M101: My Second Deep Sky Attempt - Processing Question

[feilimb]

Last night I managed to capture somewhere between 60 and 90 minutes of 4 minute subs on M101, at ISO800, F/6.3 on a Canon 450D hooked up to my C8 SCT.  The quality of the subs seems decent, with some exceptions when cloud rolled in or when the guiding became poorer (I only noticed at the end of the session that there was heavy dew on the objective of my mini guide scope).

I didn't capture any flats, darks or bias frames, although I did perform dithering between each sub.  The subs didn't look amazing at first glance, but I tried stacking the best 70% in DSS at pretty much all the default settings, and then opening in my trial PixInsight and doing a crop and dynamic background extraction.  That is pretty much where I am at, I haven't done anything else yet, but when the DBE finished processing I was delighted to see the image below jump out at me (with STF applied).

I haven't gone any further with the image processing yet as I spotted a lot of strange small red, green and blue 'splotches' in the image.  I am wondering if these are some kind of artifact which is a result of the stacking? Has anyone seen these before, or know how they might be eradicated before I go any further with processing?

Any tips greatly appreciated!  I can upload the subs to Dropbox if that is useful. 

Update: I have added a Dropbox link to the subs I captured here: https://www.dropbox.com/sh/ftnpfsfcqs7353x/AAA02RFxWhE3MTjoFiq_2Piua?dl=0

 

[Uranium235]

Looks like too many sampling points to me, all you need (to start off with) is 8:

1 in each corner (or as near as you can without landing on a star)

1 each in top & bottom middle, left and right middle. But in this case I would place the left middle marker lower to totally avoid that bright star.

[gnomus]

The coloured spots are probably a result of hot or dead pixels.  They probably show you exactly how you dithered.  Whilst there is an argument for not using darks, you should definitely use bias frames (you can shoot these after the event) and flats.  I would suggest trying darks too.  After calibrating, but before alignment, try cosmetic correction.  There are numerous tutorials and videos out there showing you how to do this.

[gnomus]

Your image could do with a shot of SCNR Green too.

[feilimb]

Thanks gnomus and uranium for the tips. I went back to look at the original stacked image and I see the splotches in that image too, so I think you hit the nail on the head regarding dead pixels and it showing the dithering which took place.

As you say I should have taken bias and other calibration frames, but am still on the learning curve and was trying to stick to using just light frames for my first few efforts.  Bias frames seem to be the easiest to obtain so I really should take some of these on my next session at the least.

I'm going to go back to DSS again and try some other stacking options to see if I can get an image which doesn't have these artifacts at the end.

[gnomus]

I would say that both Bias frames and Flat frames are fairly essential.  Darks with a DSLR are a challenge because it is very difficult to know what is happening to the temperature on the sensor during a long exposure.  And who wants to waste nice dark skies shooting hours of darks?  Some folks will shoot their darks in a cold dark place during the day (the garage maybe) but it is a problem.  You could shoot bias now and reprocess the subs you have already taken.  If you haven't taken the camera off the scope (EDIT: And don't shift focus) you could also do flats later.  

[feilimb]

Thanks Gnomus - regarding taking flats later, in theory could one take the scope of the mount at the end of imaging session with camera still attached, and take flats eg. the next day (assuming camera position and focus are left untouched) ?  I haven't dabbled in taking of flats yet so need to get up to speed on how to take them.

I went back to DSS this evening and chose to stack with drizzle (x2) and kappa sigma clipping options.  After using Harry's Astroshed newbie videos as a guide this is what I have now achieved after a couple of hours basic processing (still getting to grips with it all).  I'm pretty happy with it  although I know there is much to improve as I learn more.

[gnomus]

7 hours ago, feilimb said:

Thanks Gnomus - regarding taking flats later, in theory could one take the scope of the mount at the end of imaging session with camera still attached, and take flats eg. the next day (assuming camera position and focus are left untouched) ?  I haven't dabbled in taking of flats yet so need to get up to speed on how to take them.

...

 

Yes that is correct.  There are different ways of doing them.  The classic way is throwing a T-Shirt over the objective, and pointing the scope at the sky.  They don't take that long - you need about 20-25 x 4-5 second exposures.  They are extremely important.

Your second image is a significant improvement, don't you think?

[wimvb]

Great improvement!

[ollypenrice]

I don't stack in DSS or PI but a hot pixel filter is a fantastic tool to be applied to each sub before combining. This can be automated in all the stacking programmes of which I'm aware.

As the others have said, I'd regard bias and flats as important but darks can work for you or against you. If using a DSLR a large dither (around 12 pixels) is recommended by Tony Hallas to combat 'colour mottle' in the background sky where signal is low.

Olly

[steppenwolf]

I don't use PI myself, I stack in MaxIm DL and occasionally in DSS but SD Mask (MaxIm) or Kappa Sigma Clip (DSS) will remove the hot pixels shown in your earlier image if you dither as you have done. I have an on/off relationship with Darks but always use Bias and Flat frames as these always contribute positively to the final image. I agree with Olly that a hot pixel map can help with these artefacts but these can often over-correct leaving a dark pixel which can be intrusive. However, dithering and using a stacking method that removes outlier pixels can be very successful and doesn't leave any visible artefacts but you'll need in excess of four sub-frames for the process to really kick in.

To get going with Flats, take a set and re-use them for a while - not ideal but considerably better than no Flats at all!

[feilimb]

Thanks all, my next project after this thread is definitely to try to take a few flats even if it is just a trial run with a different camera orientation etc. 

With bias frames - can these be library type images that you can reuse again and again, or should they be taken at same temperature and conditions as light frames? I guess camera orientation and telescope connection is not important for bias frames as the exposure is so short? 

[steppenwolf]

Again, I have a library of Bias frames for my cameras but I refresh them every few months. Library Darks are more problematical, especially with a DSLR camera as you don't have any temperature to match to - the sensor will warm up differently from session to session depending on the ambient conditions. Library Darks are fine for cameras with set-point cooling but again, I refresh these every few months.

Definitely try and nail those Flats!

[gnomus]

Bias: Shoot 100-200 at the shortest exposure your camera will do.  Lens cap (or better, body cap) on and in darkness preferably.  Orientation doesn't matter and camera does not even need to be on the scope. Once stacked your Bais_Master can be used again and again.  You might want to redo it very 3-6 months.  Simples. 

[Stub Mandrel]

Amazing how going back to an image after a bit of advice can perk it up. Nothing like a fresh pair of eyes (or several pairs) for spotting what I can't see after ages looking at the same image.