Focus issue or stacking issue?

[Pryce]

Hello!   I did 4 hours on the elephants trunk tonight and even though I focused with a bahtinov mask(Perfectly I thought) I've run into an issue.


In the in individual subs my stars look like this.

But when I stacked and did a quick and dirty edit they ended up looking like this



So my question is, did I mess up the focus or did I do something wrong during stacking?

 

[PeterCPC]

Did all the lights look the same - at the start and at the end? Focus can change.

[Pryce]

They did! I rechecked my focus twice during the 4 hours. The single sub example was from my last sub of the night.   But yes, all the subs look the same

 

[Pryce]

Here are my first and last subs side by side.

And here are two subs from around when I rechecked focus

[PeterCPC]

Have you tried just stacking the lights with no calib subs?

[Pryce]

Just now, PeterCPC said:

Have you tried just stacking the lights with no calib subs?

I have not tried that.   Wait one

[Pryce]

I should have waited until I woke up before I started processing..  Starting this now was a big big mistake! 🤣

[Pryce]

A stack of lights only still produced the same effect. @PeterCPC

Edit,  and now I'm going to bed before it's too late.  Looking forward to reading what the masters come up with when I wake up!

Edited January 14, 2021 by Pryce

[TerryMcK]

The first and last look slightly out of focus to me. Hot pixel seen in red at the top left of the frame. 

On another note reduce your ISO to 800 or less. There is no benefit from higher ISO apart from introducing more noise.

[ollypenrice]

That really is very odd. The subs you've shown us here look OK. We do need to be sure that no rogue subs from the run have crept into the stacking. This has to be the best guess but it may not be the case.

Next, what software did you use for stacking and which algorithm did you select? And is there any chance that you asked the software to do anything else during the stacking? Some additional processing step?

Olly

 

Edited January 14, 2021 by ollypenrice

[PeterCPC]

2 hours ago, Pryce said:

A stack of lights only still produced the same effect. @PeterCPC

Edit,  and now I'm going to bed before it's too late.  Looking forward to reading what the masters come up with when I wake up!

So it's not an issue with the calibration frames. As Olly says, how did you stack? Did you get any dew?

Edited January 14, 2021 by PeterCPC

[vlaiv]

That has something to do with cosmetic correction / hot pixel removal or similar

Try just regular average stack without fancy options - stars should look fine in that.

[Pryce]

I still got the same result with using Average on the stacking of the lights.

And no, I don't believe I had any dew for the night.



 

[Pryce]

I just tried a stack where I selected the recommended settings in DSS, and I believe that solved the haloing issue.

Buuuuuuut that brought another problem.  A weird gradient that weren't present in the other stacks. (Bottom right)

[Pryce]

Man, this hobby is hard 😂

[vlaiv]

46 minutes ago, Pryce said:

I still got the same result with using Average on the stacking of the lights.

Turn off all "advanced" options - like hot pixel detection and removal.

 

[Pryce]

Roger that!  Stacking,  wait one.

[Pryce]

The stars are now round,   but liek mentioned above,  it now has a weird gradient.  And now it's even worse

[vlaiv]

8 minutes ago, Pryce said:

The stars are now round,   but liek mentioned above,  it now has a weird gradient.  And now it's even worse

When doing star detection - don't use median filter - and raise threshold. How many stars are detected in test run?

[Pryce]

3 minutes ago, vlaiv said:

When doing star detection - don't use median filter - and raise threshold. How many stars are detected in test run?

[vlaiv]

DSS usually freaks out like that when it can't find alignment stars or does poor job of matching up stars.

Change parameters in hope that it will properly solve alignment info - maybe increase threshold, or reduce it - see if either helps.

[Pryce]

I really can't fiigure this out..  I've tried a bunch of different parameters, lowe and higher treshold, ticked and unticked settings.

Whenever I do get sharp stars, I get the weird gradient.. And then I get rid of the gradient, I get the star haloes.   I'm so confused.

Well,  time to go get new data and see if that's the issue I guess! 😂

[vlaiv]

Just now, Pryce said:

I really can't fiigure this out..  I've tried a bunch of different parameters, lowe and higher treshold, ticked and unticked settings.

Whenever I do get sharp stars, I get the weird gradient.. And then I get rid of the gradient, I get the star haloes.   I'm so confused.

Well,  time to go get new data and see if that's the issue I guess! 😂

Try different stacking software as well?

 

[Pryce]

That's a good idea!   

Anyone in perticular I should give a go?

 

[vlaiv]

Just now, Pryce said:

That's a good idea!   

Anyone in perticular I should give a go?

 

Hopefully someone will come along and recommend good alternative.

I can name 3 - two that I heard people using, and third that I use myself, but that one I would not recommend as it is too complicated for regular use.

Siril and Iris are two often mentioned (I think I've installed both but never used either for stacking)

I use ImageJ / Fiji - but that is too scientific for everyday use as you need to manually do calibration and choose bunch of parameters when registering images. You also need to do manual stacking - so not really user friendly for astronomy.