NGC281 Pacman Nebula - Advice needed

[imakebeer]

I've been making steady progress in the last few months and made steady progress with M31, M33, M42 & M45. So I thought I'd have a go at the NGC281 Pacman Nebula. But I'm really struggling to get much out of it so far - I'm not sure if the problem is simply that it's a really faint target and I just need way more integration time, or if a standard DSLR won't cut it and I need to try a modified/dedicated camera and/or use HOO and/or SHO??? Or maybe the problem is simply light pollution??? I'm in a Bortle 5 - I set the scope up in the back garden away from the streetlights out front, but then my neighbour's landing light is shining down on the scope which really isn't ideal either!

SW150PDS + Nikon D5500

126 x 60s lights at ISO 400, plus 20 each darks and bias [EDIT: and 20 flats too!]

Processed in Siril then tweaked some more in GIMP. This version is heavily cropped because I wondered if the background extraction in Siril was causing issues, but I'm not sure that's the case.

Can anyone offer any pointers? Cheers.

Edited February 16, 2023 by imakebeer
Flats

[Vroobel]

Hi,

If it's standard body, I advise to send it to modification by removing filter from the front of sensor. Apart from that, definitely you should by a dualband filter, like Oprolong L-eXtreme (pr L-Ultimate, if yo can...) 2" and perform much, much longer subs, like 240s.

What software do you use for the post processing?

[almcl]

You don't mention flats in your calibration frames and there does appear to be some vignetting, so might be worth planning to take some of those?

Pacman is quite an Hα rich target so a modified DSLR (or an astrocam) might well pick up more.

Also, and I am not familiar with your camera, but ISO 400 and 60 seconds seems a bit short.  With my astro modded Canon (different camera, I know), I used ISO 1600 and 180 second lights on this target.

[WolfieGlos]

I imaged this one last year with a stock Canon 77D, and I got a very similar result (but using flats) and was disappointed. The stock camera is just not sensitive enough to the Ha in emission nebula.

I was also advised not to use a lenhance/extreme filter with a stock camera, only a modded camera. I haven't tried myself so others may shoot me down on this.....

[imakebeer]

1 hour ago, Vroobel said:

What software do you use for the post processing?

Thanks. As noted in the OP (😉) Siril and GIMP. The former is sometihng I've been recommended by others on here.
 

1 hour ago, almcl said:

You don't mention flats in your calibration frames and there does appear to be some vignetting, so might be worth planning to take some of those?

Pacman is quite an Hα rich target so a modified DSLR (or an astrocam) might well pick up more.

Also, and I am not familiar with your camera, but ISO 400 and 60 seconds seems a bit short.  With my astro modded Canon (different camera, I know), I used ISO 1600 and 180 second lights on this target.

Sorry, I forgot to mentioned flats in the OP, now corrected - I did 20 flats too. I seem to get diferent areas of lighter/darker "haze" across my images, I suspect there is a lighter patch just out of the frame which might explain what looks like vignetting. I can see vignetting in the original image but then this version has been massively cropped way beyond this.

The reasoning behind doing 60s at ISO 400 is this - from what others on here have told me, it's the total integration time that matters most. So whether you do 100x60s or 60x100s or 20x300s it doesn't really matter. Also, since I've not sorted out auto-guiding yet I find if I go above 60s the star trailing becomes unacceptable. Plus, when a satellite goes across your FOV it only ruins one of many shorter frames.

Re. ISO, my D5500 is ISO invariant so if I go above ISO 400 it just introduces more noise. Looking at the histogram it seems like there's enough wiggle room there - it's not like the peak is hard up against either end of the scale.

Which is not to say it wouldn't be worth trying different camera settings, just when I've tried it on M33 for example it didn't make any noticeable difference.

 

56 minutes ago, WolfieGlos said:

I imaged this one last year with a stock Canon 77D, and I got a very similar result (but using flats) and was disappointed. The stock camera is just not sensitive enough to the Ha in emission nebula.

Hmm... OK, interesting... As noted I did flats, just forgot to mention it in the OP. So maybe in part the issue is a stock D5500 just isn't up to the job for this target??? 🤔

[wesdon1]

1 hour ago, imakebeer said:

I've been making steady progress in the last few months and made steady progress with M31, M33, M42 & M45. So I thought I'd have a go at the NGC281 Pacman Nebula. But I'm really struggling to get much out of it so far - I'm not sure if the problem is simply that it's a really faint target and I just need way more integration time, or if a standard DSLR won't cut it and I need to try a modified/dedicated camera and/or use HOO and/or SHO??? Or maybe the problem is simply light pollution??? I'm in a Bortle 5 - I set the scope up in the back garden away from the streetlights out front, but then my neighbour's landing light is shining down on the scope which really isn't ideal either!

SW150PDS + Nikon D5500

126 x 60s lights at ISO 400, plus 20 each darks and bias [EDIT: and 20 flats too!]

Processed in Siril then tweaked some more in GIMP. This version is heavily cropped because I wondered if the background extraction in Siril was causing issues, but I'm not sure that's the case.

Can anyone offer any pointers? Cheers.

Hi old friend! Mate, I have found that whenever i'm imaging a target that has my scope pointing anywhere near the bright LED street lights outside my building, it plays havoc with my subs and final stacked image. The amount of artefacts, is unreal. I would also suggest, maybe do slightly longer subs, say 120-180 seconds, if your guiding accuracy will allow it? Finally, another member mentions cranking up your ISO from 400 to 1600? In my limited experience, shooting higher ISO's when there's substantial light pollution, that isn't properly dealt with by means of suitable LP filter/NB filter, will likely only make your subs blown out from excessive LP? 

Hope you find a solution!

Wes

[Vroobel]

I can also advise the Affinity Photo which is cheap and works with wonderful plugins: GradientXterminator, NoiseXterminator and StarXterminator. They cost together more than the software, but they do awesome job. The Affinity Photo removes all satellite or airplane trails. Recently I performed only 14 subs 600s each, some contain the trails, but I cannot see any in the final picture. 

It's not always true that the total integration time matters if you have lot of short subs. 

[WolfieGlos]

14 hours ago, imakebeer said:

The reasoning behind doing 60s at ISO 400 is this - from what others on here have told me, it's the total integration time that matters most. So whether you do 100x60s or 60x100s or 20x300s it doesn't really matter. Also, since I've not sorted out auto-guiding yet I find if I go above 60s the star trailing becomes unacceptable. Plus, when a satellite goes across your FOV it only ruins one of many shorter frames.

Hmm... OK, interesting... As noted I did flats, just forgot to mention it in the OP. So maybe in part the issue is a stock D5500 just isn't up to the job for this target??? 🤔

Regarding satellites or aircraft, that's interesting as I've never scrapped a frame with them passing overhead! I don't understand the whole process about how they get removed in stacking (I should probably read up on it a bit more!), but I never get any trails left in my images. The only ones I do scrap, is when you get the large plume left behind a plane that slowly dissipates, essentially like a cloud. I also use Siril and Gimp.

Regarding the camera, it's not that it's not up to the job, it's just not very sensitive to those wavelengths of light so will require a lot more time to bring it out. It'll be fine for galaxies and reflection nebula, less so for emission nebula. I didn't want to modify my camera, so I got mine from astronomiser. This page on his website http://www.astronomiser.co.uk/intro.htm shows a graph, and for the camera in question (a Canon 350), it is only circa 20% sensitive to that light. By modding the camera, it becomes circa 90% sensitive.

This was mine with a stock Canon 77D, Evostar 72ED, ISO 1600, 67 x 120s (bortle 4).

14 hours ago, Vroobel said:

It's not always true that the total integration time matters if you have lot of short subs. 

Really? How does this work, as I thought more time = more signal and less noise? Do you have to increase the ISO to do this? I guess the DSLR shutter count would go through the roof 🤣

[Elp]

Regarding exposure time think about it logically. Your camera sensor is just sitting there waiting for photons to hit each pixel. Photons are likely to hit a pixel whether you expose for 1 minute or 10 minutes, the difference only being if you expose a single image for longer a pixel which receives more photons than a shorter exposure image will appear "brighter" or appear to have more signal because the pixels full well has been filled up more (pixel has been stimulated more by received signal). This is also the reason why sometimes a larger pixel sensor will be of benefit if your scope is suited to it, a larger pixel has greater probability of photons hitting it so signal will appear to be stronger than with a smaller pixel sensor.

You can account for signal accumulation with shorter subs by taking a factor more of them in total, by doing this such things like guiding errors affecting subs will be likely be reduced in percentage as you have more images to work with. The only issue as far as I can see for taking lots of shorter images is disk space.

Edited February 17, 2023 by Elp

[imakebeer]

9 hours ago, WolfieGlos said:

Regarding the camera, it's not that it's not up to the job, it's just not very sensitive to those wavelengths of light so will require a lot more time to bring it out. It'll be fine for galaxies and reflection nebula, less so for emission nebula. I didn't want to modify my camera, so I got mine from astronomiser. This page on his website http://www.astronomiser.co.uk/intro.htm shows a graph, and for the camera in question (a Canon 350), it is only circa 20% sensitive to that light. By modding the camera, it becomes circa 90% sensitive.

Thanks, I think that makes sense and I had a feeling this might be part of my issue. I need to understand better the different types of nebula.

I also have a Canon 450D I bought cheap off eBay last year when I was getting into AP - I might consider modifying it myself or change it for a modified camera. Not sure I want to do this with the D5500 as it's my main family camera. Main issue with the Canon is being older I think it has a slower processor as it seems to take much longer than the Nikon to take and send and image to the laptop.

8 hours ago, Elp said:

The only issue as far as I can see for taking lots of shorter images is disk space.

Meh... storage is cheap 😉

On 16/02/2023 at 17:37, Vroobel said:

If it's standard body, I advise to send it to modification by removing filter from the front of sensor. Apart from that, definitely you should by a dualband filter, like Oprolong L-eXtreme (pr L-Ultimate, if yo can...) 2"

I'm intrigued by these filters. They're not cheap - a 2" L-eXtreme is a significant fraction of the cost of my OTA alone. Not that I started this hobby to save money, but still...

The blurb about these filters on the FLO websites implies I can just bung one in - I assume inbetween the focuser and DSLR? - and all will be well with the world. Is it really that simple?

[WolfieGlos]

17 hours ago, imakebeer said:

I also have a Canon 450D I bought cheap off eBay last year when I was getting into AP - I might consider modifying it myself or change it for a modified camera. Not sure I want to do this with the D5500 as it's my main family camera. Main issue with the Canon is being older I think it has a slower processor as it seems to take much longer than the Nikon to take and send and image to the laptop.

Yeah, that’s the same reason I didn’t want to modify my 77D so I bought a modded 800d. It made sense for me as both cameras have the same digic 7 processor, the same battery (important!) and sensor size. I believe if you modify it yourself and it goes wrong the camera cannot be fixed, I didn’t want to risk it but others have done it successfully 🙂

[imakebeer]

Hey @Elp , can I pick your brains on this a little more...

Up the top of this thread @Vroobel has suggested a couple of Optolong filters - they're not cheap but, well, ok...

In the course of my googling I found this thread from ~9 months ago in which you said you'd tested the L-Pro, L-Enhance & L-extreme...

In your experience do you think any of these would offer an improvement to my Pacman based troubles? 

[Elp]

Lpro, no as it's a general purpose light pollution filter. Lenhance (triband, but really dual band as it's hbeta transmission pass isn't really much use) and lextreme (dual band) maybe.

But I think your main issue is your unmodified camera, I've just tried to find a few quantum efficiency graphs for your camera, didn't find one specific to your model but some of the Nikon's (most other DSLRs/mirrorless don't fare better as they're designed primarily for daytime use) showed roughly 40-50pc transmission at 656nm which is HA, and O3 fared a little better as theres a combination of blue and green getting qe signal at around 500nm. The stage 1 astro mod for DSLRs involves removal of the IR filter which will open up the ha/red wavelength response. If however you look at qe graphs of colour astro cameras you'll see they are far more sensitive sometimes in the region of 70-90pc in HA, so you can imagine how much more nebulosity they'll capture. As red is the predominant wavelength of nebulae in RGB colour space this is where you'll see the most benefit (if you look at RGB channels taken via an lenhance or extreme you'll see the red has vastly more signal than the green or blue when imaging most emission nebulae, there are a few exceptions). So maybe you need a modded DSLR for improvement, but I'd go for an astro camera, you can image with uncooled ones too as I do most of the time, if you can afford cooled go cooled.

One thing I've noticed about all the optolongs mentioned, any stray light hitting the objective or any haze from light glow around a target and your flats calibration will be an issue, those filters cast a strange red/green pattern across images due to this, I've experienced it with all my telescope setups and camera lenses and different cameras. You can somewhat mitigate this problem via synthetic flats and limiting your background level push, but the latter is what you want to do to reveal that nebulosity.

Maybe worth a shot with the filters if you can find one cheap, you'll likely keep the lenhance or lextreme for the occasions you image certain ha and o3 rich emission nebulae as they do work wonders when they work, there are obviously other manufacturer ones out there, Antlia being another good one.

Another thing you can try without spending a penny is to use starnet if you haven't to create a starless image. You can push starless far greater in something like GIMP and then recombine with the smaller stars.

Edited February 18, 2023 by Elp

[Vroobel]

@imakebeer, you can find ypur Nikon D5500 here:

https://astrophotography.app/nikon.php

and compare it to i.e. my Canon 6D. The Nikon D5500 has QE = 58%, but smaller pixer, around 4um, while the Canon 6D has QE = 49(50)%, but its pixel has ~3x bigger surface.

I use modified C6D, which is one of the best cheap full frame OSC. Even if I buy a cooled mono camera with all stuff arround, I will always keep the C6D with the L-eXtreme.

If you are going to buy a dual-band filter, I would suggest to not compromise by buying some cheap replacement. You can consider also IDAS products or other, even more expensive, but remeber: "who buys cheap, buys twice".

BTW, I experience a greenish cast each time (easy removable by levels), but never any patterns. 

 

[imakebeer]

@Elp @Vroobel thanks for the advice and suggestions, much appreciated 👍

I'm gonna take a look into modified DSLRs, starting potententially with modifying my 450D myself - need to check out the tutorials for this on YouTube. I'll also take a look at StarNet as this seems like it could be useful more generally anyway from my initial look.

[Vroobel]

Be careful with the modifications. It's easy to damage the camera and even if not damaged, you can face a tilt caused by the sensor screwed wrong. You will have a permanent coma on one side or corner.