Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

collimation error or drift?

  • Please log in to reply
22 replies to this topic

#1 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 22 September 2019 - 07:20 PM

I took this image on friday night and the stars are not perfect. It was taken with and 8 inch SCT and a hyperstar and Zwo 1600 colour camera. The mount was an HEQ5 and I used pole master to get the alignment right. These were about 80 3 minute subs. What I cant tell is if the non circular stars are due to collimation error or tracking error or something else? 

Any tips for improvement most welcome. 

Attached Thumbnails

  • soulnebulastacked_ABEsundaysmall.png


#2 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1937
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 22 September 2019 - 07:46 PM

I would love to try and diagnose it, but the image is way too small to see anything useful.

Can you resize the original to about 1400 pixels wide and repost?


  • skydivephil likes this

#3 fcathell

fcathell

    Apollo

  • *****
  • Posts: 1172
  • Joined: 30 Jul 2017
  • Loc: Tucson, AZ

Posted 22 September 2019 - 09:17 PM

Wow - if I could have gotten wide field images like that in the "old days" with hypered film, I wouldn't give a hoot about the star imperfections! Would typically use a 250 to 400mm fast telephoto lens on an SLR which was piggy-backed on a tracking scope.

 

Frank


  • skydivephil likes this

#4 rgsalinger

rgsalinger

    Fly Me to the Moon

  • *****
  • Moderators
  • Posts: 5108
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 22 September 2019 - 09:18 PM

What you need to do to get sensible help is to post a link to several images that have received no processing at all. That's the only way to even see what might be going on with your system. You don't mention a guiding system. If these are unguided then everything is almost certainly explained by bad tracking. 

 

At the same time, the easiest way to separate optical issues from guiding issues is to take short exposures and gradually lengthen them. If you have tracking problems then the longer exposures will have more eccentricity and probably also poorer FWHM. In addition, it's likely that you will have the eccentricity along one axis. It's rare that I ever see real evidence of field rotation in three minute subs and so, you either have an optical problem (not necessarily collimation) or you have a guiding issue. Short exposures will help sort it out. 


  • jdupton and skydivephil like this

#5 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 04:42 AM

I would love to try and diagnose it, but the image is way too small to see anything useful.

Can you resize the original to about 1400 pixels wide and repost?

I had to reduce the image size to upload it to cloudy nights. But then I realised I could upload it to flickr and post the link so here it is https://www.flickr.c...57711024815906/



#6 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3989
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 23 September 2019 - 04:46 AM

My guess is some field rotation caused by bad polar alignement, combined with sensor tilt.


  • skydivephil and OldManSky like this

#7 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 04:49 AM

What you need to do to get sensible help is to post a link to several images that have received no processing at all. That's the only way to even see what might be going on with your system. You don't mention a guiding system. If these are unguided then everything is almost certainly explained by bad tracking. 

 

At the same time, the easiest way to separate optical issues from guiding issues is to take short exposures and gradually lengthen them. If you have tracking problems then the longer exposures will have more eccentricity and probably also poorer FWHM. In addition, it's likely that you will have the eccentricity along one axis. It's rare that I ever see real evidence of field rotation in three minute subs and so, you either have an optical problem (not necessarily collimation) or you have a guiding issue. Short exposures will help sort it out. 

HI , here is a link to an image that has received no processing , it is a singe sub ahtough I did hit screen transfer function in pixinsight  just to brighten it up . https://www.flickr.c...57711024815906/



#8 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 04:51 AM

My guess is some field rotation caused by bad polar alignement, combined with sensor tilt.

I have posted a sub now which may help diagnosis, wondering if you still have the same opinion ?https://www.flickr.c...57711024815906/



#9 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 04:55 AM

What you need to do to get sensible help is to post a link to several images that have received no processing at all. That's the only way to even see what might be going on with your system. You don't mention a guiding system. If these are unguided then everything is almost certainly explained by bad tracking. 

 

At the same time, the easiest way to separate optical issues from guiding issues is to take short exposures and gradually lengthen them. If you have tracking problems then the longer exposures will have more eccentricity and probably also poorer FWHM. In addition, it's likely that you will have the eccentricity along one axis. It's rare that I ever see real evidence of field rotation in three minute subs and so, you either have an optical problem (not necessarily collimation) or you have a guiding issue. Short exposures will help sort it out. 

it is unguided yes 



#10 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3989
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 23 September 2019 - 05:23 AM

I have posted a sub now which may help diagnosis, wondering if you still have the same opinion ?https://www.flickr.c...57711024815906/

Well thats odd. The sub is much sharper and doesnt show field rotation - could the stacking procedure somehow not allow for rotation of subs? I dont understand how sharp subs like that would be stacked to produce a 'field rotation' stack!

 

The upper right still has funky stars, so something going on there, probably sensor tilt or something out of alignment.


  • Stelios and skydivephil like this

#11 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 06:20 AM

Well thats odd. The sub is much sharper and doesnt show field rotation - could the stacking procedure somehow not allow for rotation of subs? I dont understand how sharp subs like that would be stacked to produce a 'field rotation' stack!

 

The upper right still has funky stars, so something going on there, probably sensor tilt or something out of alignment.

Thanks,  maybe I can try a different stacking program. Im using Nebulosity. I use save each file translation , select the stars, throw away no good images, then I use the save stack alignment method none, Std dev 1.5 . Maybe one of these settings is wrong? Also what is sensor tilt?



#12 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1937
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 23 September 2019 - 07:52 AM

"...alignment method none..."?

I'm not a Nebulosity user, but to align your stack you need to select some kind of alignment method...


  • psandelle likes this

#13 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3989
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 23 September 2019 - 07:52 AM

Thanks,  maybe I can try a different stacking program. Im using Nebulosity. I use save each file translation , select the stars, throw away no good images, then I use the save stack alignment method none, Std dev 1.5 . Maybe one of these settings is wrong? Also what is sensor tilt?

No alignment method? Well, theres one problem solved! You need to align them. Thats why the field rotation shows up in the stack.

 

Also if std dev is the number of std deviations beyond which to reject, then 1.5 is way too low. try 2.5 or 3. That'll reduce the noise quite a but as it will reject less.

 

I would suggest using deepskystacker - very simple and works brilliantly.


  • psandelle likes this

#14 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 09:50 AM

No alignment method? Well, theres one problem solved! You need to align them. Thats why the field rotation shows up in the stack.

 

Also if std dev is the number of std deviations beyond which to reject, then 1.5 is way too low. try 2.5 or 3. That'll reduce the noise quite a but as it will reject less.

 

I would suggest using deepskystacker - very simple and works brilliantly.

 the alignment method is done in translation. You should not align in the next step, the stack, as its already been aligned in the translation stage, right?



#15 Alex McConahay

Alex McConahay

    Cosmos

  • *****
  • Posts: 8177
  • Joined: 11 Aug 2008
  • Loc: Moreno Valley, CA

Posted 23 September 2019 - 10:19 AM

I cannot speak to Nebulosity's terms for alignment, but generally, there are three ways to align.

 

Translate: This moves a star up, down, left or right. The movement is 90 degrees from the top or bottom edges of the frame. 

Rotation: This moves the whole field around. In other words, a star can move in a curve, not just at 90 degrees to the edge of the frame. 

Scale: The frame itself can be stretched inward and outward. (Grown or shrunk)

 

In practice, all three of these can be used at the same time. Furthermore, in sophisticated software, the individual centroids of the various stars can be used so that the corners, for instance, are scaled, rotated, or translated differently from the center of the frame. 

 

All this is done by using a master frame as where the stars should be, and having the stars of the other frames properly aligned to that one master frame.

Alex


Edited by Alex McConahay, 23 September 2019 - 12:47 PM.

  • happylimpet, skydivephil and OldManSky like this

#16 rgsalinger

rgsalinger

    Fly Me to the Moon

  • *****
  • Moderators
  • Posts: 5108
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 23 September 2019 - 11:11 AM

While anything is possible, it's unlikely that you can go 3 minutes with an 8" SCT and get round stars consistently. I know that I can't do that on a much "better" mount. The second thing that's been mentioned is that the normal processing for color images is to debayer them, calibrate them, align them and then integrate the ALIGNED images. Sounds like you are skipping alignment. This will also give you trouble when it comes to the star shapes. So, my advice is to get an off axis guiding system or take 30 second exposures and read carefully the Nebulosity documentation about how to process one shot color images. You do not need new stacking software. Lots of people successfully use Nebulosity to stack color images with good results.

Rgrds-Ross


  • skydivephil likes this

#17 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3989
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 23 September 2019 - 02:32 PM

sounds like, as said above, youre just performing a translation on the image, without the necessary rotation.

 

the only reason i suggested DSS is that its straightforward - not to say nebulosity isnt usable too.


  • skydivephil likes this

#18 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 23 September 2019 - 08:49 PM

sounds like, as said above, youre just performing a translation on the image, without the necessary rotation.

 

the only reason i suggested DSS is that its straightforward - not to say nebulosity isnt usable too.

Ok i tried DSS and I think I do see an improvement, but maybe its just my imagination. Here is the original stacked in Nebulosity 

https://www.flickr.c...57711024815906/

 

and here is the same using DSS

https://www.flickr.c...57711024815906/

 

what do you think, much difference or not really?


Edited by skydivephil, 23 September 2019 - 08:49 PM.


#19 Galaxyhunter

Galaxyhunter

    Mercury-Atlas

  • *****
  • Posts: 2573
  • Joined: 02 Jan 2006
  • Loc: Northern Illinois

Posted 23 September 2019 - 08:52 PM

Much better.waytogo.gif


  • skydivephil likes this

#20 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 7700
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 23 September 2019 - 09:09 PM

Ok i tried DSS and I think I do see an improvement, but maybe its just my imagination. Here is the original stacked in Nebulosity 

https://www.flickr.c...57711024815906/

 

and here is the same using DSS

https://www.flickr.c...57711024815906/

 

what do you think, much difference or not really?

Much difference! However, you still have lousy stars a top right, but they improve as you go towards bottom left, which is a likely sign of tilt. Something  is not quite square to the rest of the equipment. Not familiar enough with Hyperstar to make suggestions. Not usually the easiest problem to fix. 


  • skydivephil likes this

#21 skydivephil

skydivephil

    Ranger 4

  • -----
  • topic starter
  • Posts: 369
  • Joined: 01 Oct 2016

Posted 24 September 2019 - 04:49 AM

Much difference! However, you still have lousy stars a top right, but they improve as you go towards bottom left, which is a likely sign of tilt. Something  is not quite square to the rest of the equipment. Not familiar enough with Hyperstar to make suggestions. Not usually the easiest problem to fix. 

Thanks , I'm not familiar with tilt, can you explain it?



#22 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 7700
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 24 September 2019 - 01:21 PM

Thanks , I'm not familiar with tilt, can you explain it?

I'm not an expert on the subject, but basically, your sensor is not entirely perpendicular to the optical axis. Not sure how this happens with Hyperstar, but in a refractor for example it can be caused by the focuser slightly deforming under the weight, or some accessory not threading properly -- and also by actual angle in the sensor when threaded correctly. 

 

ZWO actually makes a tilt adjuster -- no idea if it would help in your case or not. 


  • skydivephil likes this

#23 ntph

ntph

    Sputnik

  • -----
  • Posts: 25
  • Joined: 20 Oct 2012
  • Loc: Sudbury, ON

Posted 24 September 2019 - 10:06 PM

I use Nebulosity and agree with your processing steps. For those who haven’t used it, aligning and combining is done in two steps, one being to identify alignment star(s) and the second to then stack those “without” aligning ( since i’s already done to each separately). ( at least that’s my explanation—Craig Stark would really explain it!). I haven’t actually looked at the images in question—just clarifying this alignment terminology. Hard to look at these things on a phone...I use and like Nebulosity and will continue to use it, mostly for an internal standard as I take on learning PI.
It will be interesting to follow this thread...


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics