The Quest for High Resolution Deep Sky Imaging Presentation

Just watched it a bit ago, it was excellent and covers all the major points. I’ve been pushing shorter subs , deep culling, over-sampling and +60 degree elevation imaging for a long time. It’s nice to see confirmation that I’m on the right track.

Bortle 6 Guy

Just because your sky is somewhat bright does not mean the seeing is bad. I get good seeing at SRO however the sky is only Bortle 4. So it is not as dark as possible.

Only collect imagery near zenith when there is no moon. Maybe a light pollution filter?

Bob

Robert Majewski · Feb 16, 2026, 09:58 PM

Bortle 6 Guy

Just because your sky is somewhat bright does not mean the seeing is bad. I get good seeing at SRO however the sky is only Bortle 4. So it is not as dark as possible.

Only collect imagery near zenith when there is no moon. Maybe a light pollution filter?

Bob

Hi Bob,

Thanks. I do image with Monochrome and LRGB Ha OIII SII filters.

Light pollution filter for B$W camera with filters?

See below my latest image fo Arp 210 or NGC 1569 with a total of 19 hours and 15 minutes.

I am imaging from 30 - 35° over horizon and the clear nights are sparse… So I should start later and fnish earlier.

I am at 1950m over sea level in center México, 22° North 101° West which we could call a semi-arid climate

https://en.wikipedia.org/wiki/Semi-arid_climate

What do you think. This is at 3200mm focal length (a DIY CDK Mewlon 250S (acqusition year 2006)) and a 0.245” arcseconds per pixel… Please ignore the chromatic aberration at the stars as it comes from the DIY flattener using Edmund Optics lenses) Have not yet found a script in PixInsight for killing it . Also seems to be that PI does not align correctly my separate LRGB images…

Thanks Rainer

Arp210_LRGB.jpg

John Stone · Feb 17, 2026 at 06:35 AM

As I understand things you take a block of 64 sub-exposures, toss 12, integrate the remaining 12 (maybe apply a bit of deconvolution) and call it a super-sub..   Repeat over and over then stack the super-subs in PixInsight.

So what happens as the seeing varies throughout the night?  Perhaps looking all the subs as a whole you would decide to keep all 64 subs in some blocks and maybe throw away the entirety of other blocks.

Have you considered keeping all your subs and then post-processing them all at once at the end of acquisition?  i.e. stack them all in PixInsight and use the Sub-Frame selector to toss your 1/2.

Also, in this case why not stack them all using PixInsight's weighed stacking scheme?  My understanding is that the result of this is nearly identical as what you get by manual culling.

Have you done an analysis about how much more noise you have in your final stacks by the higher contribution of read-noise from all those short exposures against the trade-off in integrating more seeing variations in longer subs followed by a deconvolution (like BlurX) to recover the lost detail?

Lastly, have you considered the possibility of moving your scope to a site with better/more-frequent good seeing?  A place like Howling Coyete Remote Observatories which frequently is sub-arcsecond?

John

I generally have the mount sit a one spot for 64 frames and then move to another “dithered” position. Since it takes the mount a few seconds to accurately settle to the new position, doing it on each exposure is not a good idea. You do not want to be exposing the next frame while the mount is still settling in. I rate the sharpness of the images in this block of images by computing the relative Strehl of the entire image (4K) and take 50%. ie 32 of them using my custom software to align them using a routine that uses cross correlation and phase correction of the entire image. I then integrate that set of 32 shifted images and save a “super sub” FITS image.

Yes, I have considered rating the entire data set to create maybe improved “super sub”s. However since the entire set includes dithered positions, which are farther from a chosen reference image might reduce the accuracy of alignment. I am not sure about this. I might want to give this a try. In principle I agree this is a good idea however the devil is in the details.

What’s good about the super subs is that I might only have 50 of these in PixInsight to work with.

No, I have not done that tradeoff because seeing conditions vary from night to night and it just takes a long time to collect the data. Also, since the program I use to collect the is my own, it is not very complex in terms of switching filters, targets etc. I only shoot filter and one target per night. I also do not like to go much past the meridian because of fear of a mount crash.

SRO is a very good site in terms of seeing. It hard for me to travel and any times one does, new problems pop up. As an example: I live in Las Vegas so I never had a dew problem. However when I had the rig moved to SRO dealing with dew was a big deal.

Bob

Robert Majewski · Feb 17, 2026, 05:12 PM

Rainer Ehlert · Feb 16, 2026 at 10:18 PM

Robert Majewski · Feb 16, 2026, 09:58 PM

Bortle 6 Guy

Just because your sky is somewhat bright does not mean the seeing is bad. I get good seeing at SRO however the sky is only Bortle 4. So it is not as dark as possible.

Only collect imagery near zenith when there is no moon. Maybe a light pollution filter?

Bob

Hi Bob,

Thanks. I do image with Monochrome and LRGB Ha OIII SII filters.

Light pollution filter for B$W camera with filters?

See below my latest image fo Arp 210 or NGC 1569 with a total of 19 hours and 15 minutes.

I am imaging from 30 - 35° over horizon and the clear nights are sparse… So I should start later and fnish earlier.

I am at 1950m over sea level in center México, 22° North 101° West which we could call a semi-arid climate

https://en.wikipedia.org/wiki/Semi-arid_climate

What do you think. This is at 3200mm focal length (a DIY CDK Mewlon 250S (acqusition year 2006)) and a 0.245” arcseconds per pixel… Please ignore the chromatic aberration at the stars as it comes from the DIY flattener using Edmund Optics lenses) Have not yet found a script in PixInsight for killing it . Also seems to be that PI does not align correctly my separate LRGB images…

Thanks Rainer

Arp210_LRGB.jpg

Rainer

Having vegetation on a flat plain could be good for seeing. However, I do not know for sure.

Bob

Thanks 👍️

The yellow pin with Observatorio NHSA is my location 🏁

📷 image.png

and a close up and yes my neighbours where so kind with the vegetation and panted some palm trees killing my whole southern hemisphere sky. And since 3-4 weeks the neighbour to the left illuminates his whole house and it looks like a Hotel on the Beach.

📷 image.png

Robert Majewski · Feb 17, 2026, 05:51 PM

Rainer

Have you consider placing your rig at remote site? I moved my rig from Las Vegas to SRO and it is much better there.

Bob

Thank you.

My financial possibilities do not allow me to do that and on the other side 50% of the fun is seeing my equipment from 2-3 meters away. Everytime I sit here it is realxinf until the moment comes when I start imaging 🤣 but I have to say that with patience and learning a bit more of the tricks in PI my images are now a bit better then before when I was only able to be here on the weekends every 2-3 weeks. I retired in November 2018 and since then it is a 24/7 job sitting in my cave

https://app.astrobin.com/u/NHSA_Observatory?i=309s3h#gallery

Joseph Biscoe IV · Feb 17, 2026 at 07:58 PM

Robert Majewski · Feb 16, 2026, 05:42 PM

I gave a talk about the methodology I use to obtain high resolution imagery of deep sky objects on The Astro Imaging Channel.

The link is https://www.youtube.com/watch?v=5gSn1NltNn4

Any comments?

I am working through it now. I am curious as an 8” RC owner. Why in the presentation do you say the RC needs a corrector because it uses two conical mirrors? Is this another term for Hyperbolic Mirrors that I was not aware of? To date the scope hasn’t needed a corrector at all either.

Joseph

The RC is fine on axis but off axis it has astigmatism and field curvature. So if you want to shoot targets and have sharp imagery across a wide field you need a field corrector.

Bob