Back to forums

Rescue mission: stacking of NGC 5128

Processing Deep Sky NGC5128 Help Needed NGC Objects
7 replies•226 views
Daniel Petzen avatar
Daniel Petzen avatar
I'm at my wits' end and would like to ask for some help to save the data from a session imaging NGC 5128 that I thought would be an absolutely stunning image.

I'm new to big telescopes, stacking and post-processing. I have  Celestron 9.25 with a ASI294MC and I'm just about got a grasp of the basic, starting to get some decent images. I normally take my subs at 200 gain with 4 or 2 minute exposures.

I had very clear night and NGC 5128 was very high in the sky. I saw its visual magnitude was 6.8 and was a bit worried I would over-expose my subs, so decided to drop the gain to 120 and the exposure to 2 minutes.

The subs started rolling in and looked absolutely great in Blink, so I thought 'Yes! Spot on. IOTD here I come!'.

I got three hours of integration time. Subframe selector told me I may want to drop 7 of them, so they were all quite high quality.

I enthusiastically tried to stack my subs in PixInsight, which rejected two frames and then got stuck on plate solving and eventually failed to integrate. PI has had loads of hissy fits before, so I already have Astro Pixel Processor for stacking. APP warned about the corrupt data, but soldiered through an produced a discolored horrible end result. I also tried DSS, but it produced an unusable green blob.

I tried to process the APP image, but the colours were all wrong (despite SPCC) and it just looked terrible.

The conclusion may be that I have a bunch of under-exposed subs that can't be stacked because they are, well, rubbish, but the thing is that each sub looks absolutely amazing. I actually process a single two minute sub and it came out as a pretty impressive image. A two minute sub!

Sorry about the long story. If you made it this far, then perhaps you have the tenacity to help me out?

I know I've not provided much solid facts, so please let me know what information you need. I've zipped up all the data (lights, flats, dark flats and darks). They can be found here: https://drive.google.com/file/d/1Qk6n03Y2itL__uz2pLButyQPup4yOU99/view?usp=drive_link (please let me know if you can't access them).

I know there is an amazing image in there, so I just refuse to let it go. Any help would be much appreciated.

Thanks in advance,
Dan
Helpful Engaging
Quinn Groessl avatar
Quinn Groessl avatar
Somebody else will try before I get done with work to try and stack them. Just a note though, in WBPP in the lights tab I’ve unchecked the astrometric solution one. For basically the past year it’s been hit or miss (mostly miss) whether it worked during WBPP for me. Without seeing your logs it’s hard to say what went wrong, but I would disable that and try again. If everything’s still cached it should go pretty quick.
Helpful
Daniel Petzen avatar
Daniel Petzen avatar
Quinn Groessl:
Somebody else will try before I get done with work to try and stack them. Just a note though, in WBPP in the lights tab I’ve unchecked the astrometric solution one. For basically the past year it’s been hit or miss (mostly miss) whether it worked during WBPP for me. Without seeing your logs it’s hard to say what went wrong, but I would disable that and try again. If everything’s still cached it should go pretty quick.

Thanks for the tip on how to disable plate solving! I've googled this loads of time and never found out how to disable it. Perhaps I'm just lousy at  googling!

I re-ran the stacking in PI. I use the old output directory, but it doesn't seem to have cached anything, so it took quite a while.

It's getting very late here now, but I stuck around for one more try to stack the subs. It still fails during integration (but no long delays in failing to plate solve!). The full log: https://drive.google.com/file/d/1D4U25e9BObt5vq9lAfnrq9BFv14SHOk9/view?usp=sharing

I need to get some sleep, but I'll check in again in 8-10 hours. I really appreciate the help!
Respectful Supportive
andrea tasselli avatar
andrea tasselli avatar
Summary so far:

Darks: BAD/WRONG (No use)
Flats: Useless/Wrong (No use)
Collimation: Off

Obviously, PI doesn't "throw fits". If you feed it crap you get crap back, simple as that. And the crap, in this case, is the darks which aren't darks at all and for the life of me I can't image how they were produced. The flats followed the same route but at least weren't just working and I can't speculate the reason why, but there you go. I did without and after throwing out 4 outliers here is the final stack up of the remaining 79 light:



I don't know about you but I'd reckon this is hardly IOTD material...
Daniel Petzen avatar
Daniel Petzen avatar
Bro, jeez... Take it down a peg or two.

Anyhow, thanks for your "informative and constructive" feedback. I'm going to skip past the part where you say all my images are crap, to where you say my darks are crap and "aren't darks at all and for the life of me I can't image how they were produced.", as I'm still keen on getting some advice to improve my astrophotography.

The darks were done with the same gain and exposure time as the lights. They were created at the same camera temperature (-20) and in complete darkness (the camera was in a chilly bin). If I'm doing it wrong or should be doing it any other way, then I'm all ears.

Your verdict on the flats is somewhat ambiguous (the word 'crap' is missing in this particular sentence), but I thing the gist of it was they were also 'crap'. They were produced using N.I.N.A.'s Flats Wizard and the white t-shirt method. I positioned the light source so that the exposure ended up around 2.5 seconds in the 1-5 second exposure span I've configured the wizard with.

I've considered at getting a light panel to produce better flats, but had to spend my savings on other equipment. Yet again, I'd be keen to get some advice around producing flats.

Thanks for stacking and processing the subs! Yes, I can see that the result is quite far from what I was expecting.

The main reason I got a bit excited was when I grabbed a single sub to and did a bit of processing to see what it could look like:

So, my thinking was that if that is a two minute exposure, then a full stack of them would look amazing.

Clearly, I was wrong, but it's all about learning.

Cheers,
Dan
andrea tasselli avatar
andrea tasselli avatar
Yes, let's be constructive here (not that I see anyone else doing any heavy lifting here). Your darks aren't darks at all. Just some grey flat RGB files, incidentally produced by SharpCap, not NINA which was used for the other stuff. Why they are actually gray slabs, beats me but I don't have a clue. Just use NINA for everything. But at any rate pinning the blame on PI is just obviously wrong and to expect anything other than unusable blob coming on the other end is just wishful thinking. In fact I wondered why it allowed the procedure to go on but I expect these are the joys of using WBPP. So don't use it. 

Flats, you're technique is wrong because it doesn't correct lights. What is wrong I don't know and I can't know as I am not there. But you'll have to work on it. Dusk flats would the first thing to try. No t-shirt with diffusing lights behind.

Finally, your colors are wrong. Could that be there is no IRcut filter?
Daniel Petzen avatar
Daniel Petzen avatar
That'd be great, and thanks for your help.

I did the darks before I started using NINA, PI and APP, and I originally used them with DSS and AstroSurface. It worked OK with these applications, but that doesn't mean they're right for PI and APP, as I'm now starting to suspect based on your input.

The darks are bayered, so if I debayer (and stretch) one of them, then they look like this:

Does that look about right?

I do have a terrible amp glow on my ASI294MC, as you can see.

When it comes to the flats, well, that has been an endless struggle that never seems to end. I've watched so many instruction videos and I'm still, clearly, not getting it right.

If you, or someone else, could recommend any good videos or instructions, then I'd love to hear about them. I'll get a proper light panel if need be (once I have the money and they are back in stock).

As for the IRcut filter - no, I don't have one. This was done entirely without filters.

I've done a quick read-up on UV/IR-Cut filters and, yes, I clearly need to get one of those for when I'm not targeting nebulae with my Optolong L-eXtreme filter!

That sounds like it could make a huge difference, thanks. I've had problems with the colours being off, just as you say. I'm getting one ASAP.

Cheers,
Dan
andrea tasselli avatar
andrea tasselli avatar
Dark looks all right and remember never to debayer them nor any calibration master file. Dusk flats should be as easy as it gets but you need that IRCut filter first. Just wait for the sun few minutes before setting and point the scope to a clear patch of sky  and adjust (or have NINA do it) the exposure length so you cover between 25% to 50% of the dynamic range. Shoot 20 of them, get your flat darks and create the master flats. Do not track while recording the flats.
Helpful