Could this be sensor tilt?

[Gumminess8083]

Ok, so, I recently took these images of the heart nebula and, while processing, I realized that some stars were oval-shaped, but only in the bottom left corner of the image.

Top image is the bottom left corner, bottom image is the right corner.

I am shooting from a Redcat 51 through a ZWO ASI294MM PRO and a ZWO filterwheel with Astronomik filters. If it matters, I'm also focusing through an autofocus motor.

In this specific case, I was shooting narrowband, but I have also included an image I took with my green filter (to add RGB stars to the final image), and the problem seems to be present there as well.

Sample images; all of them have this same problem.
https://drive.google.com/drive/folders/1m8M4SOKaKj0EK6-BjCPDDdFMCPIhYcfy

Could someone help me figure out if this is caused by sensor tilt or some other factor?

Thanks in advance!

[knobby]

Looks like tilt ...

Use ASTAP as it's free, shows an approximation of tilt.

This is your Ha from the google drive.

[Gumminess8083]

10 minutes ago, knobby said:

Looks like tilt ...

Use ASTAP as it's free, shows an approximation of tilt.

This is your Ha from the google drive.

Thanks for confirming my doubts and for letting me know how to do that!

In your opinion, would I be better off trying to fix the problem myself, or should I contact ZWO support and ask for a replacement? (I bought this from a third party reseller)

[knobby]

Is everything screw fit ? don't know the redcat, but could the tilt be anywhere else ?

[Gumminess8083]

1 hour ago, knobby said:

Is everything screw fit ? don't know the redcat, but could the tilt be anywhere else ?

The only moving parts are the spacers, camera and filter wheel.

The redcat does have a built in tilt plate, but using that would force me to keep the camera in the same orientation at all times.

I would love to run more tests, but the forecast tells me that it's going to rain for the next 2 weeks

[Gumminess8083]

1 hour ago, knobby said:

Is everything screw fit ? don't know the redcat, but could the tilt be anywhere else ?

https://drive.google.com/drive/folders/1SasrJO-5gHq4gyVQnYGWShdAhXL5uB4d

I have uploaded here another picture I took with the same scope but different camera and shorter image train (no filter wheel).

 

This does seem to have some tilt as well, but not in the same way (?)

[alacant]

Hi

Before suspecting the camera, perhaps better to dismantle, reassemble and have a look at the tilt again at various camera angles.

Often, simply screwing stuff back together will cure (or change) the tilt.

Cheers

 

Edited November 1, 2022 by alacant

[Gumminess8083]

1 hour ago, alacant said:

Hi

Before suspecting the camera, perhaps better to dismantle, reassemble and have a look at the tilt again at various camera angles.

Often, simply screwing stuff back together will cure (or change) the tilt.

Cheers

 

Thanks for the suggestion!

As soon as I have a clear night, I'll try taking everything apart and putting it back together to see if it makes a difference

[pipnina]

Can you post a raw file or use the PixInsight abberation inspector tool to let us see the center of frame, and the four corners easily?

Some of the stars look like their core is off enter, which might be coma?

[alacant]

15 hours ago, Gumminess8083 said:

I have uploaded here another picture

Much better. The bloat introduced when the fw is attached is probably a combination of astigmatism and tilt. Without the fw, it is visible top right. But hey, almost nada and easily corrected in software.
fw and attachments: Clean the threads with very fine emery cloth whilst in the proximity of a vacuum cleaner tube. 

Cheers and HTH

 

Edited November 2, 2022 by alacant

[Gumminess8083]

10 minutes ago, alacant said:

Much better. The bloat introduced when the fw is attached is probably a combination of astigmatism and tilt. Without the fw, it is visible top right. But hey, almost nada and easily corrected in software.
fw and attachments: Clean the threads with very fine emery cloth whilst in the proximity of a vacuum cleaner tube. 

Cheers and HTH

 

Unfortunately this picture was taken with a completely different camera (my old and now defunct Sony a6000), so it was only meant to verify whether the problem lied with the scope or the new camera (ASI294MM PRO)/optical train.

 

I'll try to run some new tests as soon as I get a clear night

[alacant]

45 minutes ago, Gumminess8083 said:

taken with a completely different camera

OK, post a shot of the train with the 294. Guessing: the camera is screwed directly to the fw. As the spacing doesn't matter on the 51, you could try e.g. adding a tube on the camera side. Same using a different spacer on the telescope side. Trust the orthogonality of nothing!

Cheers

[newbie alert]

Problem with these sensor tilt measurement softwares is they don't measure the full sensor, only so far from the axis.  The tilt in this case is in the bottom left corner, top right seems fine.. so to determine if it's indeed a tilted sensor or focuser sag/ misalignment or even a slight optical misalignment is a tough call... You can eliminate if the sensor is tilted by building yourself a jig as Olly had on his thread... Stuart on here has just made a video on the subject and has confirmed that it's cured his issues

https://youtu.be/Gw2oTwb4GjY

Edited November 2, 2022 by newbie alert
Added info

[Gumminess8083]

2 hours ago, alacant said:

OK, post a shot of the train with the 294. Guessing: the camera is screwed directly to the fw. As the spacing doesn't matter on the 51, you could try e.g. adding a tube on the camera side. Same using a different spacer on the telescope side. Trust the orthogonality of nothing!

Cheers

 

This is my imaging train, as per what was written in the ASI camera's manual (excuse my unorthodox autofocus mounting mechanism 🤣)

You are correct in assuming that the camera is threaded directly into the filter wheel!

[Gumminess8083]

44 minutes ago, newbie alert said:

Problem with these sensor tilt measurement softwares is they don't measure the full sensor, only so far from the axis.  The tilt in this case is in the bottom left corner, top right seems fine.. so to determine if it's indeed a tilted sensor or focuser sag/ misalignment or even a slight optical misalignment is a tough call... You can eliminate if the sensor is tilted by building yourself a jig as Olly had on his thread... Stuart on here has just made a video on the subject and has confirmed that it's cured his issues

https://youtu.be/Gw2oTwb4GjY

Stuart did in fact message me yesterday with a link to his video, so I'll have to give it a later today!

Just in case, I did notice that my redcat had these screws:

 

 

Since in the imaging train these come after the filter wheel and just before the connection to the scope itself, could I possibly use these to try and fix the tilt (provided it's actually caused by the camera and not the cope itself)?
Unfortunately this would prevent me from using Stuart's method, as the tilt adjuster is connected directly to the scope and it cannot be removed, as far as I know

Edited November 2, 2022 by Gumminess8083

[gaz81]

This isn’t exactly super technical… but I had (maybe still have) tilt issues. 
 

Mainly I think because I took things apart but didn’t put it back together again properly. So pretty bad stars in one corner. But that corner would change when the camera was rotated. So seemed not to be the camera itself (1600MM pro). 
 

So I had a laser collimator kicking around, I collimated that as best I could and then fitted it to the scope with a paper target on the end… and it was way off centre. Reseated things and it’s now more or less in the middle… I also turned it carefully in the scope to make sure it didn’t draw a big circle! just need to clear nights to test it. 
 

Maybe not the best method ever… but will have to wait and see. 
 

Anyway, thought it might help. 

[Gumminess8083]

OK, I finally have an update!

After months without a single clear night, I was finally able to take some pictures yesterday and the day before

I have included all the measurements in the following album: https://imgur.com/a/T4GqSd8
 

All pictures were taken with the same setup, in the same position and without moving it. The first night was perfectly clear, while the second was quite hazy, but I decided to shoot anyways, just to see if I could get something decent out of it. I have also included some old pictures before "calibrating" the sensor tilt, to see whether it made a difference.

Now, I am not an expert, far from it, but, from what I can see, considering that the camera has been clibrated and should now be flat, the problem still seems to be present...

Could it be that the problem lies with my tracking? I tried using the guiding help tool in PHD2 for the first time, and it seems that my error is around 3400 ms.

If anyone has any further ideas, I'd be happy to try them out

Edit:
https://drive.google.com/drive/folders/1-H2JRX_zegJOhZemnRRRL9razu7LcvnZ?usp=share_link
 

I have uploaded some of my new shots from these past 2 days in here

Edit 2:
@gaz81 Thanks for the suggestion! Should I still attempt to collimate my scope, even if it's a refractor? I read somewhere that refractore don't usually need to be collimated

Edited January 15 by Gumminess8083

[gaz81]

Probable depends on the scope… I have more ‘affordable’ stuff where there isn’t any real adjustments it’s just more a case of refitting the focuser etc. 

[ollypenrice]

On 15/01/2023 at 12:19, Gumminess8083 said:

Edit 2:
@gaz81 Thanks for the suggestion! Should I still attempt to collimate my scope, even if it's a refractor? I read somewhere that refractore don't usually need to be collimated

Sometimes, but not all that often, refractors have easily collimatable front cell assemblies in which the whole group of elements can be tilted as one. I've seen some Altair Astro triplets which use antagonistic pairs of screws, parallel with the scope's axis and accessible from the front. These scopes are sold under other brand names as well. I've also collimated the front doublet element of a TeleVue Genesis on which the technique is to back off three radial screws round the lens cell, move the cell by hand, not with the screws, and fix it in position with the screws. I personally would only attempt these operations on a bench while observing an artificial star, and on an instrument with user-friendly adjustment facilities - which are the exception rather than the rule, I think.

What I would not do at any price is attempt to collimate individual elements within the cell.

If the scope gives a good star test I wouldn't do it at all.

Olly

[Stuart1971]

On 01/11/2022 at 15:09, Gumminess8083 said:

Ok, so, I recently took these images of the heart nebula and, while processing, I realized that some stars were oval-shaped, but only in the bottom left corner of the image.

Top image is the bottom left corner, bottom image is the right corner.

I am shooting from a Redcat 51 through a ZWO ASI294MM PRO and a ZWO filterwheel with Astronomik filters. If it matters, I'm also focusing through an autofocus motor.

In this specific case, I was shooting narrowband, but I have also included an image I took with my green filter (to add RGB stars to the final image), and the problem seems to be present there as well.

Sample images; all of them have this same problem.
https://drive.google.com/drive/folders/1m8M4SOKaKj0EK6-BjCPDDdFMCPIhYcfy

Could someone help me figure out if this is caused by sensor tilt or some other factor?

Thanks in advance!

There is also bad astigmatism there too, as you can see by the star shapes, they look like bloated crosses, try focusing about 2/3 from the middle of the image to even out the curvature, it should help, I don’t think that is all tilt at all….👍🏻

[Gumminess8083]

9 hours ago, Stuart1971 said:

There is also bad astigmatism there too, as you can see by the star shapes, they look like bloated crosses, try focusing about 2/3 from the middle of the image to even out the curvature, it should help, I don’t think that is all tilt at all….👍🏻

Is there any way to do it with an autofocuser? Or even to fix it at all?

I usually let an autofocus routine take care of it, so I would have to find a way to change the settings on there.

 

Anyways, looks like this weekend is gonna be pretty clear, so I will do some more tests, such as rotating the camera and taking some more measurements with shorter exposure times, so as to eliminate other variables such as tracking, etc.

 

Hopefully I'll be able to sort everything out, eventually!

 

Thanks again, everyone. I'll report back if I have news

[Stuart1971]

56 minutes ago, Gumminess8083 said:

Is there any way to do it with an autofocuser? Or even to fix it at all?

I usually let an autofocus routine take care of it, so I would have to find a way to change the settings on there.

 

Anyways, looks like this weekend is gonna be pretty clear, so I will do some more tests, such as rotating the camera and taking some more measurements with shorter exposure times, so as to eliminate other variables such as tracking, etc.

 

Hopefully I'll be able to sort everything out, eventually!

 

Thanks again, everyone. I'll report back if I have news

What software do you use, as NINA had region of interest autofocus, so you can choose an doughnut type area…looks like this, area inside the dough it is used for focus, and the inner and outer lines can be altered 

 

Edited January 16 by Stuart1971

[Gumminess8083]

20 hours ago, Stuart1971 said:

What software do you use, as NINA had region of interest autofocus, so you can choose an doughnut type area…looks like this, area inside the dough it is used for focus, and the inner and outer lines can be altered 

 

I do indeed use NINA, so I'll try that out next time!


I also have a small update—I remembered that I had taken some test shots while trying to figure out the best exposure time for my target the other night, which were taken before the meridian flip, while my normal LIGHTs were taken after.

Here's a comparison:

S filter before meridian flip

S filter after meridian flip

H filter before meridian flip

H filter after meridian flip

O filter before meridian flip

O filter after meridian flip

 

If the problem was caused by sensor tilt, I would see the "gradients" in FWHM always in the same corner, since the image would always be aligned with the camera.
In my current situation, the gradient seems to be flipped by 180° after the meridian flip, suggesting that this is not caused by sensor tilt nor collimation errors, but by sommething else altogether!

My best bet would be the guiding, as my mount is not that good at tracking currently. I should get a belt conversion kit for it soon, though, so maybe that will help me fix this problem.

...This is as far as my knowledge can take me. If there's some other factor at play, I cannot figure out what it could be, beside astigmatism.

Edited January 17 by Gumminess8083
a word

[Stuart1971]

2 hours ago, Gumminess8083 said:

I do indeed use NINA, so I'll try that out next time!


I also have a small update—I remembered that I had taken some test shots while trying to figure out the best exposure time for my target the other night, which were taken before the meridian flip, while my normal LIGHTs were taken after.

Here's a comparison:

S filter before meridian flip

S filter after meridian flip

H filter before meridian flip

H filter after meridian flip

O filter before meridian flip

O filter after meridian flip

 

If the problem was caused by sensor tilt, I would see the "gradients" in FWHM always in the same corner, since the image would always be aligned with the camera.
In my current situation, the gradient seems to be flipped by 180° after the meridian flip, suggesting that this is not caused by sensor tilt nor collimation errors, but by sommething else altogether!

My best bet would be the guiding, as my mount is not that good at tracking currently. I should get a belt conversion kit for it soon, though, so maybe that will help me fix this problem.

...This is as far as my knowledge can take me. If there's some other factor at play, I cannot figure out what it could be, beside astigmatism.

It could also be some sort of flexure in the imaging train or focus tube, as after the flip the weight will be shifted and so would the tilt….

[pipnina]

32 minutes ago, Stuart1971 said:

It could also be some sort of flexure in the imaging train or focus tube, as after the flip the weight will be shifted and so would the tilt….

Could be the front cell or the focuser aren't quite done up tight? I've had that issues before...