Help! Doubled stars (and trails) in my images?!

[Anthony RS]

Hello!

I've been having this issue for a while now, and I've been trying to find the culprit with no success. This is driving me crazy. For some reason, I've been having weird stars in my image, where it's not a typical star trail, but the stars are actually doubled or tripled in more than 50% of my images. I've attached an image with different previews from different images, one of the images in these previews has good stars while the others have weird stars which seem to go in different directions in each image.

Here are some notes:

-I'm using a celestron AVX with the 8 inch newtonian telescope, Canon 500D with SKywatcher aplanatic coma corrector, 60mm guidescope with zwo asi 120mc-s.

- My guiding looks normal, I'm having the issue with guiding error of 0.8" and my image scale is 1"/px. I've had no problems before even when my guiding error was 1.2"

- The dec graph seems to swing from north to south sometimes due to backlash but the error is always below my image scale. I think I've tried to guide with dec guiding off and had the same issue but I'm not sure (been a while since I've tested). Plus, I believe that if the problem was strictly due to the Dec behaving strangely, the star "trails" wouldn't be in a different direction in different images. Correct me if I'm wrong please. EDIT: After reviewing all the images. It seems like the trail of doubled stars is mostly in the same general direction (in most images but not all).

- My RA is always below 0.8"

- I'm positive my PHD2 settings are correct in terms of focal length, pixel size etc...

- I'm using a cheap 8 inch newtonian (celestron). I always collimate before the imaging session. I even checked collimation after the imaging session just to check if the mirrors are moving during imaging, causing this problem, but collimation always seem to be spot on after I'm done.

- I have tried to extremely tighten everything and all the screws you can think off. All the adapters, guidescope rings screws, literally everything. 

- I tried a different camera.

- I've tried different exposure times from 180s to 300s, and different PHD2 parameters. I have yet to try exposure times of less than 180s.

- It gets cold really fast during the night at the location where I'm imaging, maybe that could be causing the scope or the mount to cause this problem in some way.

- I tried to image with dithering off. I first suspected that maybe APT was sending dither commands randomly during the image (although that should appear as a larger error in PHD2 guiding error)

- I thought it might be the focuser or focuser's draw tube moving due to the weight of the camera but that would mean that I'm losing focus so I've checked focus several times during the night with a Bahtinov mask and it never changed.

SO could the problem be the mount behaving strangely and not showing in PHD graph? some movement in the scope? bad collimation? flexure somewhere in the imaging train or guidescope? 

I'm really our of ideas so I was hoping someone could have some ideas or experience with this. The image I've attached might help in the analysis hopefully.

I'd appreciate any insights and recommendations.

Clear skies,

Anthony

Edited July 28, 2019 by Anthony RS

[almcl]

If the guiding graphs are good with no sudden excursions, and the scatter plot in PHD2 is reasonably round yet you get star trails, flex between guide scope and imaging scope is a possibility.  

You mention guide scope rings, are these the ones with plastic tipped locating screws?  If so, these alone are sufficient to introduce the sort of flex displayed in your post above.  Solid guide scope rings might help, or, better, switching to an Off-Axis Guider (there are Canon specific ones) would help to eliminate flexure between guide scope and imager.  An OAG will also help if the mirror is moving in its cell as, for example, when the scope tracks through or close to the zenith.

Can you post a guidelog from a session where you got the weird stars?  

[david_taurus83]

I'd say its guiding error or as above, flexure. Your imaging at 1" so in an ideal situation youd want your guiding to be sub 0.5"RMS and that includes the peaks! Can you check a few guide logs? Ignore the average RMS and take note of the peak errors. It only takes one jump to 2.0" to give issues.

[Anthony RS]

27 minutes ago, almcl said:

If the guiding graphs are good with no sudden excursions, and the scatter plot in PHD2 is reasonably round yet you get star trails, flex between guide scope and imaging scope is a possibility.  

You mention guide scope rings, are these the ones with plastic tipped locating screws?  If so, these alone are sufficient to introduce the sort of flex displayed in your post above.  Solid guide scope rings might help, or, better, switching to an Off-Axis Guider (there are Canon specific ones) would help to eliminate flexure between guide scope and imager.  An OAG will also help if the mirror is moving in its cell as, for example, when the scope tracks through or close to the zenith.

Can you post a guidelog from a session where you got the weird stars?  

Yep the rings have those plastic tips. My main issue is that I've been imaging with the same gear without any problems and suddenly I got this. Why would flexure suddenly appear out of nowhere. I always make sure the screws are extremely tight to avoid flexure. I will try to post some logs but the problem is that I was messing with the scope while guiding (checking for loose stuff, cable drag, etc...) so the log will have huge spikes that we won't be able to distinguish from actual guiding spikes and me messing up with the scope.

[Anthony RS]

16 minutes ago, david_taurus83 said:

I'd say its guiding error or as above, flexure. Your imaging at 1" so in an ideal situation youd want your guiding to be sub 0.5"RMS and that includes the peaks! Can you check a few guide logs? Ignore the average RMS and take note of the peak errors. It only takes one jump to 2.0" to give issues.

I've never had this issue even before I hypertuned the mount. I used to have >1.2" error with huge 4" spikes in the RA, and never got this exact issue. Now i have sub arcsec guiding with no spikes (very few spikes now and then but they do not always coincide with the bad images). Plus my reasoning is that even if there are spikes, they are usually really short and not enough to register such a bright second star?! It looks more like the scope moved entirely during image acquisition with no guiding correction which does not seem to be the case according to the guiding graph. 

[almcl]

2 minutes ago, Anthony RS said:

 Why would flexure suddenly appear out of nowhere. I always make sure the screws are extremely tight to avoid flexure. 

Can't answer that one, I am afraid, but it happened to me after some weeks of blameless guiding, suddenly, there it was.   It might have been 'mirror flop', or guide scope flexure or a combination.  If the PHD2 graph is solid for an image with bad stars, my money is on one or both of the two foregoing.

Incidentally, now that I've had a chance to plate solve your image of the Veil, it looks as though the two bottom images are erring in Dec, while top left has jumped in RA:

Balancing'east heavy' can be tricky close to the zenith (if that is where the scope was pointing) and can produce this effect, but so too can failing to rebalance after a meridian flip (ask me how I know!)

[Anthony RS]

1 minute ago, almcl said:

Can't answer that one, I am afraid, but it happened to me after some weeks of blameless guiding, suddenly, there it was.   It might have been 'mirror flop', or guide scope flexure or a combination.  If the PHD2 graph is solid for an image with bad stars, my money is on one or both of the two foregoing.

Incidentally, now that I've had a chance to plate solve your image of the Veil, it looks as though the two bottom images are erring in Dec, while top left has jumped in RA:

Balancing'east heavy' can be tricky close to the zenith (if that is where the scope was pointing) and can produce this effect, but so too can failing to rebalance after a meridian flip (ask me how I know!)

That's very interesting Thank you! Most of my images tend to be similar to the 2 bottom images so perhaps I'm having some Dec issues. But what's confusing me is that my guide graph looks normal 😕  Here's a screenshot I took, unfortunately I don't have one showing the whole graph and the logs won't be of much help. I guess I'll have to narrow it down to flexure and/or dec error. Thank you that really helped.

[Anthony RS]

6 minutes ago, almcl said:

Can't answer that one, I am afraid, but it happened to me after some weeks of blameless guiding, suddenly, there it was.   It might have been 'mirror flop', or guide scope flexure or a combination.  If the PHD2 graph is solid for an image with bad stars, my money is on one or both of the two foregoing.

Incidentally, now that I've had a chance to plate solve your image of the Veil, it looks as though the two bottom images are erring in Dec, while top left has jumped in RA:

Balancing'east heavy' can be tricky close to the zenith (if that is where the scope was pointing) and can produce this effect, but so too can failing to rebalance after a meridian flip (ask me how I know!)

Btw how does pointing to the Zenith cause this effect if east heavy balancing isn't as it should. If you mean it could cause the RA axis to wobble, then that should appear in the guiding graph but it's not. 

[Anthony RS]

11 minutes ago, almcl said:

Can't answer that one, I am afraid, but it happened to me after some weeks of blameless guiding, suddenly, there it was.   It might have been 'mirror flop', or guide scope flexure or a combination.  If the PHD2 graph is solid for an image with bad stars, my money is on one or both of the two foregoing.

Incidentally, now that I've had a chance to plate solve your image of the Veil, it looks as though the two bottom images are erring in Dec, while top left has jumped in RA:

Balancing'east heavy' can be tricky close to the zenith (if that is where the scope was pointing) and can produce this effect, but so too can failing to rebalance after a meridian flip (ask me how I know!)

One more thing if you don't mind, how did you plate solve the image and how did you know where the DEC and RA directions point?

[almcl]

7 minutes ago, Anthony RS said:

 If you mean it could cause the RA axis to wobble, then that should appear in the guiding graph but it's not. 

at the scale your screenshot shows, you might not see RA wobble.  Try reducing the Y axis figure to 2 and seeing what the RA line is doing?

 

5 minutes ago, Anthony RS said:

One more thing if you don't mind, how did you plate solve the image and how did you know where the DEC and RA directions point?

I fed your image to nova.astrometry.net and after it solved the image  http://nova.astrometry.net/user_images/2965169#annotated  (took a few minutes) I followed the link to the Worldwide Telescope.  The submitted image is overlaid on the view and the RA and Dec axes can be superimposed from the sidebar.  For amusement (can you tell it's overcast here?) I also got All Sky Plate solver to run it, using a guess for the focal length and image scale but its result (after 76 seconds) isn't as easy to read on my monitor (although it's great for solving in the field as it doesn't need an internet connection).

[Merlin66]

You can superimpose your solved image in Cartes du Ceil and then click on any/ all star images to get the detail.

 

[david_taurus83]

1 hour ago, Anthony RS said:

That's very interesting Thank you! Most of my images tend to be similar to the 2 bottom images so perhaps I'm having some Dec issues. But what's confusing me is that my guide graph looks normal 😕  Here's a screenshot I took, unfortunately I don't have one showing the whole graph and the logs won't be of much help. I guess I'll have to narrow it down to flexure and/or dec error. Thank you that really helped.

Change the y scale to 4 to give you a better idea of what's going on.

[glowingturnip]

another thing to check - don't choose a guide star close to any other star in PHD, eg don't choose one of a binary pair.  PHD can get confused between the two stars and keeps flipping from one to the other as its centroids change.  PHD won't report anything wrong, and the guiding graphs will look good, but the subs will all show a smear in the same angle and magnitude as those binary guide stars.

[Anthony RS]

8 minutes ago, glowingturnip said:

another thing to check - don't choose a guide star close to any other star in PHD, eg don't choose one of a binary pair.  PHD can get confused between the two stars and keeps flipping from one to the other as its centroids change.  PHD won't report anything wrong, and the guiding graphs will look good, but the subs will all show a smear in the same angle and magnitude as those binary guide stars.

I didn't think of that thank you! But i doubt this is the issue since I've tried many different targets and different stars to guide on. But I'll keep that in mind. Thanks!

[alan potts]

I don't know the reason but for about half hour one night I had something very similar, I had chosen the guide star (may have been a double), asked the program to then select and have not seen it since. I didn't see anything in the guide graph but was looking on a much larger scale Y was 4, it never went outside 1second. On 16 mine appear a stright line.

Alan

[michael8554]

I see "choose guide star" in several of the preceding replies. 

You don't choose the star, PHD2 does that with the Auto Select. 

That way you don't get double stars selected, or stars too close to each other, or bloated stars, or underexposed stars. 

Of course, if you think you know better than PHD2 what constitutes a suitable star, go right ahead....... 

Michael 

[Anthony RS]

6 hours ago, michael8554 said:

I see "choose guide star" in several of the preceding replies. 

You don't choose the star, PHD2 does that with the Auto Select. 

That way you don't get double stars selected, or stars too close to each other, or bloated stars, or underexposed stars. 

Of course, if you think you know better than PHD2 what constitutes a suitable star, go right ahead....... 

Michael 

I always let PHD choose the star. That's not the issue.

[glowingturnip]

as do I, but it has chosen doubles for me in the past