53 replies to this topic

[imtl]

Andrey

 

Its good that you used guiding assistance. I have a few points for you. Hope it helps you.

 

1. Your Dec is more than twice as large as your RA. That means two things. Either you still have some serious Dec problems or you are still not guiding in RA.

 

2. Your guide star looks extremely weird. Did you properly focus your guide scope? In case you are using an OAG you might have a spacing problem. If its a guide scope then something weird is going on over there. You seem to be having some serious aberrations there. Make sure you are focused with your guide scope and try and see why you get that star shape. It might throw PHD off from doing proper guiding analysis and from actually guiding. 

 

3. Your image looks a bit out of focus. 5 minutes sub for a very good mount (Ioptron 120?) would be able to yield good stars. But, if you are not in good focus then it will hide any tracking and guiding issues. But that is because you're not in focus!

The other issue is that with 5 minutes L you are probably saturating a lot of the stars, which again causes bloating and therefore will hide guiding errors.

 

If I can suggest something is that you pull back, get your focus in order, do 60-120s subs unguided so you can judge if your mount is working properly. Then add guiding to the puzzle. And take 60s images of large star fields so you can judge any errors you have. Either from camera, or mount or whatever.

 

No point of imaging many nights while your system needs adjustment and tweaking. Learn your setup, go step by step and then go to the big stuff. 

 

PS. now that I think about it. Are you doing proper calibration of PHD2?

 

That's my opinion anyways, good luck

Eyal

Edited by imtl, 08 May 2020 - 01:09 AM.

[Andrey Kobelev]

Eyal, thanks much for your advice! I appreciate it!

 

Yeah, learning curve is not easy  

as for guiding star - Yeah it is also good moment - it is Lodestar X2 - it has not square pixels, it might be the reason - I do not not. Or it might be again wrong driver. I did not collimated guiderscope - just using it as it came from the store - it is  6-8 months old. Do not even touch it as for now.

 

As for the picture - yes, still learning. might be fucsing better in the future. But it is step up from my images two days ago. the scope also might contain some aberrations, since I collimated it by myself I am pretty much beginner and it was my first collimation of the scope ever.

 

So, I expect it will be a lot of re-iterations of reviews guiding-scope-guiding-scope, etc.... I agree the picture is not perfect.

 

thanks much!

 

Best,

Andrey

[imtl]

Eyal, thanks much for your advice! I appreciate it!

 

Yeah, learning curve is not easy

as for guiding star - Yeah it is also good moment - it is Lodestar X2 - it has not square pixels, it might be the reason - I do not not. Or it might be again wrong driver. I did not collimated guiderscope - just using it as it came from the store - it is  6-8 months old. Do not even touch it as for now.

 

As for the picture - yes, still learning. might be fucsing better in the future. But it is step up from my images two days ago. the scope also might contain some aberrations, since I collimated it by myself I am pretty much beginner and it was my first collimation of the scope ever.

 

So, I expect it will be a lot of re-iterations of reviews guiding-scope-guiding-scope, etc.... I agree the picture is not perfect.

 

thanks much!

 

Best,

Andrey

Andrey,

 

I encourage you to keep trying and especially since you seem to be enjoying yourself also keep the good spirit. Good results come with persistence and high motivation.

 

I'm a beginner like you and trying to help out.

 

So, a few things.

 

1. It seems like you are in the middle of an imaging session and also writing stuff here. I suggest you focus on one and leave the other for another time. If you want to write and understand what is going on then do it when you finish your session. Focus on the scope and mount when you are out imaging.

 

2. I am having a bit of trouble following what you were writing about your guidescope. And I think its probably because you are in the middle of imaging so need to focus on that. However, What do you mean that Lodestar2 has no square pixels? Also, I'm am guessing that "collimating the guidescope" refers to my comment about calibrating PHD2. So,..

 

Lets get things in order so we can try and help you. 

 

A. You need to check your guide scope and camera to see what is going on. Focus your guide camera the best you can. You're using Lodestar x2 and a 430FL guide scope. What is your imaging scope? That way we can **** your imaging/guide scope ration. After focusing your guide camera properly you should just reset all your PHD2 settings and erase any calibration that it did and start from scratch. 

 

B. Slew to a star in the area of where celestial equator crosses the meridian and do a PHD2 calibration. Read the manual how to do that. If you have questions then ask.

 

C. Once calibration is done. Slew to an object you would like to image and run guiding assistance. Let it do its thing and see what are the results. PA, backlash etc. If all looks okay then accept the results and star guiding. Do 60s exposures on a star field and what the results. If all looks good then increase to 120s. Then increase to 180s. Keep going until you start seeing problems. That way you will know you CURRENT LIMIT. 

 

D. Once you get that then its time to start tweaking and improving.

 

E. If you get to about 5 minutes exposures with no issues with everything in focus then you are off to a very good start. This will allow you to image many many objects.

 

F. Forget about dithering for now. Focus on the basics.

 

That is what I would do if I were you.

Eyal

[Andrey Kobelev]

Yes, stars are saturated   how do you determine the best exposure length? I actually paid almost no attention on star saturation, wanting to take out more details of the M101 Galaxy.

[imtl]

That is a different topic to discuss. There is a LOT of information about this on CN forums. Very much depends on your imaging gear and what objects you are trying to image.

 

Let keep this post focused on your guiding and mount. I believe there are fundamental things to address first before starting to discuss best exposure lengths etc. You should open another post on that if you want. Its better to keep things focused in a post in my opinion.

 

Eyal

[james7ca]

The images of your guide star looks somewhat misshapen (almost like an unresolved double star). I noticed that in your first post but attributed it to just being a one-off fluke, but you're second image shows an even more pronounced effect. Can you explain that? Guide stars don't need to be round (they often aren't with an OAG), but you say you are using a separate guide scope which should produce a fairly round-looking guide star.

 

It may not make any practical difference but if it were me I'd try to understand why the guide star image is misshaped.

 

As for your indicated RMS error in R.A, I'd agree that your 0.12" seems somewhat suspect. But, that's coming from a 0.03 pixel measurement which is probably pushing the limits of any practical guiding scale. I'd generally be a little leery of any measurement that is notably below 0.1 pixels (guiding scale). Maybe you can trust down to 0.05 pixels, but probably not much further.

 

In any case, as has been mentioned earlier the real "test" is in how your images look. If you are getting small and round stars then your guiding is good enough regardless of what is indicated by PhD. Similarly, if you stars aren't round then it probably doesn't matter much if your PhD measurements are "good" (although non-round stars aren't always the result of poor guiding -- optical aberrations being a common cause).

 

Then there is your seeing conditions. It's much easier to guide when the seeing is good, not so easy with poor seeing.

[ChrisWhite]

As others have mentioned, your DEC error is still twice what your RA error is.  That is probably why your stars are not round in your exposure.  It's peculiar to me that DEC is worse than RA.  Most of the time its the other way around.  DEC really isnt moving when you are imaging, just RA as it tracks the sky.  DEC only needs to move to correct for things like polar alignment error, etc...

 

When looking at your guide graph there is always an overcorrection with DEC.  So additional PHD settings can be tried.  Increase the min/mo.  0.1 is pretty small of a setting.  You could also lower the aggression.  Try something like 60 or 70 instead of 100.  This will temper the correction strength.

 

If this is an iOptron CEM60, I would suggest tightening the thumbscrew that engages the DEC axis.  Maybe there is a little slop there.  If it's the CEM120, I'm not sure how that mount works, but if there is a similar adjustment I recommend that. 

 

If it's a mount other than iOptron that you can adjust the meshing of the gears, I recommend that you make that adjustment and try to tighten things up a little bit.

[Andrey Kobelev]

Fellows, thanks much!

 

I agree with what you all said. Perhpas my next step to dig deeper is a form of the star @ guider. I agree looks strange. However, I autoselected the star - it was PhD choise. The form of the star can perhaps affect all PHD2 algrithms and abuse them somehow,m bringing that unsual RMS result. this is definitely my next step to discover and py much more attention.

 

As for not perfectly round stars on the image obtained - PI estimated this session in the range of 0.5-0.65 eccentricity. But I agree - although they are a way better i had previously - they are not yet  perfect.

 

here might be two reasons for that:

1. I might collimated the scope terribllu wrong. form of the stars on the picture is function of the scope not only of the mount.

2. Here is another one issue - I have not even touched here. it is a way complicated for the begineer - still in discovering mode - making more confirmations  - so not even want to mention it for now.

 

So, hitting a reasonble form of taguiding star at PHDs - is my next objective. Then will come DEC overcorrections and settings.

 

Thanks much!!!

 

Best,

Andrey

[StephenW]

It would be really helpful to see the PHD2 guiding log, including the GA run.

 

As Chris mentions, DEC should not be moving at all, so the fact that it is 2x worse than RA is very surprising.   

 

With perfect RA tracking (which you basically have with the encoders on RA), both DEC and RA should be limited by your seeing.  The latest graph you posted indicates that your seeing averaged over 2s has an amazing (unbelievable?) RMS of 0.2", with a peak of only 0.9".   Where are you imaging from? The top of Mauna Kea?

 

As a test, I would suggest increasing your DEC min-move to 0.4 and increasing your exposure to 4s and seeing how DEC behaves.

 

And again, uploading an actual guide log would go a long way to help understanding what is going on.

[imtl]

I think before Andrey fixes his guide star issue there is no point to look at anything. If the star is causing PHD2 to spit out silly things then non of this is really relevant to anything.

Eyal

[StephenW]

I don't think there's any reason to think PhD is spitting out "silly things". The only question I would have is whether the image scale in PHD2 is accurate (and so whether the reported 0.2" RMS is accurate), but that would have zero impact on DEC being 2x RA RMS.

FWIW PHD2 is actually very good at determining the guide stars centroid, even if it is out of focus or not fully round.

Andrey - fell free to try out some different settings and then upload the guide log so we can see what is actually happening.

[Andrey Kobelev]

Yep, I do believe scaling is accurate.  the guiding results were obtainted by the following:

 

1. Guiderscope AT 72II ED w/ 430 FL (it is added to PHD2)

2. Lodestar X2 w/ ~ 8.4 um pixel

 

So it is indeed about 4'' per pixel.

 

Anyway - I fully agree w/ Eyal I must have a deeper look at the guiding star shape as it appears in PHD2. Perhaps talking to other Lodestar X2 guider owner will help.

 

thank you!

 

Best,

Andrey

[imtl]

I don't think there's any reason to think PhD is spitting out "silly things". The only question I would have is whether the image scale in PHD2 is accurate (and so whether the reported 0.2" RMS is accurate), but that would have zero impact on DEC being 2x RA RMS.

FWIW PHD2 is actually very good at determining the guide stars centroid, even if it is out of focus or not fully round.

Andrey - fell free to try out some different settings and then upload the guide log so we can see what is actually happening.

 

If the star shape looks like that then I would say that PHD is not doing right. This is not a case of soft focus that PHD can handle. The star shape is completely off. I agree with you it has nothing to do with the DEC issue.

 

Things need to be investigated from the basics here before continuing in my opinion.

[StephenW]

>If the star shape looks like that then I would say that PHD is not doing right

In what way is PHD2 not "doing right"?

 

>I agree with you it has nothing to do with the DEC issue.

Ok - so what will change in the guiding results if the non-round guide star shape is "fixed"?

 

In cases like this it is important to focus on the facts: the non-round guide star-shape is not affecting RA guiding at all.  What makes you think it is affecting DEC?

Edited by StephenW, 08 May 2020 - 11:16 PM.

[Kevin Ross]

Yep, I do believe scaling is accurate.  the guiding results were obtainted by the following:

 

1. Guiderscope AT 72II ED w/ 430 FL (it is added to PHD2)

2. Lodestar X2 w/ ~ 8.4 um pixel

 

So it is indeed about 4'' per pixel.

 

Anyway - I fully agree w/ Eyal I must have a deeper look at the guiding star shape as it appears in PHD2. Perhaps talking to other Lodestar X2 guider owner will help.

 

thank you!

 

Best,

Andrey

Didn't you say you're also using a 2x barlow?

[Kevin Ross]

>If the star shape looks like that then I would say that PHD is not doing right

In what way is PHD2 not "doing right"?

 

>I agree with you it has nothing to do with the DEC issue.

Ok - so what will change in the guiding results if the non-round guide star shape is "fixed"?

 

In cases like this it is important to focus on the facts: the non-round guide star-shape is not affecting RA guiding at all.  What makes you think it is affecting DEC?

The guide stars (in the star profile window, as well as the stars in the main view) look almost doubled in one direction vs the other. The guiding is doing the opposite of what people expect (expect RA to be worse than Dec, but instead getting unbelievably perfect RA, with Dec nowhere near as good).

 

It might be a coincidence. But we don't know. Start at the beginning. Find out why the guide stars look so weird, when using a guide scope (not OAG). Fix that, then re-evaluate guiding.

[imtl]

>If the star shape looks like that then I would say that PHD is not doing right

In what way is PHD2 not "doing right"?

 

>I agree with you it has nothing to do with the DEC issue.

Ok - so what will change in the guiding results if the non-round guide star shape is "fixed"?

 

In cases like this it is important to focus on the facts: the non-round guide star-shape is not affecting RA guiding at all.  What makes you think it is affecting DEC?

Stephen,

 

Kevin already responded to this but I will say that, if the guide star looks that bad then trusting the PHD algorithm to do well is something that needs to be proven. So far I have not seen any proof of that. 

 

This is not a case of "non-round" star or soft focus. You can clearly see that there is some serious aberration there. So I do not understand what is the point you are trying to make actually. Clarifying that will help me to understand you. Stating that PHD will work no matter what doesn't mean much. It is clearly not doing right. Either by user error or by bad guide star or both. So where do you see that PHD is working well? I'll be happy to learn new stuff on PHD. I'm a beginner myself.

 

You cannot say that it is not affecting RA guiding at all because first, again this is not just a bit of an oval star. This is completely deformed star. And second, there is no real guiding in RA. If you really think he is getting 0.1" RMS in RA then okay. That is your opinion. There is no evidence for that and definitely not in his image. If this is true then great, I would be very happy to learn how to get 0.1" RMS in RA with a pickle shape guide star and out of focus system. So basically I can just focus my system by hand and use my eyes as a guide camera. Not need to spend thousands of dollars. I might actually do a decent job myself. Not joking! That is how they used to do it decades ago before all this technology!

 

If you want to share an example like this that you succeeded in getting good images with actual tracking like this on this kind of a guide star I'll be extremely happy to learn from your experience.

 

Eyal

[rgsalinger]

I've been using PHD as well as several other guiding methods for 10 years now. Star shape is not critical. It's always better to have sharply focused stars but I routinely get .3 or .4 arc seconds or RMS error with awful looking stars. However, this is a guide scope. If the guide stars are funky in a guide scope then there is an optical problem of some kind, the guide camera is faulty, or the scope is moving (less likely). 

 

Second, there is no guiding method that can calculate the centroid to an accuracy of 4/.12 - 1/32nd of a pixel. So that value is certainly wrong - it's an artifact of some kind. Of course I don't know what it is. About the best you can expect is about 1/8th of a pixel with accuracy. Remember, at 4 arc seconds/pixel, if the guide scope is focused properly will be just one pixel. 

 

Third, the Lodestar X2 has problems if you do not bin it 2x2. It has what the manual describes as "interlacing issues". This gives stars a venetian blind effect and can affect guiding. I know I've had this problem. I suspect that this is the cause of the funky stars but I can't tell. A full guide frame would be helpful. 

 

So, I'd like to see full specifications on the system and a full PHD guide log and then I bet that a number of use can figure out what the problem is. 

[james7ca]

...Third, the Lodestar X2 has problems if you do not bin it 2x2. It has what the manual describes as "interlacing issues". This gives stars a venetian blind effect and can affect guiding. I know I've had this problem. I suspect that this is the cause of the funky stars but I can't tell. A full guide frame would be helpful. 

 

So, I'd like to see full specifications on the system and a full PHD guide log and then I bet that a number of use can figure out what the problem is. 

That's correct, some models of the Lodestar do output interlaced video and some have claimed that can cause problems. But even in the worse case I'd think that could never cause more than a fraction of a pixel of error. However, that doesn't really explain the apparent drift in Dec, although it certainly wouldn't help. But, maybe the doubled appearance of the guide star is being caused by the interlacing.

 

I have one of the original Lodestar guide cameras but I no longer use it with separate guide scopes because it produces too coarse of a guiding scale with its large pixels. So, I only recommend the Lodestar for use with OAGs where you have plenty of image scale (like on an f/10 SCT). In the case of a long-focus scope with an OAG the Lodestar's large pixels are an advantage and you can certainly get away with a 2x2 binning to further increase the camera's sensitivity (and eliminate any issues with its interlacing).

 

In terms of image scale, yes, the AT72ED will produce an Airy disk that is just under 4 arc seconds (in theory) which seems uncomfortably close to the guiding scale of just over 4 arc seconds. That could be one of the reasons why the RMS error appears so low in RA, although PhD is reporting a FWHM of over two pixels which would suggest enough sampling to guide on. But, I agree that an RMS of 0.03 pixels is suspect.

Edited by james7ca, 09 May 2020 - 02:02 AM.

[imtl]

That's correct, some models of the Lodestar do output interlaced video and some have claimed that can cause problems. But even in the worse case I'd think that could never cause more than a fraction of a pixel of error. However, that doesn't really explain the apparent drift in Dec, although it certainly wouldn't help. But, maybe the doubled appearance of the guide star is being caused by the interlacing.

 

I have one of the original Lodestar guide cameras but I no longer use it with separate guide scopes because it produces too coarse of a guiding scale with its large pixels. So, I only recommend the Lodestar for use with OAGs where you have plenty of image scale (like on an f/10 SCT). In the case of a long-focus scope with an OAG the Lodestar's large pixels are an advantage and you can certainly get away with a 2x2 binning to further increase the camera's sensitivity (and eliminate any issues with its interlacing).

 

In terms of image scale, yes, the AT72ED will produce an Airy disk that is just under 4 arc seconds (in theory) which seems uncomfortably close to the guiding scale of just over 4 arc seconds. That could be one of the reasons why the RMS error appears so low in RA, although PhD is reporting a FWHM of over two pixels which would suggest enough sampling to guide on. But, I agree that an RMS of 0.03 pixels is suspect.

James,

 

I got confused by your last comment. I thought the AT72ED was the guide scope. Not the imaging scope.

Eyal

[Andrey Kobelev]

Fellows,

 

I substituted Lodestar X2 for my QHY 178m.  Although singal is not that good as w/ Lodestar - it seems that PHD2 guiding is OK now!!!!!!!!!

 

Today's guiding and RED 10 mins single frame.

 

Thanks much!

P.S. - I know here are a lot of things to do - to play w/ settings etc.

P.P.S. Ross can keep both his hats

 

Best,

Andrey

Attached Thumbnails

• 
• 

[Andrey Kobelev]

Yep

1. No Barlow ( I removed it as it was suggested in the previous post). So only 430 FL

2. AT 72II ED is a guider scope. Imaging scope is different. it is refractor a bit bigger

3. Now scaling: Guider - 1.2'' per pixel. Imaging scope about 0.7'' per pixel

 

 

Perhaps the issue was undersapling..

 

Best,

Andrey

Edited by dron2015, 09 May 2020 - 02:20 AM.

[james7ca]

James,

 

I got confused by your last comment. I thought the AT72ED was the guide scope. Not the imaging scope.

Eyal

The AT72ED is the guide scope (as I understand).  However, the guiding scale is just over 4 arc seconds per pixel with the Lodestar and the Airy disk formed by the AT72ED is also around 4 arc seconds so as Ross suggested the guide star (in theory) would be just a one pixel sample. (which is what I termed being "uncomfortably close"). 

[imtl]

Yep

1. No Barlow ( I removed it as it was suggested in the previous post). So only 430 FL

2. AT 72II ED is a guider scope. Imaging scope is different. it is refractor a bit bigger

3. Now scaling: Guider - 1.2'' per pixel. Imaging scope about 0.7'' per pixel

 

 

Perhaps the issue was undersapling..

 

Best,

Andrey

Happy for you Andrey. You can do a 2x2 bin with your guide camera and you would still be fine with imaging/guider scale. It will help you a lot with get more guide stars.

 

Your DEC aggressiveness is still too high in my opinion. take it down to 30-50 and see how it goes. 

 

So I guess star shape that looks like a split sausage does affect guiding...

Good luck

Eyal

[imtl]

The AT72ED is the guide scope (as I understand).  However, the guiding scale is just over 4 arc seconds per pixel with the Lodestar and the Airy disk formed by the AT72ED is also around 4 arc seconds so as Ross suggested the guide star (in theory) would be just a one pixel sample. (which is what I termed being "uncomfortably close"). 

Thanks. Got you now.