Strange tilting issue (changes from sub to sub)

[04Stefan07]

I am getting a very strange tilt issue with my new imaging train. The tilt changes from sub to sub, it would make more sense if it was consistant throughout all subs.

The attached picture are 6 subs taken with the Ha filter over the space of around 10-12 minutes. As you can see the tilt gets worse and then improves and I can't work out why! 

My current spacing is:

• Camera: 6.5
• EFW: 20
• T2-T2: 2
• Spacer: 11
• M42-M48: 16.5

= 56mm

I know it's 1mm over, I was waiting for an adapter as I didn't have the correct one to get it to 55mm exact so next I will try 55mm. If that doesn't work I will add a tilt adapter but it makes more sense to add one if I am getting a consitent tilt throughout all subs.

Using a WO GT81 IV, 1600MM, 36mm Baader Narrowband filters.

Any idea what this could be or anything I can try? If you need any more information regarding my setup, let me know!

Thank you

[04Stefan07]

And just to throw it in there.

Here are four subs with OIII, similar issue.

[04Stefan07]

OK so upon further investigation I noticed my draw tube for the focuser had a bit of wobble in it. Seemed to be an issue that has popped up with various William Optics telescopes. I managed to tighten the tension screws and now there is no more play.

Next clear night I will give it another test and see if it improves. If not, what would be the next step?

[symmetal]

It's quite normal for there to be slight differences shown between subs when using CCDI. To get a more representative picture it's better to take 6 or more subs, load them into CCDI, select them all and then do a curvature/3D test which will then show the average results.

To check for tilt you should use a wideband filter, preferably L, so you can take short subs of say 15s to minimize other factors affecting the results. As you're using narrowband filters I assume your exposures are many minutes at least. Your varying results are more likely due to periodic error and/or guiding rather than tilt changing between subs.

Alan

[04Stefan07]

On 12/04/2022 at 12:51, symmetal said:

It's quite normal for there to be slight differences shown between subs when using CCDI. To get a more representative picture it's better to take 6 or more subs, load them into CCDI, select them all and then do a curvature/3D test which will then show the average results.

To check for tilt you should use a wideband filter, preferably L, so you can take short subs of say 15s to minimize other factors affecting the results. As you're using narrowband filters I assume your exposures are many minutes at least. Your varying results are more likely due to periodic error and/or guiding rather than tilt changing between subs.

Alan

Here is a combination of 6 subs Ha at 120 seconds. Still some significant tilt.

[04Stefan07]

9 Subs LUM filter at 15 secs, combined.

Tilt was in all one direction but one sub the tilt shifted.

[04Stefan07]

One last test.

10 LUM Subs, 15 secs each (combined).

I added a 1mm spacer, readjusted my field flat to ensure its at 7.1mm as per spec and recalibrated my filter wheel to ensure they were sitting correctly.

Close inspection in Pixinsight on each corner it looks like I might have to add another 1mm or so maybe? Have added the corners of one sub, stars look to be point outward which indicates it needs more space? Thoughts?

[symmetal]

Stars pointing towards the centre does tend to indicate more spacing required. You need to add spacing if there are filters and camera protect glass panels between the field flattener and the sensor, usually 1/3 the total thickness of the glass in the way, around 1.0 to 1.5mm.

Your CCDI examples are showing variations with the point of best focus being near the edge in some of them. This will give worse focus results at the opposite edge which may be diagnosed as tilt by CCDI. The 9 sub luminance one looks like it's almost focused equally around the center, just slightly better nearer the top. If you get the FF spacing optimum then the corners should get better anyway.

If you're using an autofocuser it's optimum focus position is set by the algorithm used, whether it's biased towards the centre or over the whole frame, and the distribution of suitable stars in the image, which can vary the point chosen.

Alan

 

[04Stefan07]

Appreciate the reply and advice mate.

I had a look at the spec of the filters I am using and they are 2mm thick. I am using the ASI 1600MM which takes 6.5mm. So if it's 1/3 of 2mm then I will need to add about 0.67mm correct?

So by the sounds of it the focus needs to be pretty much bang on to give more accurate results otherwise the software will interpret it as tilt.

Using the EFW btw.

Some of my results show a dark blue ring but in the middle its very bright, what does that mean? I checked the 3D plot and the middle of the image looks like a little mountain.