M31/NGC6960 technique vs seeing

[Gerhard]

Hi all,

Just another update on my progress thus far...

I had some good nights recently, with rather variable results, so wanted to share, and hope to receive some feedback on  my musings as to the causes of the differences.

First I went high up (1600 m) on a quite clear night, with also good seeing, and I was able to take this:

Then, last week, I went less high (1000 m), good clear night, but with some more light pollution, but (IMHO) with better setup:

The latter image was done doing drift alignment, AND PEC on my HEQ5 SynScan. The former only with drift alignment.

I also tried the M31 on the second occasion, but it was much lower than the NGC6960 (which was close to zenith) and it has terrible traces and other anomalies: I had to dump a lot of subs... :-(

So, as you might see, the stars in the second image are much better, but the light was better on the first image...

I am obviously still also on a steep learning curve, so that might also have contributed...

Question: how much of the quality of the stars in the second image do you think is due to the mount being better balanced pointing to zenith?

If this is a significant amount, that would mean that for now I should limit myself to zenith objects, to obtain quality subs..... :-/

Second question: do you think making flats would help with the obvious lighting problems on the second image?

Friday maybe I'm going out again, and will have another try at the NGC6960, or maybe its sister NGC6992... There's a full moon, though.... :-/

Thanks!!

Gerhard.

[NGC 666]

First Answer: I think, though my mount (EQ5) doesn't have PEC, that most of the trailing (and I assume more bad subs) is due to having no PEC on in the M31 image. Because PEC helps correct errors which would cause trailing. From my experience, balancing the scope/weights on the mount is not terribly important for good stars, though you should balance east heavy, meaning make the setup slightly heavier on the east facing side - that way the gears are constantly under pressure and won't backlash. Another factor to consider is the position of the object: objects nearer to the celestial poles rotate in the sky slower than those at the celestial equator, so there's more room for error if the mount is inaccurate (not so much a zenith thing). WInd ect also contribute.

Second Answer: Flats definitely help to remove the gradient of the skyglow, especially notable in the Veil Nebula image. The full Moon completely ruins any attempt at deep sky imaging when imaging nebulous objects (star clusters are okay). A PS plugin called GradientXTerminator helps tremendously with the gradient, too.

[Gerhard]

First Answer: I think, though my mount (EQ5) doesn't have PEC, that most of the trailing (and I assume more bad subs) is due to having no PEC on in the M31 image. Because PEC helps correct errors which would cause trailing. From my experience, balancing the scope/weights on the mount is not terribly important for good stars, though you should balance east heavy, meaning make the setup slightly heavier on the east facing side - that way the gears are constantly under pressure and won't backlash. Another factor to consider is the position of the object: objects nearer to the celestial poles rotate in the sky slower than those at the celestial equator, so there's more room for error if the mount is inaccurate (not so much a zenith thing). WInd ect also contribute.

Second Answer: Flats definitely help to remove the gradient of the skyglow, especially notable in the Veil Nebula image. The full Moon completely ruins any attempt at deep sky imaging when imaging nebulous objects (star clusters are okay). A PS plugin called GradientXTerminator helps tremendously with the gradient, too.

hmmm... in fact NGC6960 is a bit lower than Andromeda (30, vs 41 dec).... but the difference is not huge...

Wind here in Italy is fortunately almost never a factor, when the sky is clear...

I tried to make some new flats, following carastro's page's instructions... but they did not have any significant effect on the final stack... but most likely it's my lack of post-production abilities...

I do not have PS....

[StargeezerTim]

Hi there,

Nice images! In terms of getting help and suggestions, its best to provide details of how many subs, exposure time, iso settings etc as these things all affect the quality of the outputs. also what software you use to stack and process the images. I have done a quick edit in PS and got rid of the vignetting. You could also do this in GIMP (free!). Its the gradient removal process found in one of Doug Germans videos where you create a copy of the picture, blur it, then subtract it from the original. 

Keep it going...Tim. 

[Gerhard]

Ok, you're right, sorry I didn't provide the stack info. :-)

M31: stack of 30 subs (if I discarded all subs with traces, I would have been left with only 1 or 2) of 73 seconds, 15 darks and 40some bias, no flats.

NGC6960: stack of 14 subs (weeknight, so little time to make more subs), 10 darks, and 30 odd bias, no flats.

All subs RAW, ISO800

Stacking and stretching with DSS, post with ACDSee. In post for now I only look at lighting, and exposure & contrast, and the likes. I am not experienced with layers.

I did download GIMP (and now am gonna have a try, trying to duplicate what you did), but don't use it yet...

I can't help but notice though, that in the process the veil lost a lot of its detail... Looks like the gradient removal affects the subject as well... :-/

Two other quick questions: as I gather flats can be taken at a different temperature as the lights, but also at different ISO? and different Focal length? IOW, should I take flats on the telescope, or can I do it with the normal lens attached to my DSLR?

Second question: I usually already start packing up when I start doing darks, so I take the OTA off of the mount and place it on the ground while still taking the dark subs. As until now my DSLR rotates automatically, the orientation of the darks sometimes varies. In your knowledge, does this screw up the stacking?

From now on I will avoid this possible problem: I made a note on my setup cheat-sheet to always turn off automatic rotation before starting to shoot (I use this camera for normal photography as well. As soon as I will buy a new one, the one I use for AP will be modded, and I will leave automatic rotation off indeterminately)

[StargeezerTim]

Hi Gerhard,

In my experience, gradient removal does affect the image detail and colour a bit, so flats are better! Having said that, you could bring out the colours and detail again now. I havn't done any processing, just removed the gradient.

I take flats with the camera still in place and focus on the scope. I use AV mode and keep the same ISO. Darks it doesn't matter about keeping it on the scope. You could take the flats, then remove the camera from the scope. Cap the camera (with the plastic cover) and take your darks while you are packing up!

Had another twiddle. Putting a bit of colour back also tends to bring the gradient back a bit. Better than it was though!

[Gerhard]

I had a go as well in GIMP (more like, a few goes...), now I finally understood how to remove the gradient.

I also practiced on other images, older than this one, and it works quite well, although images with big object, such as M42 or M31 are quite difficult...

[StargeezerTim]

Great stuff!

This is all about 3 steps forward, 2 steps back... overall its a steep learning curve but worthwhile... Tim. 

p.s. Proper gradient removal programmes will do a better job, but Gradient Exterminator is $50 and DBE (in Pixinsight) much more!

[NGC 666]

If you supply the full stacked tiff file we can get a better stab at reducing that light pollution.

[Gerhard]

In this dropbox directory:

https://www.dropbox.com/sh/iu0mfg5rf9hhdn8/AAAvMt4_4EihAHrLcg_OOGSea?dl=0

are some of my images. M52.tif, and M81_82.tif are from yesterday evening (with bright moon, so not too great). They include the stretch I did in DSS.

If I haven't suffered a complete mental breakdown, a NGC6960_nostretch.tif should appear later on, as the pc @home should still be chugging away with the upload. Otherwise I will reload it this evening.

You'll see also some of my tries with gimp from yesterday (all with filenames including "gimp").

Thanks for taking an interest! :-)

[Gerhard]

Now the NGC6960_nostretch.tif is present, as is the NGC6960.tif, which contains my stretch, as I usually do in DSS.

[NGC 666]

Hi,

I had a go at the Veil Nebula image. It seems as though we have both stretched as much data as possible in the image, otherwise making it brighter would just make the noise unbearable. So more subs are required. The gradient was also pretty successfully removed with GradientXTerminator, and light pollution with the levels tool in Photoshop. I tried a bit of noise reduction, too.

The colour balance seems fairly accurate, and the stars are sharp.

The image is, of course, of low quality as to be able to attach to the post.

Hope this helps.

(Your M31 image is great, by the way.)

[Astrosurf]

Nice images. They're difficult subjects too, as they're faint (except M31's core). I notice some coma. It would be worth getting a coma corrector. Have you thought about guiding?

Alexxx

[Gerhard]

Hope this helps.

(Your M31 image is great, by the way.)

Very nice! A lot cleaner image as the one we made with GIMP before.

Thanks about the M31! I am VERY glad with that one, I will be returning to the same spot in september for sure (www.StarParty.it). Hope to make some more images of that level with PEC and better drift alignment.

obviously more subs are needed for the NGC6960. I made only 15 (which fortunately came out all well) because it was a weeknight, and clouds were also coming in...

I'm guessing that sooner or later my trusted ACDSee and GIMP are not going to be enough... PS seems very complicated, though... Only yesterday did I learn my first elaboration with layers in GIMP...

[Gerhard]

Nice images. They're difficult subjects too, as they're faint (except M31's core). I notice some coma. It would be worth getting a coma corrector. Have you thought about guiding?

Alexxx

Thanks! I like a challenge! :-D I will try for the other half of the veil nebula, maybe this friday (will be a disaster due to full moon, but, hey.... :-) )

I thought about a coma corrector from the very beginning when I got my Newton, as it is a common problem with newtons, and, as it turns out, mine is no exception...

Guiding is definitely on my todo list.

everything (surprise surprise) depends on budget, though... lots of things to buy, not much €€ to spend...

I am currently borrowing a Barlow from the prez of my astroclub, which I use primarily for the drift alignment and PEC, to see anomalies better. It's also very nice for planetary visual, and with my 5mm EP it results in 300x which is about the maximum attainable for my Newt (F/5 150/750)

I need to buy a tent etc for future nights staying outside (it already happened once, and I could borrow a tent, but need one for myself..)

Guiding is a definite must for the future...

Maybe my astroclub has some things I can buy for not too much (guide scope & camera), but I firstly need to buy me a laptop... Then when I have those things, I don't know if my battery will suffice... it's only a small one...

What do you guys think about OAG? Is it a valid alternative? cheaper? easier? more expensive in the long run? I haven't got a clue...? :-)

[NGC 666]

Apparently using Off Axis Guiding reduces the amount of light that goes to the imaging camera. Also, it's a problem finding bright enough guide stars in the main scope's small FOV, and you'll probably end up using extension tubes to focus both the camera and the autoguider simultaneously.

There is less flex between the tube and the autoguider when using OAT compared with a guidescope, and you are zoomed in more with an OAG, so I'd assume it'd be more accurate at guiding - but I think the advantages don't make up for the aforesaid drawbacks.

Even though I've never guided, I advise a guidescope.

[Gerhard]

hmmm. you have a good point. some time back I read about a TOAG which also had a filter wheel, which seemed at the time a very valid thing to consider... Also because lots of beautiful DSO objects are Halpha, so a filter wheel for the future CCD seems inevitable... (but maybe I need to inform more...)

well, one step at a time... guiding is still talking in the future...