Can anyone explain why my stars are shaped like this?

[5haan_A]

Hi,

Was imaging last night and whilst processing images I noticed this weird star shape when I zoomed in. It's something I didn't see when previewing the images whilst they were being captured because at first glance the stars seemed quite round. Is this some sort of tracking PHD 2, or PA issue? Usually when I have had PA issues my star shapes are much more linear, that's to say there is a bloat on one edge of the star. 

Attached is a PHD2 log fire as well

 

 

PHD2_GuideLog_2019-11-18_1954556.txt

[steppenwolf]

This looks like either a snagging cable, some machining debris in the drive gear or a loose component.

Snagging cable would be quite likely.

[kirkster501]

^^^^^Agreed.

Or wind was blowing and gusting a bit from one direction.  This has happened to me.

[ollypenrice]

What we see here are pretty good star shapes but with every star duplicated, the doubles offset from each other and superimposed, one larger than the other. This would be explained by the mount tracking well but with a sudden nudge of the OTA position during the sub. It is not drifting slowly, it is recording two quite distinct stellar images and spending longer in one position than the other. We are, therefore, looking for something capable of causing this effect.

The first thing we need to know is the orientation of the camera relative to RA and Dec to see if the angle between the double images aligns with one axis or the other. I image with my camera orthogonal to RA and Dec (a good practice for several reasons) and, as I expected, we find that your doubles lie precisely along one axis. I asked Registar to rescale your image and apply it over mine:

So your problem is a single-axis one and from the star charts we can see it's RA. As well as the possibilities mentioned above, backlash in the drives could very easily cause this. The mount could spend some time resting on one side of the backlash then flop over quickly to the other, so recording two stellar images.

While wind or snagging cannot be eliminated I would not expect either to affect only one axis or to hold the scope still for long enough to record a round star. Personally I think backlash in RA would actually give the best fit with the evidence but it is never possible to be sure.

Olly

 

 

 

[almcl]

You have quite an amount of periodic error (which PHD2 is mostly guiding out):

This might be associated with the worm period of the mount and might benefit from a bit of 'tuning'.

But another problem might be the 3 second guiding exposures.  Here's one example of an almost 5 arc sec error which develops between exposures:

That's an extreme example, but there are plenty of other, slightly smaller, ones.  Balancing 'east heavy' and dropping exposure to 1.5 secs (or such lesser amount that the guiding assistant suggests) might help.  

Your PA error at 7' isn't brilliant but probably isn't affecting this particular issue.

[5haan_A]

Hi guys,

Firstly, a big thank you for contributing to helping to understand why those star shapes could occur. 

I had a look at two random light frames that I captured. These frames have had some preprocessing done namely flat and dark subtraction. I think they present another twist in the story, because they don't seem to contain the same star shape as what the final stacked image showed, which leads me to believe that the problem occurred during the stacking.

 

What do you guys think? Does it rule out some of the potential problems that have been identified in this thread?

 

Attached are two light frames taken in .fit format

 

Best,

 

 

recon_norm_pproc_L_2019-11-18_20-30-30_Bin1x1_120s__3C.fit recon_norm_pproc_L_2019-11-18_21-30-13_Bin1x1_120s__2C.fit

[geordie85]

Have you checked every sub? Even if there is just 1 sub taken like the first image you posted, it will affect the entire stack. 

Best just to examine every sub to see when you had the jump and delete it

[alacant]

2 hours ago, 5haan_A said:

they don't seem to contain the same star shape as what the final stacked image showed,

Ah, ok. So the double stars appear only in the stacked image?

That happens when the individual frames have not been aligned, resized and rotated prior to stacking, so the stars on each frame coincide. Or maybe in this case, one errant frame - perhaps one with insufficient stars to be included in the registration process- has been stacked alongside the correctly registered frames.

Cheers and HTH.

Edited November 19, 2019 by alacant

[ollypenrice]

10 hours ago, alacant said:

Ah, ok. So the double stars appear only in the stacked image?

That happens when the individual frames have not been aligned, resized and rotated prior to stacking, so the stars on each frame coincide. Or maybe in this case, one errant frame - perhaps one with insufficient stars to be included in the registration process- has been stacked alongside the correctly registered frames.

Cheers and HTH.

This is certainly possible but it means that we must attribute the alignment of the duplicate stars along the lines of RA to chance. (Had the OP's camera been orthogonal to RA and Dec we might have suspected the stacking software of having erred in one up-down alignment which wouldn't be surprising.) However, coincidences do happen.

Olly

[5haan_A]

Hi,

So here is a stacked file with what I deemed were the naughty frames out. Looks much better and the star shapes look a lot more regular. This is still a work in progress but now my nights work looks like it might bear some fruit.

 

Thanks,

 

[steppenwolf]

How many frames had the fault? Can you attach the faulty frames here for analysis?