Screen Grab of M27 in OIII 120s Single Sub @ F/6.8 - An Argument Against Longer Subs With CMOS Cameras

Incredible for a 120s sub with an NB filter...  Hard to argue against a picture like that :)

Thanks for sharing that Steve. This post illustrates why I all but gave up at home and went remote with Roboscopes. It's not the telescopes we have access to alone, it's the sky. It is absolutely shocking how much of a difference good skies make to integration and SNR.

I recently took some frames on P3 of NGC6559 and my Lum subs are actually slightly blown out at 120 seconds - 60 would have been better. it's difficult to describe just how fast F3.3 / 3.6 is under dark skies is - and I've been learning, quickly, that shorter is better.

Most piers will run unguided at 120seconds or less, which means the dithers are so much quicker. In addition, you get more total dithers which is absolute gold to SNR. More dither, more signal.

The noise of modern CMOS sensors are so low compared to CCD we just do not need the longer subs we used to.

As an example, take M51 - a very popular target and lets use Steve's ODK 16 he's used on M27 above. I have attached a picture of the angular size (green circle) used to calculate the surface brightness of 22.9 mag/arcsec2.

Using Steve's ODK16 at home in Bortle 7 ish skies, (without taking in to account dithers) and using 120s sub frames I would need to take 1,217 images or integrate for 40.5 hours to get an SNR of 20.

To get the same SNR at Roboscopes in average conditions (not peak), I'd need 14 hours of subs or 426 x 120s frames. Staggering difference.

Shooting at 300 seconds instead of 120 seconds does a few things. Using the same equipment, target and sky (Roboscopes) above and just changing the sub length means we get slightly more signal and only need 13.92 hours of data (instead of 14.17) to hit our SNR. So technically we are better off with long subs - but in the real world that is simply not true.

By the time you take in to account the hundreds of extra dithers we have been able to get and the fact that we have lost less frames due to clouds nuking 2mins of data at a time rather than 5mins at a time, you end up with a better image. Plus - it will be unguided so dithers recover much much quicker.

Sorry for the word salad, I typed this up quickly.

I don't like that my italics are in red - it wasn't supposed to be that emphasised and in-your-face!!

Deep Sky Astrophotography With CMOS Cameras by Dr Robin Glover

This is a great video about exposure times.

Deep Sky Astrophotography With CMOS Cameras by Dr Robin Glover

This is a great video about exposure times.

 

Thanks for sharing, I have not seen this before.

Its an awesome video and I have to say that he's a very good at making the information easy to assimilate, even for a dullard like me :)

Steve

That video is awesome, I've watched it before and It taught me that from the UK I should be doing 7 second exposures lol. I think the main point here is that unless you are lucky imaging at 1s or less it's all about compromise. No one is going to realistically do 0.3s from bortle 9.

The Optimum examples he gives are not all realistic for real world use.  When we are are talking about CMOS and bortle 1 the noise is soooooo low it almost doesn't matter what you do, you are going to get a stonking image with enough total exposure. The benefits of shorter subs such as more dithers, less risk and no guiding outweigh any marginal gains in my view.

If anyone has time to take a look at these sheets I'd appreciate it... I'm working away! https://drive.google.com/drive/u/0/mobile/folders/1EnRw69iA5_kTBxObnU9CKj5GilnZW8VJ?pli=1

That video is awesome, I've watched it before and It taught me that from the UK I should be doing 7 second exposures lol. I think the main point here is that unless you are lucky imaging at 1s or less it's all about compromise. No one is going to realistically do 0.3s from bortle 9.

The Optimum examples he gives are not all realistic for real world use.  When we are are talking about CMOS and bortle 1 the noise is soooooo low it almost doesn't matter what you do, you are going to get a stonking image with enough total exposure. The benefits of shorter subs such as more dithers, less risk and no guiding outweigh any marginal gains in my view.

If anyone has time to take a look at these sheets I'd appreciate it... I'm working away! https://drive.google.com/drive/u/0/mobile/folders/1EnRw69iA5_kTBxObnU9CKj5GilnZW8VJ?pli=1

https://www.youtube.com/watch?v=rQAXVwE_nvY

Agreed

Here is another Youtube video based on the above mentioned maths, I do enjoy how Cuiv bends the hard and fast math to suit his real world applications and constraints like storage and processing power not to mention the KISS rule he seems to enjoy like myself :)

Manuel is doing some "Sciencey" stuff with our data at present so I have been told, I am really interested to see what he comes up with if I am honest. Although it will no doubt mostly be way above my simple brain!

Steve

And we must not forget about gain

Choosing the right gain for Deep Sky imaging with CMOS cameras

And I the QHY268c, a camera that has the most confusing gain settings. If I set the camera in high gain mode with 2CMS at 57 the exposure calculator in NINA will suggest an exposure time of around 180s while using my ALP-T narrow band filter. Then if I set the gain at 90 NINA will suggest a time around 900 seconds. After a bunch of experimenting a gain of 67 and exposures times between 120 and 180 seconds work well. With my AM5 mount I don't like going beyond 180 seconds. The AM5 is an OK mount, but it is a bit sad that my little iOptron CEM26 tracks way better then a mount that cost twice as much.

And we must not forget about gain

Choosing the right gain for Deep Sky imaging with CMOS cameras

And I the QHY268c, a camera that has the most confusing gain settings. If I set the camera in high gain mode with 2CMS at 57 the exposure calculator in NINA will suggest an exposure time of around 180s while using my ALP-T narrow band filter. Then if I set the gain at 90 NINA will suggest a time around 900 seconds. After a bunch of experimenting a gain of 67 and exposures times between 120 and 180 seconds work well. With my AM5 mount I don't like going beyond 180 seconds. The AM5 is an OK mount, but it is a bit sad that my little iOptron CEM26 tracks way better then a mount that cost twice as much.

Good morning :)

Gain we are looking at I assure you, when finished we hopoe to have a new page for the website with lots of useful things to help with planning your imaging run with much less wasted time and better results

https://www.cloudynights.com/topic/877919-wd-17s-harmonic-equatorial-mounts-servo-motor-direct-drive/#entry12787974

I have one of these coming from the manufacturer to test, if I am happy I have the option to become the EU distributor. If you look at the technical details it should be at least 100% better than the AM5 ever will be

I am quite excited about this little mount :)

Steve

Very interesting, specs on this mount look great...  Compatibility with Asiair is also a major plus...  Cuiv's review will probably make or break the sales of this mount in the west, if its positive I would jump on that EU distributor option...