Help on stacking

[Xen0n]

Hi all.

Yesterday, I eventually managed to get my tracking to work. It's my first attempt at astrophotography.

I had good hopes to get a decent shot of Pacman Nebula (NGC 281) as APT showed me a pretty good image in preview (so I guessed that stacking 20 of them and a few blacks could only give better results):

The imaging sessions is:

- 20 RAW 120 seconds shots at 1600 ISO

- 4 Blacks with identical settings

The hardware is a Skywatcher 130/650 on a motorized EQ6 with camera guiding.

But the result after stacking them is quite disappointing (lightened result - original TIF added to the post):

I tried the following method to filter out pollution:

http://www.skyandtelescope.com/astronomy-resources/astrophotography-tips/remove-light-pollution-astro-images/

But the result is not even close to what APT gave me on a single shot:

I did this session in my backyard, the light pollution is quite high. I'm using a de-filtered 1000D without light pollution filter (on my list of stuffs to buy).

Would someone have advises for me to get a result which is closer to what APT gave me but improved with the stacking process?

 

Thanks in advance.

Pacman.TIF

[alacant]

Hi. Excellent effort. Here's a quick stretch and denoise using GIMP's largest hammers. I don't think it's gonna be rescuable as it is unless you stack flat frames too. Meanwhile, one of these along with the flats will help enormously with the orange mush. HTH.

[wimvb]

Welcome to the forum, Xen0n.

Here's my attempt with PixInsight:

2 passes of DBE to get rid of most of the vignetting and gradient

Background neutralisation & Colour calibration

Masked stretch

Curvestransformation + colour saturation

Star reduction

Noise reduction

(click to enlarge)

To properly remove the vignetting, you do need to add flats to the calibration frames. The strong red colour in the integrated image is due to light pollution and very likely an effect of your modified camera.

The streaks in the background can be removed by careful calibration (finetuning the darks and a bit more aggressive sigma clipping during stacking).

There's also a satellite trail that can be removed by sigma clipping, or easier by removing the subframe from the stack.

[wimvb]

Btw, here's a quick test with the new arcsinh stretch method of @sharkmelley. Imo, it's clipped too hard, but that's just me figuring out this new stretching method.

Otherwise almost the same process as above, but a little less star reduction and no noise reduction at the end.

[carastro]

Yes, the data is there, but will be pretty useless to process without flats or a very experienced person's magic.  

But we've all been there, so that's your next step to master.

Carole 

[Xen0n]

First impression: WOAW!

Thank you all for your quick replies and advises. It may seems trivial for you but as it's my first attempt ever, this is blowing my mind and gives me hope for future sessions.

Regarding the satellite, I thought I had removed this sub from the stacking process (but apparently not).

For the filter, here's what I had in mind: http://www.astroshop.eu/broadband-filters/astronomik-cls-ccd-xt-clip-filter-canon-eos-aps-c/p,54608

Now I must figure out how to apply your techniques:

- quick stretch and denoise using GIMP's largest hammers (Alacant)

- 2 passes of DBE to get rid of most of the vignetting and gradient, Background neutralisation & Colour calibration, Masked stretch, Curvestransformation + colour saturation, Star reduction and Noise reduction (Wimvb)

I'm a total newbie in this domain. Would you have tutorials/videos or books to recommend?

Next time, I'll also do flats. I read somewhere that a white picture displayed on a tablet in front of the aperture while taking pictures at ~2/3 of the histogram does the job. Do you have any opinion on that?

Again, many thanks!

Julien

[wimvb]

7 hours ago, Xen0n said:

Next time, I'll also do flats. I read somewhere that a white picture displayed on a tablet in front of the aperture while taking pictures at ~2/3 of the histogram does the job. Do you have any opinion on that?

When taling flats, it's imortant you keep the exposure in the linear part of the sensor response curve. You may have to experiment, but I would start with pixel values about 50 % of max scale.

Btw, the processing that I did was with PixInsight. Not cheap, but a great program. Pixinsight.com

[Xen0n]

Hi all.

It's been a long time since my first attempts. I upgraded my mount with OnStep project and I now have a Goto mount.

I started shooting M101 without using any guiding (just plain old sideral speed on the RA axis). The result was quite good, here is a 120 seconds exposition:

No blur so I launched a 50x120 seconds imaging plan.

Then I realized that there was a drift between each shot. Here is a simple stacking of a few shots in photoshop using 50% transparency:

What I don't understand is: why does each single 120 seconds shot is sharp while I'm having a "shift" between shots?

Event weirder to me: I attempted a stacking with Deep Sky Stacker and I get the following result:

M101 is sharp, many bright stars are sharp but fainter stars are blurred...

Any idea on what could create such effect?

Thanks in advance.

 

best regards.

 

[Demonperformer]

Suggest you get a copy of Making Every Photon Count - obtainable from our sponsors - worth its weight in gold for the advice in it.

[wimvb]

I'venever seen a good explanation for the "jumps" between subs. My personal theory is that it's caused by "stiction". The vibration of the mechanical shutter and mirror in the dslr causes the mount to move slightly. Maybe someone else can come up with a better explanation.

The streaks you see in the stacked image are caused by hot pixels that were not calibrated out properly. For an astro camera, you would use darks to remove these. But for a dslr this may not necessarily work, as the camera isn't cooled. Increasing the aggressiveness of the pixel rejection during stacking can reduce the effect. As @Demonperformer suggested: time to get a good read. Order the book he recommends, it really is worth it.

Clear skies and have fun.

[newbie alert]

Where was the histogram on the individual subs,as  iso 1600 combined with the noise level of the camera combined  with the light pollution maybe just too much.. take it the moon was up too?

 

[Stu Wilson]

Had a quick go, Gradientxterminator  and couple stetches.

 

Stu

Pacmanredo.tif

[Stu Wilson]

And then had another go ……….

[Xen0n]

Hi all.

Thanks for the answers and the book reference. i'll get a copy of it very soon.

Thanks also for the 2 reworks of pacman, it'll be my new wall paper.

I will make new attempts as soon as possible.

[almcl]

I had the 'walking noise' similar to that in your M101 shot.  In my case it was cured by very much more vigorous clamping between guide scope and imaging scope, to prevent the two differentially flexing and by large scale dithering between frames.  This helped get rid of  the hot pixels that were the source of the streaks.  

The blurred stars in the outer areas might be coma, do you have a Coma Corrector in your imaging train?

[Xen0n]

On 03/07/2018 at 13:41, almcl said:

I had the 'walking noise' similar to that in your M101 shot.  In my case it was cured by very much more vigorous clamping between guide scope and imaging scope, to prevent the two differentially flexing and by large scale dithering between frames.  This helped get rid of  the hot pixels that were the source of the streaks.  

The blurred stars in the outer areas might be coma, do you have a Coma Corrector in your imaging train?

Hi, thanks for your answer.

During this test I was not using the guide scope, just "plain old" stellar speed on the RA axis with a good polar alignment.

I'm not using a coma corrector. Moreover, for this test, as far as it seemed to be failed, I did not take flats, the result is only light frames without any darks, flats or anything...