NGC7000 , my first DSO attempt

[MHaneferd]

@Lee_P The satelite photo was taken 21:38 (CET)

By the way, Here in Oslo, it is not the light pollution that bother me most, it is rather the weather 😳

 

[MHaneferd]

16 minutes ago, Lee_P said:

FYI your 120s sub looks better than my 120s sub

Not sure about that. You have the little galaxy upper left, but my focal length are clipping it 😳

[iantaylor2uk]

I generally use unity gain on my ZWO 071MC Pro camera (which is 90 on my camera) and usually image NGC 7000 at 1 min subs if just using a uv/ir filter but I go up to 3 mins if I’m  using an L-enhance filter. I also usually just cool the sensor to 0 or -5 C.
 

[MHaneferd]

@iantaylor2uk interesting. What is the reason for extending the sub time depending on filter type?

[iantaylor2uk]

The L-enhance filter only lets through H alpha, H beta and OIII wavelengths so greatly improves the signal to noise ratio of emission nebula (and is also great at reducing light pollution) but the overall signal is lower so you have to go to longer subs. There is a spreadsheet for ZWO cameras where you can put in details of your telescope (aperture and focal length), sky brightness level, and it tells you min exposure times, also for different filters - you need around 3x longer subs for the L-enhance filter compared to no filter.

[Lee_P]

11 minutes ago, iantaylor2uk said:

The L-enhance filter only lets through H alpha, H beta and OIII wavelengths so greatly improves the signal to noise ratio of emission nebula (and is also great at reducing light pollution) but the overall signal is lower so you have to go to longer subs. There is a spreadsheet for ZWO cameras where you can put in details of your telescope (aperture and focal length), sky brightness level, and it tells you min exposure times, also for different filters - you need around 3x longer subs for the L-enhance filter compared to no filter.

I'd be interested in a link to that spreadsheet, if you have it handy. I've always shot 120s regardless of filter (i.e. no filter or L-eXtreme) and I've gotten good results -- happy to change my method if another one is more effective! Not quite sure of the logic though, as you get good signal from a high total integration time rather than individual subs, right? I'm tired, my brain might not be working properly...

Edited September 25, 2021 by Lee_P

[MHaneferd]

18 minutes ago, iantaylor2uk said:

There is a spreadsheet for ZWO cameras

Thanks.. after hefty googeling, I found these ones for my ASI2600, 

https://drive.google.com/drive/u/0/mobile/folders/1CnKISk-4X0pTo6emReMIO099JlgPHIT6?usp=sharing

 

[iantaylor2uk]

I downloaded the spreadsheet from the link below:
https://www.cloudynights.com/topic/753380-gain-and-sub-exposure-calculator-spreadsheet-for-zwo-cameras/

[iantaylor2uk]

I should have added, as it isn’t obvious, but if you aren’t using a filter, or just using a uv/ir filter choose green from the filter list.

[Lee_P]

Thanks! I think that my background sky level is so high that 120s for L-eXtreme is way more than three times the optimal exposure when not using any filter anyway, and the limiting factor is storage space and PC processing power. So in effect my current method (120s for everything) is still valid for me given my local sky conditions. If I had much darker skies then it would be a different story.   

[iantaylor2uk]

If your background light pollution is high, you should be using shorter subs. I have a 7 year old i3 dual core desktop and I can still  stack 400 subs in deep sky stacker in 30-45 mins or so.

[Lee_P]

1 hour ago, iantaylor2uk said:

If your background light pollution is high, you should be using shorter subs. I have a 7 year old i3 dual core desktop and I can still  stack 400 subs in deep sky stacker in 30-45 mins or so.

I think we need to be careful about the phrasing here: it would be accurate to say that I could use subs shorter than 120-seconds, but not necessarily that I should.

Keep in mind that total integration time is the most important factor. Especially if you’ve got badly light-polluted skies, you really need to tease out that signal from the noise. I aim for 20 hours per target.

The background sky level has a huge influence on the “optimal exposure” length, which is about 10 seconds for me using an L-eXtreme. That would be 7200 subs. 7200! I could do that but don’t think that I should.

So, I shoot 120-second subs. One 120-second sub is equivalent to 12 10-second ones. And with 120-second subs, I now need 600 of them to make it to 20 hours. Still a lot, but manageable. 120 seconds isn’t some magic number reached at through calculations, but rather a reasonable compromise for me to get the benefits of shooting short subs without filling my harddrive or melting my PC. FYI I have a modern computer and 48GB of RAM. I integrate using PixInsight, which seems to be more resource-intensive than DSS. I generally leave it integrating overnight.

When I got my camera I found Dr Glover’s presentation about CMOS exposure times, and ran through the calculations – it’s only just occurred to me that that’s what the spreadsheet is doing.

I think that we’re both right in our approaches: you trebling your sub length with an L-eXtreme is likely the right thing for you to do as you’ve got darker skies so have got more headroom, so to speak. @MHaneferd is in Bortle 8/9, similar to me, so will likely find that the same kinds of settings as I use will work well.

That’s my understanding of it all anyway 😊

 

[iantaylor2uk]

Yes - I agree with you that these recommendations are a minimum and you can go longer if you wish - the whole point however is that there is not a lot to be gained by going longer (apart from having less subs). Shorter subs means that if you have to lose any (due to satellites or bad guiding) there is less of an overhead. 

I think you misunderstand though the effects of light polluted and dark skies - if you are in a light polluted area the spreadsheet (and Dr Glover's presentation) would guide you to shorter subs, whereas if you are fortunate enough to be in dark skies, you can use much longer subs. 

There are plenty of people around who use short subs (and thousands of frames) - I saw a good M57 where the subs were only 1 sec each: https://www.astrobin.com/345864/?image_list_page=2&end_date=2020-05-02&nc=AnonymousUser&page=3

I used to have a 12" f/4 Newtonian, and I got a good image of M51 using just 10 second exposures (I live on the outskirts of Chester where it is a Bortle 6 sky) - the photo below is a stack of just 180 subs (only 30 mins integration time) - I didn't use a coma corrector so some of the stars at the edges won't look too good: https://photos.app.goo.gl/eDNKesm8ZSNoLCPX9

[Lee_P]

3 hours ago, iantaylor2uk said:

Yes - I agree with you that these recommendations are a minimum and you can go longer if you wish - the whole point however is that there is not a lot to be gained by going longer (apart from having less subs). Shorter subs means that if you have to lose any (due to satellites or bad guiding) there is less of an overhead. 

There is a lot to be gained by going longer if the recommended sub length is just a few seconds, but you want to get a total integration time in the tens of hours -- see my comments later about file sizes. I understand the benefits of short subs; I've written about that here. By the way, don't worry about losing subs due to satellite trails, stacking algorithms remove the trails efficiently.

 

3 hours ago, iantaylor2uk said:

I think you misunderstand though the effects of light polluted and dark skies - if you are in a light polluted area the spreadsheet (and Dr Glover's presentation) would guide you to shorter subs, whereas if you are fortunate enough to be in dark skies, you can use much longer subs. 

I'm not misunderstanding this. If you re-read my previous comment, you'll see I say "The background sky level has a huge influence on the 'optimal exposure' length, which is about 10 seconds for me using an L-eXtreme" and "you trebling your sub length with an L-eXtreme is likely the right thing for you to do as you’ve got darker skies so have got more headroom, so to speak." I'm in agreement with you! 

 

3 hours ago, iantaylor2uk said:

There are plenty of people around who use short subs (and thousands of frames) - I saw a good M57 where the subs were only 1 sec each: https://www.astrobin.com/345864/?image_list_page=2&end_date=2020-05-02&nc=AnonymousUser&page=3

I used to have a 12" f/4 Newtonian, and I got a good image of M51 using just 10 second exposures (I live on the outskirts of Chester where it is a Bortle 6 sky) - the photo below is a stack of just 180 subs (only 30 mins integration time) - I didn't use a coma corrector so some of the stars at the edges won't look too good: https://photos.app.goo.gl/eDNKesm8ZSNoLCPX9

Sure, I've seen many images like this. They're excellent proofs of concept. But again I reiterate what I've said previously, which is that from heavily light-polluted skies you benefit a lot from long total integration times. Let's say you want to get a total integration of 20 hours just from 1 second subs. That's 72,000 images. Each image from my 2600MC is 50MB. That's about 3.4TB of harddrive storage. Just for the raw subs. OR shoot 120-seconds and then we're talking under 30GB -- and you're still getting most of the benefits you get with shooting ultra-short subs. If I had dark skies and so could get away with lower total integration times, I might lower my sub length, and perhaps get into the realm of "lucky imaging". The same for if I had no restrictions on harddrive space or processing power. But I don't, and find that 120-seconds is a sweet spot for me. That's the sub length I used for most of these shots.

Edited September 26, 2021 by Lee_P

[Ouroboros]

21 hours ago, MHaneferd said:

@Ouroboros Is there a formula for exposure times? Or is it in the video you mentioned? Link?

I will try out a lot of exposures at gain100 @2min as described by @Lee_P. I just hope that the PC dont melt..

I was thinking of either give the NGC7000 another shot, by moving to the cygnus wall, or try the pin-wheel galaxy.

All credits to @StuartT for giving me these coordinates to the wall;

RA 21h 0m 2s
Dec +43d 41m 42s
 

Yeah the formula is about 49 minutes into Robin Glover’s video.  You really need to watch the vid to get the context though. 

[MHaneferd]

Thanks @Ouroboros. I just listened to Dr.Robin Glovers video as a podcast while painting my cabin today. I defiantly have to look at it on YouTube. I am not sure if a formula will help getting the exact sub-shot-time, since when I look at my sky. It seems to be more light polluted in the southern direction than the northern.. And, I also feel that there is a ratio on how many files my computer can handle, compared to noise / exposure time ratio.. So, I will try to follow @Lee_P's two minute exposure times. I will test both lower and higher to see the results...

Anyways, the Andromeda shots last night, got a bit blurred due to clouds. So after removing the most blurred, and the ones with satellite tracks (6 of them), I ended up with 49 frames out of 100. all at 2 minutes subs. I got the blacks over the night, but when I was going to take the Flats. The sky where quite moisture and wet, so I gave that up. 

So this is the result of 49 frames at 2 minutes subs, with gain 100 + 100 darks. Stacked with ZWO ASIDeepStack_1.5 . As you can see, there are some dust spots. The result has not been edited in any ways, it is just the JPEG from the ASI Deepstacker:

[Lee_P]

@MHaneferd a great result, especially considering it's just the raw stack! (Also, good idea deciding to test various exposure times for yourself and seeing what works best for you!)

[MHaneferd]

@Lee_P, I think it can be better. It’s what you are saying the longer integration time, the  better signal to noise ratio can be achived. As I had to scrap 50% of the frames, and I suspect the remaining percentage also have som dizzy cloud in them, I will test another go for two minute subs over several hours. Just have to wait for clear sky again.. going to rain for 14 days now 😳

[Lee_P]

11 minutes ago, MHaneferd said:

@Lee_P, I think it can be better. It’s what you are saying the longer integration time, the  better signal to noise ratio can be achived. As I had to scrap 50% of the frames, and I suspect the remaining percentage also have som dizzy cloud in them, I will test another go for two minute subs over several hours. Just have to wait for clear sky again.. going to rain for 14 days now 😳

You can always add "just one more night" of data into the stack..! I'm currently imaging M31 as well, and am at about 20 hours. Nothing but rain forecast now, too

 

[MHaneferd]

1 hour ago, Lee_P said:

just one more night

But, if you do several nights.. How do you manage flats? Do you take a set of flats and darks after each night, or do you do that at the end? 

[Lee_P]

19 hours ago, MHaneferd said:

But, if you do several nights.. How do you manage flats? Do you take a set of flats and darks after each night, or do you do that at the end? 

Darks: I always shoot 120-second subs at -10 deg C, and have a Darks library matching that. So I use the same Darks for every imaging project.
Flats: I take one set of 20 Flats per imaging project (i.e. the same set of Flats covers approx. 600 subs / 20 hours integration time). I've got a DIY Flats panel I use for this: http://urbanastrophotography.com/index.php/2021/09/27/diy-flats-panel/ I take these 20 Flats altogether whenever is convenient; normally before an imaging run one night. This reminds me that I haven't taken the Flats for my current M31 project yet, so I should do that soon!

 

Edited September 27, 2021 by Lee_P

[MHaneferd]

@Lee_P But, if you use flats for severa hours of imaging, the dust particles should stay at same spot over all 20 hours? I am just thinking that if I move the scope in and out for two sessions on the same M31.. the dustparticles will probably move. And how will the flats be intrepeted then in the stacking??

[Lee_P]

2 hours ago, MHaneferd said:

@Lee_P But, if you use flats for severa hours of imaging, the dust particles should stay at same spot over all 20 hours? I am just thinking that if I move the scope in and out for two sessions on the same M31.. the dustparticles will probably move. And how will the flats be intrepeted then in the stacking??

My telescope is a "closed system" so to speak -- easier being a refractor -- so it's not that easy for dust to get in. Also, I don't change anything (like swapping filters, removing or rotating the camera) during a project. So far I've used this Flats method for about a dozen long-integration images, each of them involving data acquisition over many many nights, and have never had any issues. Try it and see if it works for you too?

[StuartT]

this is turning out to be a very interesting and informative discussion. I'm learning a lot!

Just in praise of shorter subs... as long as you have a decent CMOS camera I am finding that shorter subs work just fine. In my case 35s sec being the limit of my mount (because I basically CBA with guiding).

[bobro]

It is a very interesting discussion that I have been following over the days, though some points may be worth adding to the discussion:

> When captured photons are reduced camera read noise becomes a bigger part of the overall SNR picture - that's why recommended minimum sub exposure time increases when filters are added. Filters cut down on the signal, whereas read noise is constant for a given camera setting. Increasing sub exposure time reduces the overall read noise (less subs) for a given total imaging time. However, there is no compulsion to go beyond a sub exposure time where read noise becomes low enough to be ignored, unless there are mitigating reasons such as too many subs required due to very long overall image exposure time. I've been experimenting with this on my setup using Robin Glover's calculations, finding that (not surprisingly) read noise shows when an Ha filter is used and also to a lesser extent with 30 sec subs. To keep things simple I've settled on 2-3 min subs as appropriate for my setup as guiding generally works fine at this period and few subs need be scrapped.  Also helps with my non-cooled camera as it reduces the variations required in a darks library.

> Plane or satellite trails are readily eliminated with kappa-sigma stacking, allowing these subs to contribute to the image.

> Dust bunnies don't move around much, allowing flats to be used for a while.

> Unless imaging is narrowband, modern cmos sensor dark current is so low that cooling isn't required in order to reduce camera noise - sky noise is usually the much larger source of noise. However, cooling (or rather fixed temperature operation) helps with providing accurately matched darks.

I hope @MHaneferddoesn't mind, but I've gone back to the start of this discussion and reduced stars on the NGC7000 image using Starnet++, increased saturation and blurred the background to reduce the noise (hence the mottling) with a quick reprocess. It shows some interesting nebulosity, though still needs some work, especially on the brighter stars. Wish I has done as well with my first images!