take possibly  lower resolution but far better SNR images to fill out the outer and fainter parts of galaxies

Equipment wise this has meant going for fast fairly long focus big reflectors (e,g, an F 4 Orion UK VX12 300mm and  F 5 SW PDS200) at an image scale of ~ 0.5 arcsec/ pixel  and also a CEM70 mount which can just about take these while remaining portable enough for field use.   These combinations provide reasonable SNR of the brighter core regions even at the short exposure times of less than a few seconds required for lucky imaging.   For lucky imaging  I have used a MONO camera (ASI294 MM) in unlocked pixel mode (2.315 uM pixels)  to j get luminance at a high sampling rate.  These core images  can sometimes be further enhanced  by deconvolution.  Then I have used an OSC (ASI294MC) to get the colour data at lower resolution (4.63 uM pixels) from the Bortle 6 - and if possible also from  darker sites.  It is then possible to combine everything together by using high dynamic range compostions of the overall luminance  and applying it to the colour data .

Taking this sort of approach  I have acquired quite detailed and extensive images of objects like M51  - and some interesting core detail on galaxies such as M63 and M94 even if the images do not show the full extent because all of the data were from Bortle 6.

I use a UV/IR cut filter but am not convinced that light pollution filters add anything.

One further fun thing to do from a Bortle 6 site though is also to take some HA data (for which light polution has little effect).  Like many others I add the HA data into the RGB data to emphasise the star -forming HA regions .  Particularly striking for galaxies like M51, M82 and M106 where the HA data really adds information - and indeed in the case of M106 even new galaxy 'arms'.  Another more ambitious one is to try using an SII filter to see HA in really distant galaxies.  A couple of galaxies in the COMA cluster are far enough away (350Mly) that the HA band is red shifted by ~ 16 nm .


So while living at a dark sky site would be ideal – overall still plenty of scope to do galaxy  imaging from more light polluted locations imo.


Tim

Thanks Yusra, Tim, and gfunk for you helpful replies. All very much appreciated.

and Tim, wow, there is gold in what you say and much food for thought.  Had never occurred to me for example how redshifting of distant galaxies might play into my poor filter choice selections. I have the asi294mm so could try bin1 lucky imaging… do you stack and process differently for this (more like in planetary imaging?)… otherwise can imagine my machine wouldn't cope with the sheer scale of data.

I'll also need to look at how within my processing software one can work the luminance to this extent and then apply colour. I own pixinsight but after several failed attempts (its like its from mars and im from venus or vice versa) continue to use APP, Startools, and PS… and all probably too superficially (eg I push buttons and pull levers til I see something I like rather than understanding why).  Sounds like whatever software I use I'll need to delve a bit deeper.

I‘m shooting from Bortle 6. I tried the Astronomik CLS for this and in my light pollution it does not help with galaxy images but just reduces the object signal and skews colors.

If you look at my images, I think you‘ll get a good idea how much integration time is needed. At Bortle 7, you need to multiply by 1.2-1.5x, approx. For something faint things like halos, low surface brightness galaxies or IFN, upwards of 20h is needed.

I find the biggest challenge to be large-scale chromatic noise. It‘s devilishly difficult to control when it overlays the galaxy. For the small-scale noise there are good tools, but against the large scale residues, the only thing that helps is to gather more light to average it out: Long integrations and high etendue. I feel my 10“ RC is still too small. This also limits the efficiency of LRGB because I can devote a maximum of 20-30% of the time to L and still control color noise. 

What you won‘t be fighting as much as in shorter FLs are the gradients: In the smaller FOV, they tend to be rather linear, especially if you shoot mono. I‘m experimenting with using wide-FOV shots for the gradient removal for the small-FOV ones. I’m not decided yet if it is worth the additional overhead. The new PixInsight LN and the PSG script both help a lot to reduce them, too.

thanks Torben for your insights. Love your gallery. That's the kind of thing I'm after and shows what is possible. Not sure my skill level is to the point I could discriminate large scale chromatic noise over small scale. A lot to learn. No doubt I am still bludgeoning issues with blunt tools that probably require a much more finessed approach. Your last point on FL helping to manage gradients makes intuitive sense and was one gain I could see from going longer.