Experienced Pixel Peeper required......

[Craney]

Evening everybody,

I have an Atik 414ex Mono and am exceedingly pleased with the results achieved so far.   The choice of DSO's has been restricted by a the small chip and the consequent smallish FOV, so I have invested in some Focal Reducers for my SW Equinox 80 to literally widen the horizons.  

BACKFOCUS !!    Being a bit slow to pick up on the obvious, I spent a couple of clear nights going for the big exposure runs with LRGB etc without checking the data.  So when I did eventually sit down to process the results, it became apparent I might need to think about backfocus a bit more.

An order promptly went off to FLO  (thanks, excellent service as ever)..... and when the various T2 extender bits arrived, I had a good session with some digital calipers and the internet and 'solved' the issue..... well... so I thought.

The question going out to the people who know about these things.... is.....  Are my star shapes below the result of an incorrect back focus  ??

..... or is there a strange internal reflection going on  ??    ( I did have a LP filter in the optical train before the Filter wheel , and this is the only bit I have not removed to eliminate it from my enquiries... and the 414ex is very sensitive).

The following shots are the results from my latest faffing.......

First, is the final stacked image from DSS. from about 10 x 3min subs...with a zoom-in to show the problem.....

 

test b.TIF

and secondly, here is a jpg of an original  TIF sub ...  with the same (ish) zoom section.

 

M31 top 3 min L3 3 min L.tif

 

Any thoughts welcome.

 

Sean.

 

 

[vlaiv]

I think it is processing artifact. Either there were some strange star core clipping due to hot pixel removal routine (cosmetic correction in DSS), or very funny curves were applied to final stack in stretching.

It will not show as much on single sub because of noise, but I got something similar looking in gimp on single sub, like this:

Your single sub looks just fine on its own and does not look like there is anything wrong with star shapes

[Big Jim Slade]

4 minutes ago, vlaiv said:

I think it is processing artifact. Either there were some strange star core clipping due to hot pixel removal routine (cosmetic correction in DSS), or very funny curves were applied to final stack in stretching.

Agree with Vlaiv here. Looks like a clipping problem during stacking. I don't use DSS, but I'll bet I could reproduce it in PixInsight if I had reasonably "hot" stars and high pixel clipping turned on.

[Craney]

Thanks @vlaiv  and @Big Jim Slade

It is reassuring to know that the optical train is OK.    I was not looking forward to night after night of 'tweaking' the distances.

I will re-stack  and  post some subs with the un-DSS-stretched result. 

I do not have the in-depth knowledge of what DSS does on the small scale.  I just load 'em up and press <play>...... and yes,  I am guilty of wrecklessly moving the sliders around in DSS to see what I have got.... impatience I guess.

[Craney]

OK then,   here are five subs from my M31 session

M31 top 3 min L3 3 min L.tif

M31 top 3 min L4 3 min L.tif

M31 top 3 min L5 3 min L.tif

M31 top 3 min L6 3 min L.tif

M31 top 3 min L7 3 min L.tif

and this is the result of DSS stacking  test C raw stack.TIF

( without any fiddling on the sliders....    The effect is still there, but not as pronounced... because maybe I only used 5 of the subs ?)

 

Please feel free to have a play, and maybe suggest which 'check-box' I have ticked / unticked in DSS which I shouldn't have.

Again, any useful insight, much appreciated.

 

Sean.

 

 

[vlaiv]

There is something seriously wrong with stack

Just look at histogram:

All the values are clustered and distinct - almost like there is only 16 distinct values in the stack (not sure how can that happen, I'll download subs now and do a quick stack of it to see what happens).

[vlaiv]

Here is result of quick stack - I did not look at individual frames (might be that there is an offending one since stars are misshapen in stack):

But result seems quite ok in comparison to stack from DSS

[vlaiv]

Yes, there are some guiding issues (or tracking if you did not guide), have a look at montage of two stars from each of five frames:

second and third show elongation in same direction - probably RA, while last two are a bit larger - first frame is "spot on".

[Craney]

Wow, thanks @vlaiv

That looks 2.2 million light years better than my first attempt !!!!....

Mis-shapen stars were the least of my worries tbh,  I may have a few tilt issues introduced due to the collection of spacers and adapters I have on the optical path.   That I can understand.    All this DSS clipping stuff   ( ? ).......   Any way, I have been experimenting with the DSS stacking parameters, and I may have improved the process.

You were right, the 'cosmetic'  settings page seems to be the source of my weird stars...    Plenty to fiddle about with there.

[Craney]

5 minutes ago, vlaiv said:

Yes, there are some guiding issues (or tracking if you did not guide), have a look at montage of two stars from each of five frames:

second and third show elongation in same direction - probably RA, while last two are a bit larger - first frame is "spot on".

 

Oh yes,  my tracking......  something to be worked on.   (  )   1 out of 5  is pretty good for me.

I have been using a Lacerta MGEN, which has been great until recently when it developed a strange fault where it keeps losing the star that I could quite clearly see on the display. Thats another story.     Well, as a result of that I have been using PHD2 for the first time..  I am the 'Dummy' that the programme was aimed at, and like DSS there are a few options I need to check out a bit more carefully,   "calibration"  I would imagine

BUT, the good news is that the BACK FOCUS looks OK.......  WheyHey !!!!!!!

 

Many thanks  @vlaiv    for your time and expertise.  It is much appreciated.        If we ever meet up at a star party I will treat you to the finest Yorkshire Ale.

 

Sean.