Back Focus Issue?

[PhotoGav]

Thank you for your forensic assessment Vlaiv. I am not convinced that I am capable of improving that collimation. I am no expert and remarkably ham fisted when it comes to collimating this scope. If I get a rush of blood and am feeling brave, and the sky clears, I will give it a go. I have some pics of this when less out of focus which I will post next time I am at my computer. Out of interest, how much better should this be? As I understand it, I am trying to get the dark central portion as central as possible within the brighter outer rings. To me that looks almost perfectly central. I am not sure that I could improve it by 12 pixels!?! Should I be able to?

[vlaiv]

1 hour ago, PhotoGav said:

Thank you for your forensic assessment Vlaiv. I am not convinced that I am capable of improving that collimation. I am no expert and remarkably ham fisted when it comes to collimating this scope. If I get a rush of blood and am feeling brave, and the sky clears, I will give it a go. I have some pics of this when less out of focus which I will post next time I am at my computer. Out of interest, how much better should this be? As I understand it, I am trying to get the dark central portion as central as possible within the brighter outer rings. To me that looks almost perfectly central. I am not sure that I could improve it by 12 pixels!?! Should I be able to?

I'm not sure. 12 pixels does not sound a lot but if you take it as percentage - might be more meaningful.

I honestly have no idea of relation between larger defocus pattern and off center and small defocus pattern and off center. If it's proportional than it's probably not that big of a deal, but if relation is not simple proportion but off center stays roughly the same - then 12 pixels difference on 50 pixels total defocus pattern is almost 25%.

When you shrink your defocus pattern you should be getting center spot as bright spot - that one should be close to center of inner "ring" - and you can see that one by using less defocus.

If it looks like this then it's ok:

But if its more like this, then I think you need to tweak it:

Do put star in exact center of the field - this is where you want it perfectly collimated. If your star in center is at slight miscollimation - this means that it will be worse in one of the corners compared to others - as there is place on sensor where it is close to good collimation - but that spot is not center of sensor (and it should be).

[PhotoGav]

Here's the slightly closer in focus version. I can see that it is ever so slightly to the left of centre. I will have a go at correcting this next time it is clear.

[martin_h]

On 24/01/2019 at 10:31, PhotoGav said:

Here's the slightly closer in focus version. I can see that it is ever so slightly to the left of centre. I will have a go at correcting this next time it is clear.

When you've done that nip round and do my RC.

[knobby]

isn't the distance from reducer to chip supposed to be 105 mm ?

[PhotoGav]

1 hour ago, knobby said:

isn't the distance from reducer to chip supposed to be 105 mm ?

Yes, exactly. That is made up of the T-Adapter (without extension), a 5mm adapter and 50mm QSI camera backfocus. Though the 50mm backfocus is a point of contention, given the minus one millimetre stated in the QSI manual to compensate for the 3mm thickness of the Astrodon filters (are they even 3mm thick - I've never checked!).

 

[PhotoGav]

I took advantage of the unexpected clear-ish skies last night and tweaked the collimation:

Is that central?! My eyes went funny and I just couldn't see the wood for the trees...

I then tried an 1800s Ha sub of my next target - Jones-Emberson 1, aka the Headphones Nebula:

I was pretty happy with that. Interested to hear what you all think! Am I now ready to just get on and image? If the weather ever clears up properly!!

 

[wxsatuser]

Looks central to me but no expert.

Get a session in when weather allows and see how it goes.

[knobby]

56 minutes ago, PhotoGav said:

Yes, exactly. That is made up of the T-Adapter (without extension), a 5mm adapter and 50mm QSI camera backfocus. Though the 50mm backfocus is a point of contention, given the minus one millimetre stated in the QSI manual to compensate for the 3mm thickness of the Astrodon filters (are they even 3mm thick - I've never checked!).

 

Ah, so the T adaptor is 50 mm, I'm a DSLR bod so was assuming the T adaptor was a T ring ?

[PhotoGav]

2 minutes ago, knobby said:

Ah, so the T adaptor is 50 mm, I'm a DSLR bod so was assuming the T adaptor was a T ring ?

Exactly. The adapter is in two pieces. Use both when not using the reducer and unscrew it to use just the ‘orange’ end with the reducer...

[RayD]

1 hour ago, PhotoGav said:

are they even 3mm thick - I've never checked!

What a great point.  What a trusting lot we are! ?

[vlaiv]

That looks like good collimation - it is so much easier to tell when one can see individual rings - just observe spacing of each in relation to others.

I think you'll still need to do a bit of tweaking on other parts (collimation is good, so you don't need to spend additional time tweaking that in my opinion) - but how much exactly - you'll be able to tell after stacking exposures to get clean image (which btw will be excellent judging by that single frame).

[steppenwolf]

This is looking much better with regard to collimation as the corners all match quite well. I think you can move on to the spacing issue now.

[knobby]

Should add that your guiding looks spot on too, 300 seconds at 1400 mm !

[PhotoGav]

2 hours ago, RayD said:

What a great point.  What a trusting lot we are! ?

Just checked and it’s 3mm +/- 0.025mm apparently!

 

1 hour ago, vlaiv said:

That looks like good collimation - it is so much easier to tell when one can see individual rings - just observe spacing of each in relation to others.

I think you'll still need to do a bit of tweaking on other parts (collimation is good, so you don't need to spend additional time tweaking that in my opinion) - but how much exactly - you'll be able to tell after stacking exposures to get clean image (which btw will be excellent judging by that single frame).

 

34 minutes ago, steppenwolf said:

This is looking much better with regard to collimation as the corners all match quite well. I think you can move on to the spacing issue now.

Thank you for the feedback Vlaiv (if that’s your name?!) & Steve. Yes, it isn’t perfect yet, close but not spot on. So, on to spacing... I had a look at that before shooting the sub and removed the two spacers, one by one. The test shots looked pretty much identical to me, as if the spacers were making zero difference. I nearly removed the 5mm spacer too just to see if it made any difference, but I didn’t... perhaps I should?! I’m not entirely sure how to proceed at this point beyond trial and error, so all help would be most welcome, please.

11 minutes ago, knobby said:

Should add that your guiding looks spot on too, 300 seconds at 1400 mm !

Good, thanks! 1800 seconds at that focal length is a bit of an ask, especially in poor sky conditions, but the Mesu performed admirably. It was reporting error in the region of 0.6” to 0.7”, which is bad for the mount, but just about within the tolerance limits of 0.75”/pixel scale. I’m hoping that it will be significantly better when the perfect clear skies come back, soon!

[vlaiv]

27 minutes ago, PhotoGav said:

Thank you for the feedback Vlaiv (if that’s your name?!) & Steve. Yes, it isn’t perfect yet, close but not spot on. So, on to spacing... I had a look at that before shooting the sub and removed the two spacers, one by one. The test shots looked pretty much identical to me, as if the spacers were making zero difference. I nearly removed the 5mm spacer too just to see if it made any difference, but I didn’t... perhaps I should?! I’m not entirely sure how to proceed at this point beyond trial and error, so all help would be most welcome, please.

I still think you need to look into possible tilt issue, before finding right spacing. Look at these close ups of stars:

This is from bottom right corner of latest sub. These stars look as good as one can reasonably expect.

On the other hand, we have bottom left corner:

These look elongated. If it were all the stars that looked that way - I would suspect guiding issue, but you have proper stars in the other corner, so your guiding is fine!

Top left corner shows this:

That is little coma tail right there.

It's present along top edge as well, but pointing in different direction:

and

With these it's harder to tell - that is why I suggested to stack images - it will become clear then.

This means that you are at about place with distance, but since your sensor is tilted - different corners are at different distance to FF/FR and hence not all corners are at proper distance. This is why I feel you need to sort the tilt out to get equal stars in each corner - and then it is easy to dial in proper distance.

Btw, I'm Vladimir, but Vlad is perfectly fine, Vlaiv is just a screen name composed out of first few letters of surname/name that I've been using since ever on net (trust me, you don't want to try to pronounce my last name - most English speaking people find it a nightmare to pronounce Vlaisavljevic )

On a separate note, what sort of FWHM do you get on your subs and final stack? 0.75"/pixel is most likely oversampling, and you might consider binning your subs (or even better - splitting them after calibration and before stacking). Also, 0.6-0.7" RMS is quite good for HEQ5 (if that is what you are using, from your sig), but its high for your working resolution. You want your guide RMS to be about the half of your desired sampling rate (you don't have to worry about guiding if you are going to bin your subs, since at 1.5"/pixel, such guiding is sufficient).

 

[PhotoGav]

Thanks Vlad!

Mmm, Sensor tilt, how do you resolve that? Is it in the camera itself or from the optical train? Is this something that has plagued all my images or just with this setup I wonder? I certainly don’t want to tinker with things only to mess up my other set ups (mainly the Esprit 100, which appears to be working well). I am in new territory here and need to tread carefully!

As for guiding - I am now using a relatively recently acquired Mesu 200 (the HEQ5 has been confined to the observatory floor!). Part of the reason for the purchase was to improve guiding to improve results with the C8. If only I could have the sky conditions to let the Mesu perform to its true ability! As for over / under sampling... I still haven’t really understood all of that....

The forecast is meant to be better tomorrow night, so I think I will gather some data and see how results are looking.

[vlaiv]

Ahh Mesu 200

Poor skies are not excuse!

That beauty should give you exceptional RMS (if it's indeed due to seeing, try increasing guide exposure length - Mesu 200 is so smooth that you can comfortably go with 10s guide exposures).

Ok, let's leave whole resolution thing on the side for now.

I doubt it's in camera if you are using it on other scopes and did not have issue so far, so that is the last thing that you should check. What sort of connection to scope are you using? Clamping connection is very susceptible to tilt - and you should be looking into threaded connection if you are not using it already.

Besides clamping, tilt can be due to telescope / focuser (this is EdgeHD right?) are you using additional focuser or just regular one on scope (that moves mirror?).

 

[steppenwolf]

10 hours ago, vlaiv said:

I still think you need to look into possible tilt issue, before finding right spacing. Look at these close ups of stars:

When I first looked at this image, I was unable to get a full resolution version to come up but when I saw Vlad's cropped versions, I had another go - it took numerous clicks to finally get it but it was there!

The coma shapes at top left and top right simply show under-correction by the field flattener. The bottom right as Vlad says is about as good as you are going to get with this telescope and flattener combination but the bottom left is more of an issue and isn't coma but neither is it a disaster! Comparing your new image with the original at the start of this thread, you have clearly made an improvement to the collimation but the pattern of star distortion, having now got in close an personal, shows that the errors, although apparently diminished, remain there in the original places which would indicate, again as Vlad has suggested, that tilt is still present. If you were sited in the Atacama desert, you could afford to spend a lot of time sorting this but from the UK it becomes much more problematic. Would I leave it there? We both know that I wouldn't but sometimes, it is just fun to get out and capture some images and enjoy them for what they are and dismiss the pixel peeping!

Tilt is always a bit of a nightmare to resolve and requires a series of tests to determine where the issue is (hence my comment about the Atacama desert!)

1. Does the tilt change with different regions of the sky, both altitude and pier side? Are you collimating and capturing your final image at similar azimuth and altitudes? I am thinking mirror shift here

2. Does rotating the camera in 90 degree stages change the distribution of the star shapes? I am thinking actual sensor tilt, filter tilt, mechanical mounting tilt, flattener lens misalignment

You see, this is why I have considered taking up crochet..............

[knobby]

I love my C8 HD but have always wondered how much the moving mirror changes collimation and focus due to tilt.

I suspect the only way for extreme consistency is to lock mirror position - use an external focuser - then collimate and sort reducer spacing. But now I'm nit picking .

[knobby]

1 hour ago, steppenwolf said:

You see, this is why I have considered taking up crochet..............

Just ordering new book for the Mrs...

Making every cross Stitch count ?

[PhotoGav]

Vlad, Steve & Knobby (your real name?), thank you for your replies and continuing attention to this issue. I think we are at an impasse here now. I am pretty sure that the misshapen stars are down to a floppy mirror. The EdgeHD 800 is almost wonderful, but the fact that focus is achieved by moving the primary mirror has got to be madness! Every time the focuser is moved, even slightly, surely collimation is altered, however minimally, but it is constant, so the whole optical system is being shafted (literally!) every time. No wonder it’s not perfect. Reflector collimation is normally performed by accurately positioning the primary mirror and holding it securely in place. Oh dear, not with the Edge!!

Wiltshire is definitely not the Atacama desert! I won’t be carrying out all the various tests to identify exactly where the issue is coming from! Mainly because I don’t think I could even resolve them. As you say Steve, it all changes depending on where the scope moves to anyway! I think I have to accept the limitations of the system I have chosen to use. It has the great advantage of being 1422mm or 2032mm focal length - excellent for catching the apparently small targets. It also has the advantage of being in my observatory right now. I would love a long refractor, but the Finance Director has an alternative opinion to that!

The images produced by this system are perfectly acceptable despite being slightly imperfect. Hopefully some of the issues will ‘stack out’ or crop out anyway!

Thank you for your help.

Now, where do I order my copy of ‘Making Every Cross Stitch Count’?!?

[knobby]

Well discussed on cloudy nights ... Lovely scope but slightly flawed .

All said though, yours is producing lovely images that should look sweet after stacking / processing.

Hope you find some cloud gaps soon.

[cfpendock]

Hello Gav.  I have just stumbled across this post and found it very interesting.  Certainly some of your problems have been experienced by myself using my CPC1100 Edge which is pretty similar to your set up - albeit with less exotic filters....

I have found two things:

1. Most of my problems disappeared when I fitted a motorised focuser onto the visual back, so the mirror was effectively always locked in one position, focus being achieved by moving the camera assembly in or out rather than moving the mirror. Probably there are good optical reasons why this is not correct, but it works for me.

2.  I have never been able to obtain the crispness of stars using the CPC rather than my Tak refractor - allowing for difference in focal length.  The stars from the CPC always seem to have a certain "softness".

Chris

[PhotoGav]

Thanks Knobby. We could all do with some clear skies... it’s getting beyond a joke now. 

Interesting Chris, thanks for your comments. Yes, the Moonlite focuser on the visual back solution has blipped onto my radar several times, but I have always ignored it as it theoretically messes with back focus and image clarity. Theoretically. It also has issues with fitting in the reducer. As for star softness, I totally agree. All SCT images have that softness. Opportunity cost of value focal length, I guess. I am saving up for a long refractor...