An issue i cant pinpoint

[Mark0]

Hello everyone, im having a problem with my imaging train. I can not quite pinpoint the issue but im thinking it must be the backfocus. Please see the attached images. They are 60s, 180s and 300s exposures of NGC281 which ive used PI's Aberration Inspector on. There is some trailing in the 300s exposure but thats not part of the problem. Stars may look blocky but thats because the frames arent debayered yet. 

The Stars on the sides and corners of the frames are elongated, i believe this is caused by backfocus not being perfect. Without the IDAS NB1 filter, this doesnt happen. This filter is 2mm thick and ive followed the rule of thumb of dividing filter substrate thickness by 3 (so 0.67mm) and adding/removing that number to/from the backfocus with spacers. Ive tried 0.67mm metal sheet i got from a metalworker friend but ive also tried 0.67mm 3D printed spacers.

Heres the problem: Nothing changes with either spacer and without any spacers. Focus was as good as i could get it with a bhatinov mask. 

Im using a TS ED80, TS Flat 0.8x reducer corrector, an IDAS NB1 Ha+Oiii filter and a self modified Nikon D3400. This also happened when i had an ASI585mc for testing. 
So, what do you guys think causes this? Is this just a limitation of my gear, or is my reducer-corrector faulty? 

EDIT: Something important i forgot to add is that i modded my D3400 and did NOT replace the stock filter with anything!

Edited August 24 by Mark0
Added important information

[alacant]

Hi and welcome
The tsflat 0.8 is very sensitive to spacing. For f6 over aps-c, we found 55.5mm.  To obviate the need to make spacing allowances for the filter, screw it to the telescope side of the reducer. Self modified so we don't know the exact spacing so probably best to use a low profile Nikon to m48 adapter and build the spacing using m48 rings from there. 

Or of course, a lot easier, just correct the stars in software;)
Cheers and CS

EDIT: Forgot to say that to remove any tilt, we removed the rotator and went for a simple push fit m54 to 2". To make it hold at all angles, we also needed to support the camera. I don't think any consumer level focuser would hold a camera satisfactorily at that distance. You should be able to do something similar for your dslr. Fortunately, relatively cheap fixes.

 

Edited August 24 by alacant

[michael8554]

Hi Mark

With these long exposures it will be difficult to separate tracking and guiding problems from the mix.

Having said that, the centre of the 60s exposure looks okay, as do some of the corners.

That suggests a small amount of Tilt.

Are they the same corners with the ASI585mc ?

Michael.

[Stuart1971]

That is not a spacing issue, as the stars are actually elongated across the whole image, and obviously the 300 second one is the worst, this is a tracking issue…

‘If it was a spacing issue the stars in each corner would either be all pointing to the middle with the centre stars being perfectly round, or pointing in a circle around the middle like the image below, and yours are neither of these…

 

[Mark0]

Michael, unfortunately i did not keep any of the raw data from the 585 but i remember well that they looked the same. I have just noticed something else tho. The Aluminium ring where lenses attach on my dslr body is not really that tight anymore. If i press on it, it moves around 0.2 to 0.4mm inwards but only left and right, not up and down. Gonna have to find out how to tighten it, preferably without taking the body apart...

[Mark0]

1 hour ago, alacant said:

Hi and welcome
The tsflat 0.8 is very sensitive to spacing. For f6 over aps-c, we found 55.5mm.  To obviate the need to make spacing allowances for the filter, screw it to the telescope side of the reducer. Self modified so we don't know the exact spacing so probably best to use a low profile Nikon to m48 adapter and build the spacing using m48 rings from there. 

Or of course, a lot easier, just correct the stars in software;)
Cheers and CS

EDIT: Forgot to say that to remove any tilt, we removed the rotator and went for a simple push fit m54 to 2". To make it hold at all angles, we also needed to support the camera. I don't think any consumer level focuser would hold a camera satisfactorily at that distance. You should be able to do something similar for your dslr. Fortunately, relatively cheap fixes.

 

That looks super solid. I have just found something else that may also play into the mix, and your solution would definitely fix both focuser tube and body lens attachment ring flexing. 

Unfortunately im not able to find any info on how thick the stock D3400 filter is. I have just messaged a seller on ebay that sells them, hoping he can answer that question for me. Thats all i did to the body tho. 

The Filter is already screwed at the Telescope side.

As for star correction in software, i found that using Deconvolution in PixInsight never gets me good results, and the Easy Processing Suit EZ Decon freezes everytime i try. Can you recommend me anything to try? 

[alacant]

1 minute ago, Mark0 said:

how thick the stock D3400 filter

If you use the build-it-up-from-a-low-profile-adapter method by adding spacers by trial and error, you don't need to know. Perhaps 53mm as a starting point and work upwards? But definitely get the flange fixed first otherwise the tilt will affect the spacing.

 

4 minutes ago, Mark0 said:

recommend me anything to try

StarTools

6 minutes ago, Mark0 said:

Filter is already screwed at the Telescope side.

Then you don't need to add or subtract anything to or from spacing.