Please check For Guiding Error Or Backfocus distance Error.

[Singlin]

Here is a compressed image and link to a non compressed image of a 600 sec sub of the horsehead nebula.

https://www.dropbox.com/s/mnlkr9cosbm5b8a/Horsehead%20600%20sub.tiff?dl=0

As you can see the stars are not great.

Is it a problem with guiding or could it be something else like backfocus distance error or flexure?

My scope is a f4 10" Quattro and camera is an Atik 383L  and Atik filterwheel and Skywatcher f4 coma corrector.

[Owmuchonomy]

On my iPad it looks like a unidirectional egginess (not radial) and thus could be oscillation in one of the axes. Can you superimpose the RA and Dec axes in your light frame? That may give us more of a clue.

[Tinker1947]

I'm no expert but it looks a bit like Coma around the outside......

[Owmuchonomy]

Coma would be radial and it doesn't look like that on my device.

[Tinker1947]

Coma would be radial and it doesn't look like that on my device.

Your properly right.....

[Singlin]

This shows the image orientation:

[Owmuchonomy]

Not sure, but could be Dec oscillation or flexure of the guide scope vs main scope. With that Mesu mount I would guess Dec oscillation is not an issue? The aberration is not aligned directly with either axis so it's not do clear to me. Is the image cropped?

[Singlin]

I did however rotate it 90 degrees and flip horizontaly.

Below is how it was when opened with Maxim dl:

[frostynixon]

Hmm, how do you attach your guider and guidescope? Is it securely bolted down? 

[Oddsocks]

Hi Singlin,

Really difficult to diagnose this one, I think there is more than one problem.

Looking at your dropbox sub and zooming in to a crop from the four corners and the centre you can see that top left the stars are nicely round and this extends a good distance towards the centre but does not reach centre.

Bottom left through centre to top right show oval stars with a common axis of deformation as does the bottom right but to a lesser extent.

It is difficult to be certain but the oval star shapes seem very slightly biased in different directions at bottom left and top right which is indicative of incorrect spacing on the coma corrector.

 

My hunch would be that the camera is not orthogonal to the image plane and is twisted with respect to the yellow axis in the image below, possibly a misalignment between the coma corrector and the camera.

 

I don't think it is the mount because the image would show the same defect across the whole field if flexure or drive errors were present, you would not get one good corner and the rest of the image showing oval stars but to help you eliminate the mount you need to take a few images at the roughly the same declination of bright star fields where you can use shorter exposures and then evaluate them, my guess would be they will show the same star shape and distribution as you have posted above.

If the camera CCD to coma corrector is orthogonal then you would next need to look at complete collimation including secondary placement to ensure the camera is orthogonal to the image plane.

You were getting good images in the past so it is difficult to know what has changed, perhaps the focuser has become misaligned or lose or you have changed something else in your setup?

William.

[Oddsocks]

As an additional check I ran CCD inspector on the single sub and the stacked image, I have posted the results for the stacked image below.

I wouldn't rely on the numbers calculated as absolute fact due to the nature of the image defect but the curvature plot does seem to confirm the tilt axis of the CCD is away from the imaging plane as shown by the yellow line that I marked in the last post.

The 3D plot appears to show the flattener to CCD spacing may be slightly out given the slight dip in the centre of the plot but this might be normal for the flattener you are using and the curve is not that extreme.

I haven't seen any other plots for your scope and flattener combination to compare it to, the only way you could test this is by adding an extra 0.5mm to the CCD-flattener distance and run a star field test image through CCD inspector again, plus a second image with 0.5mm removed from the current setting to see which way the plot changes with flattener distance.

The collimation is only very slightly out, at least according to CCD inspector.

William

 

 

[Singlin]

Great work William!

Today I will remove my focuser and re-bolt everything down and make sure it is central.

I will also add .5mm to the spacing and take a sub of a star fiel with and without the spacing this evening.

Thanks for your expertise.

[Singlin]

Attached are one  20 s,300s,and 600s sub I took last night.

My guidind graph was very zig zaggy.

[ollypenrice]

I was sure the original problem wasn't guiding because the distortions were not uniform across the chip but William has done a great analysis and covered all that.

What is happening in software to wreck your stars in the later images? These are not raw images, they've had some pesky algortithm attacking them! Note the dark rings and surrounding squiggles. Is this DSS on the rampage again? (I never go anywhere near it myself...) It's vital, first of all, to get a totally unmolested linear image out of the camera so we can see what's what. What we're looking at here smacks of deconvolution but I can't be sure.

Olly

[NickMorris]

I'm no expert but from your 2nd trio of images, the 20s looks ok maybe a slight bit of camera tilt but from the longer exposures, there are some guiding/tracking issues.  Could this be differential flexure between guidescope and main scope?

Cheers

Nick

[Oddsocks]

I think the image stacking / deconvolution defects that Olly has seen are jpg compression artefacts, too much compression has created artefacts around the stars in the posted images.

I have run the 20s exposure through CCD inspector and posted the curve and 3D plot below, they won't be that accurate because of the jpg compression artefacts in the posted source image but it shows a change in total curvature which is an over correction to the original setup, the correct CCD flattener distance will lie half way between the original distance and the one used for this latest series, but before changing anything I would like you to post an uncompressed version of the 20 second image as a TIFF, or better, a FIT file for CCD inspector to work with.

The total tilt in the system is slightly worse than before, according to CCD inspector,  but I can't confirm this until you post an uncompressed FIT or TIFF version for it to work with.

Looking visually at the 300s and 600s images the star elongation is in the same direction across the full image which suggests guiding is the problem and I think Nick is right that this is possibly differential flex in the guide scope mounting or an overcorrection by the guiding software, have you re-calibrated the guiding parameters recently or do you run auto calibrate each session?

The CCD inspector checks on the 300s and 600s showed a totally flat image plane with some tilt in the same direction as previously but I would discount the results due to the very obvious guiding errors and haven't posted the results for those images here as it will just add confusion.

So post the 20s image again as a TIFF or FIT for me to recheck with CCD inspector and look to possible problems with guide scope flex or guiding calibration.

William.

 

 

[Singlin]

The above images were stright conversions from tiff to jpeg.

Below is an attached image link of the 20s tiff version.

https://www.dropbox.com/s/8fe6sgbnwbx5fxv/Capture_120_098t.fit?dl=0

[Singlin]

I think that it could be guiding issues with my finder guider and QHY5.

I have not been able to attach an oag to my scope with success.

I am thinking of buying the carbon fibre version of the Quattro or similar to help resolve the flexing issues.

[Oddsocks]

The FIT file has produced a good CCD inspector calculation this time and now the tilt axis in the curve chart corresponds with the 3D plot so you know it is accurate, CCD inspector can be easily mislead by the compression artefacts introduced when reducing images to jpg files.

The current CCD to flattener distance is almost spot on.

The tilt axis is accurately shown with a 6% rise towards the top left edge and the image plane has a 15% total curvature which is not bad for a fast newtonian.

It might be improved with a small change in flattener to CCD distance of about -0.1mm to -0.2mm but actually seeing the effect of a 15% curve and 6% tilt in your images would be very difficult and the biggest improvement would be to fix the guiding first and worry about fine tuning the spacing later.

If you want me run the CCD inspector on subsequent images you can PM me, or post here, I will keep monitoring the thread for a while....

William

 

 

[Singlin]

I can't thank you enough William for taking the time to run these images through CCD inspector.

I will try and concentrate on the guiding issues, Although I am not sure I can change much with my setup.

When I put my Skywatcher St 80 bolted directly on to the Tube rings my guiding is worse.

I use the finder guider on it's bracket firmly locked down.

[Oddsocks]

You are more than welcome...

Bit of a long shot but I see the Quattro has a focal length of 800mm and the ST80 has a focal length of 400mm, have you tried using the ST80 to guide with a x1.5 or x2.0 barlow to bring the two scopes closer together in FL?

William.

[Singlin]

^{My quattro has a focul length of  1000mm-}

^{I found the flexure to be greater with the ST 80.}

^{I have some alluminium plate in my garage that I might use to reinforce the inside of my tube around the focuser and guide scope attachment holes-}

[ollypenrice]

The problem is that the flexure might be mirror shift. There should be no problem making an ST80 bracket stiff enough to guide at a metre. I do this routinely on our TEC140.

It is important on the ST80 to check that the focuser is tight in the main tube (three radial screws) and that the extension tube is tight along with the focus lock.

Were you unable to make the right chip distance with an OAG?

Olly

[Singlin]

Hi Olly!

I took my St 80 apart and cleaned it and collimated it as per a couple of guides I found on the internet.

When assembling it again I tightened everything down. I will however re-check.

I will also tighten slightly the bolts holding my main mirror on it's cell as I previously loosened it to stop pinching.

I can't get the correct chip distance with my Skywatcher F4 Coma corrector as the distance from the 383 sensor to my 9mm oag prism is shorter than the minimum obtainable distance from the oag prism and my QHY5 guide camera.I attached my Oag directly to my Atik filter wheel and my coma corrector directly to my Oag, This gives me precisly 55mm back focus. I then added a 1mm spacer to take into account for the thickness of´the glass of my filters.

I will perservere and try and get my St80 working as well as reinforcing the inside of my scope tube around the focuser.

Thanks for the help.

[ollypenrice]

There has been a long and complex thread on the subject of adding and subtracting this famous mm. It seems that maybe you should have subtracted a mm, not added one!

Olly