Don't assume that triangular stars mean pinched optics

Equipment Collimation Troubleshooting
15 replies•1.1k views
Andy Wray avatar
Andy Wray avatar
I posted a few times on here about my horrible star shapes ... almost triangular with some square edges etc..  Many people suggested it was pinched optics.

In the end, it just turned out to be my bad collimation; particularly the focusser positioning over my secondary (and hence my coma corrector) being tilted with respect to the optical axis.

Just wanted to share as it is not always obvious what causes these weird star shapes.

I still have work to do, but here is a recent abberation inspection following proper collimation/alignment.  This is far from perfect, but at least most of my triangular stars are gone.

Thanks to @John Hayes for pointing me in the right direction.

Helpful Insightful Respectful Concise Engaging Supportive
Chris White- Overcast Observatory avatar
Chris White- Overcast Observatory avatar
Very slight triangle shapes can also be caused by a Bayer matrix if you use an OSC camera.
Well Written
andrea tasselli avatar
andrea tasselli avatar
They are still very much triangular if you ask me.
Andy Wray avatar
Andy Wray avatar
andrea tasselli:
They are still very much triangular if you ask me.

They are better than they were, but I see what you are saying when I look at the larger stars in particular.  I'll keep working at it.
Well Written
wittinobi avatar
wittinobi avatar
have you checked your primary, i still think its pinched.
stars are still looking horrible to me,  but if you are happy, …fine.
Arun H avatar
Arun H avatar
Andy Wray:
andrea tasselli:
They are still very much triangular if you ask me.

They are better than they were, but I see what you are saying when I look at the larger stars in particular.  I'll keep working at it.

I would say even the smaller stars are still triangular. Many of us get to be very picky about star 💫 shapes!! Posting a “before” picture would help show the improvement. But the general advice I agree with - many things can cause defective stars. I thought for an earlier scope I owned that I had pinched optics first from the primary, then from the corrector, but now am told by good authority that it was something else entirely.  Working with someone that knows what they are doing, like John, is always helpful.
Andy Wray avatar
Andy Wray avatar
I'm slowly getting there with my collimation and the stars are now a bit oval with a few too many diffraction spikes.. That said, they are not too distracting on a full frame:
Haakon Rasmussen avatar
Haakon Rasmussen avatar
Hallo Andy,

i think you are right about the collimation causing bad stars. I had the same problems with my stars for a long time using only  a Concenter Eye Piece to do the collimation. It was just not accurate enough. If i inspected the image with ccd inspector it told me i had tilt. I tried to adjust my focuser, but it just got worse. I decided to not adjust the focuser any more, and bought the Ocal Electronic Collimator. I watched the video of Daniel Nimmervoll on youtube how to use it. With the Ocal Collimator i was able to center and level the secondary mirror properly, thus becoming a fairly illuminated field and round stars. Check this out:  compared to this: https://www.astrobin.com/7vwj3n/B/ compared to this: https://www.astrobin.com/dea91o/ . If you check all my resent photos you will sometimes see some star distortion because i had not checked my collimation well enough, or the telescope lost its collimation under way. The TS-UNC Newtons (800mm F4) are not particulary good on holding its collimation. Before  i had a Skywatcher 150PDS, and i had the same problems getting it well collimated with the Concenter Eye Piece. I hope this can help you to continue your quest of getting a well collimated Newton Telescope.

CS Haakon
Helpful Supportive
Andy Wray avatar
Andy Wray avatar
I'm not sure who can help with this, but looking at last night's master lights:

* The Ha and Oiii stars were reasonably round
* The Sii stars were definitely triangular
* with the sequence I was running I had passed the meridian before the Sii frames were taken and it moved back after that to taking another set of Ha frames which had roundish stars, so was not down to any change in the imaging train.

Any ideas why that might be?
Alan Brunelle avatar
Alan Brunelle avatar
I agree with most here that your stars still have a triangular component, but I also agree that you have improved things having followed your other post.

I wonder if you have taken my suggestion of directly confirming the pinched cc that I made.  You were the one who brought it up by saying that you had been cranking down on the clamp screws.  It would be a rapid test.  Get your rig pointed at you test star, preferably with your camera in a position where it can't fall out.  Start with a not too tightly clamping, take a few subs, then crank the screws like you had been doing before.  I'd like to know the result, since I noticed this problem with my scope.

The thing I can't explain is why the image I sent you with my tightly clamped cc, is why three screws yielded 4 sided stars.  I would have expected three sided like yours.
Andy Wray avatar
Andy Wray avatar
Alan Brunelle:
I agree with most here that your stars still have a triangular component, but I also agree that you have improved things having followed your other post.

I wonder if you have taken my suggestion of directly confirming the pinched cc that I made.  You were the one who brought it up by saying that you had been cranking down on the clamp screws.  It would be a rapid test.  Get your rig pointed at you test star, preferably with your camera in a position where it can't fall out.  Start with a not too tightly clamping, take a few subs, then crank the screws like you had been doing before.  I'd like to know the result, since I noticed this problem with my scope.

The thing I can't explain is why the image I sent you with my tightly clamped cc, is why three screws yielded 4 sided stars.  I would have expected three sided like yours.

Thanks Alan!  I did address that and made sure that I didn't over-tighten the CC last night and I am sure that was part of my issue.  I have two screws on my brass compression ring; interesting that that may have been exaggerating the triangular stars.  I haven't done the exact test you have mentioned, but will give it a go on another night (it's clear tonight, so I'm making the most of it).
Alan Brunelle avatar
Alan Brunelle avatar
As an aside, maybe a joke.  Maybe!  With all the great functions our imaging software now has (I use only PI), I think someone who knows how to develop and write scripts should come up with a Make Stars Round script.  Or even a global, Correct Star Abberation, that also extend the abberation defect correction to all parts of the image.  Sounds crazy!  But, hey we now have one touch star removal, etc, etc.  Who'd of thunk!

I could imagine that the script would do a number of things to achieve this:  1. Locate and record precise locations of all stars.  2. Record brightness with auto ROI and total signal for each star.  3. Record color for each star (for RGB).  4.  At each star position in either a starless version or on a zero background template, create an ideal psf (maybe user definable) based on star brightness, add back color, if RGB and viola!  5. Merge new stars onto the starless version.  I can't think of any idea as to how to correct the defect globally.
Engaging
Alan Brunelle avatar
Alan Brunelle avatar
Andy Wray:
I did address that and made sure that I didn't over-tighten the CC last night and I am sure that was part of my issue.


That is great to hear Andy.  I have set up my rig with direct threading of my cc to the focuser tube.  It will likely be a month before I can confirm that I have sufficient back focus to use it that way.  I will report.
kuechlew avatar
kuechlew avatar
Alan Brunelle:
As an aside, maybe a joke.  Maybe!  With all the great functions our imaging software now has (I use only PI), I think someone who knows how to develop and write scripts should come up with a Make Stars Round script.  Or even a global, Correct Star Abberation, that also extend the abberation defect correction to all parts of the image.  Sounds crazy!  But, hey we now have one touch star removal, etc, etc.  Who'd of thunk!

I could imagine that the script would do a number of things to achieve this:  1. Locate and record precise locations of all stars.  2. Record brightness with auto ROI and total signal for each star.  3. Record color for each star (for RGB).  4.  At each star position in either a starless version or on a zero background template, create an ideal psf (maybe user definable) based on star brightness, add back color, if RGB and viola!  5. Merge new stars onto the starless version.  I can't think of any idea as to how to correct the defect globally.

I lost a lot of my interest in daytime photography when it became too easy. Travelling through the US national parks in the mid 90s with a large format camera was a challenge and a lot of fun, while these days it's almost impossible to create a badly exposed image you can't rescue in post production. I may sound crazy but a lot of my fascination for this hobby comes from the fact that it requires quite some effort to get good images even with the advanced tools at hand. Coming home with the collected data and turning it into a stacked image is a similar experience to me as developing a film in the old days. It's always a bit like Xmas if I got a decent result. So while I appreciate tools like NoiseXTerminator or Starnet or the WBPP script in PI, I'm happy that there isn't a "push here dummy" for image development. 

Your star improvement script sounds a bit like my suggestion in another thread: plate solve the image and then replace the crappy user data by the closest hubble image :-). I would loose interest quickly. Having said so I have no problem using my Vaonis eQuinox on occasion which is a "lazy noob device". These devices may very well turn into serious fully automated imaging devices at some point in the next 10 years but for now they are just some nice toys for a relaxed evening of stargazing, when there is not enough of clear sky for serious imaging. 

Clear skies
Wolfgang
Helpful Insightful Engaging
Andy Wray avatar
Andy Wray avatar
Alan Brunelle:
As an aside, maybe a joke.  Maybe!  With all the great functions our imaging software now has (I use only PI), I think someone who knows how to develop and write scripts should come up with a Make Stars Round script.  Or even a global, Correct Star Abberation, that also extend the abberation defect correction to all parts of the image.  Sounds crazy!  But, hey we now have one touch star removal, etc, etc.  Who'd of thunk!

I could imagine that the script would do a number of things to achieve this:  1. Locate and record precise locations of all stars.  2. Record brightness with auto ROI and total signal for each star.  3. Record color for each star (for RGB).  4.  At each star position in either a starless version or on a zero background template, create an ideal psf (maybe user definable) based on star brightness, add back color, if RGB and viola!  5. Merge new stars onto the starless version.  I can't think of any idea as to how to correct the defect globally.

I lost a lot of my interest in daytime photography when it became too easy. Travelling through the US national parks in the mid 90s with a large format camera was a challenge and a lot of fun, while these days it's almost impossible to create a badly exposed image you can't rescue in post production. I may sound crazy but a lot of my fascination for this hobby comes from the fact that it requires quite some effort to get good images even with the advanced tools at hand. Coming home with the collected data and turning it into a stacked image is a similar experience to me as developing a film in the old days. It's always a bit like Xmas if I got a decent result. So while I appreciate tools like NoiseXTerminator or Starnet or the WBPP script in PI, I'm happy that there isn't a "push here dummy" for image development. 

Your star improvement script sounds a bit like my suggestion in another thread: plate solve the image and then replace the crappy user data by the closest hubble image :-). I would loose interest quickly. Having said so I have no problem using my Vaonis eQuinox on occasion which is a "lazy noob device". These devices may very well turn into serious fully automated imaging devices at some point in the next 10 years but for now they are just some nice toys for a relaxed evening of stargazing, when there is not enough of clear sky for serious imaging. 

Clear skies
Wolfgang

Funny you should say that Wolfgang:  

When I was a teenager I spent many hours in my Dad's darkroom both developing negative film and then turning that into prints with trays of chemicals.  I felt I had really cracked it when I developed my own first colour print.  Those were very magical times where you maybe had 12 photos on a roll of 120 film, so you had to get each one right.

I then got one of the first digital camera SLRs ... An Olympus C1400XL which had 1.4 mega pixels and could only store a handful of photos on each memory card.  It actually took some wonderful photos of my daughter when she was really young (a few of which got published in photography magazines); many of which are in photo frames around the house.  You'd be hard-pressed to tell that they weren't taken with a modern day camera just by looking at them from a distance, yet nowadays everyone craves for 40M pixels plus.

Astrophotography is to me very like those early days of photography where you first had to make sure you captured the frames at the right exposure, aperture and focus.  We even used different filters to bring out detail (red would be good to bring out clouds). We'd then develop the negative film (pre-processing) and then used a projector where you had to get the exposure just right onto the paper (the equivalent of stretching) before post-processing in chemicals before finally putting it in fixer.  

Maybe that's why I like astrophotography; a similar, magical challenge.
Helpful Insightful Engaging
Alan Brunelle avatar
Alan Brunelle avatar
To @kuechlew ​​​​@Andy Wray 

Completely agree.  I started on hand processing B&W, then onto home processing color slides on standard Kodak and Velvia.  Terrestrial photography only.  When my postgraduate lab had a color print processor I then did cibachrome printing.  Absolutely spectacular results.  Those prints seemed to have real 3D depth!

When digital photography came around, I took many more photos, but rarely printed them.

I got into the astrophoto thing a few years ago, starting with my daughter's dslr, using my old visual telescope.  Decided that at my age, I needed to move more quickly up to better gear and heck, earned the right to do so.  The gear and the processing, such as it is now, was fun to learn.  Much more to learn.  But I do struggle with how some of these techniques alter the data captured.  I've seen methods of data processing used in online tutorials that I refuse to use.  I always process to details that I must be able to detect in either subs or unprocessed stacked images, and try to do it as globally throughout the image as I can. 

I currently live in an area where I am limited in hours of imaging in a year.  For me that may be a good thing.  Even now, as I contemplate moving to clearer skies, I still intend to limit my artistic imaging to just a few projects, marvel on this site at what the artists here can do with their rigs, small and large, and their use of the processing software.  Instead, I hope to turn my attention to more data analysis and processing in a discovery process.  I may fail, but the journey is the challenge!

Cheers!
Related discussions
Doubled bhatinov spikes.
An interesting thing happened last night. I got double bhatinov spikes on my edge hd 8. my backspacing was previously 104-ish, and everything was ok for the most part but i figured i would try to increase backspacing to the more reported 105ish. To d...
Doubled Bhatinov spikes relate to backspacing issues affecting star shape quality.
Mar 22, 2024
RC8 Collimation using Innovations Foresight SkyWave
Hi, I will soon resume AP using my GSO RC8, that hasn't been used for 2 years and travelled a bit by car in between. Last time I checked it wasn't collimated anymore. I doubt it will be when I'll get it out of retirement (never works that...
RC collimation guide directly relevant to author's collimation problem-solving experience.
Aug 26, 2024
GUIDE: Collimating Takahashi Epsilon 130 ED w/ OCAL
We all know the reputation these scopes have when it comes to collimation. As a former RC owner I always wondered if collimation was really that difficult on an Epsilon or if people just aren't using the correct method. I've owned an OCAL v2 ...
Collimation guide for specific scope model useful for collimation techniques.
May 13, 2024