Guiding issues.

[Adina Fetche]

Dear all,

I have experienced occasionally this problem with guiding and it is ruining my imaging session for that night if I am not paying attention. I have noticed that if a cloud passes, occasionally. The first step in this drift was around 30 arcsec which is a lot and happened in a matter of seconds, so I can't blame the polar alignment( which was done with SharpCap and was excellent).  

   

I have noticed in a number of occasions that, when cloudy,  PHD2 starts looking for a guiding star and the guiding square moves randomly on the frame. Any idea how to stop PHD2 start looking randomly for a guiding star and maintain the green square in the same position? ( unfortunately I do not have a record of this bizarre behaviour, but the ultimate outcome of it is a frame looking like this) 

I have highlighted the first step( when the star trail is missing) and the period with clouds cover on the graph. I have attached the GuideLog the  DebugLog is too large.. 

 

[Bruce Waddington]

Please take a look at the explanation here:

https://openphdguiding.org/man-dev/Trouble_shooting.htm#Problems-_Lost_stars

Your problem is likely to be that you are essentially trying to guide on sensor noise rather than real stars.  You should be using a dark library or a bad-pix map and your Min-HFD value should be set high enough to filter out hot pixels and sensor noise.  Going forward, please use the requested procedure to upload both the guide and debug log files to our server:

https://openphdguiding.org/getting-help/

If you're trying to do unattended imaging, you should also be using one of the many image session manager apps that are commonly used - those are able to recover from cloud problems. 

Bruce

[mj.w...@gmail.com]

Hi Bruce

I couldn't help but notice that RA and Dec Minimum Move settings were essentially zero:

 "Minimum move = 0.010". 

About 0.005arcsecs.

And the poor Cal.

And borderline star mass most of the time.

Would  a better Cal, increasing the Min Move, Binning  (Pixel scale = 0.44 arc-sec/px), and longer exposures help ?

Michael

Wiltshire UK

[Adina Fetche]

Dear both,

Thank you for the answers. This is the link for the guidelog and debug log : https://openphdguiding.org/logs/dl/PHD2_logs_MDDi.zip.

Regarding the software I am using SharpCap, it does PlateSolving in the Imaging Session, but I did not include a PlateSolving loop to check position after each frame.  

Regarding the imaging session, in the first part of the last guiding episode( the 4 h and 30 minutes) I had a reasonably good guiding and the SNR was in the range of 10 -19 so I don't really understand why increasing the exposure would help. You said that I am trying to guide on noise rather than real stars. What would be an acceptable SNR? To me a SNR higher than 10 seems fine and even PHD2 comes with a standard setting of 6.  I mean I can go to higher values, but when the issue arises and the cloud rolls in, the SNR would still be low causing abnormal movements(assuming that it was an odd agglutination of several noisy pixels looking like a faint star causing a high amplitude movement of 30 arcses or so in the rectangle) and besides this, with these set values it usually works fine.    

As you can see  my target SNR is set 6 but during normal session I have SNR bigger than 10 as it is shown on the graph. Could the tolerance-set at 80%- be the issue?

I always thought that the PHD2 will not act if whatever it sees has a SNR lower than 6, but the first abnormal correction was done at SNR 4.2 and further corrections were done to even lower SNRs (in the 3.something range )

(please disregard what is written in the snipped SNR value, I have marked the  SNR value which triggered the move) 

I don't really understand why increasing the minimum movement would help when the clouds come( and as far as I understand it it is the minimum movement to apply correction). The poor calibration looks ugly, but from my experience it does not really influence the guiding. It is usually from a gust of wind during the calibration, I can repeat it easily, but I didn't bother at that time and, as you can see, on the guiding graph, I was having decent RMS( look at the values on the left, the SNR values were above 10(usually 16 or15)): 

Best regards

[Bruce Waddington]

I think we have multiple conversations going here.  I was responding to your initial complaint which was basically asking why PHD2 was moving the scope around during cloudy periods.  I said that you were probably guiding on sensor noise rather than a real star,  which is the case.  Now that you've uploaded the necessary logs, I can be more specific.  You have an over-sampled guiding configuration of 0.4 arc-sec/px which means your stars are very large.  During normal guiding, they appear to have an HFD (half-flux-diameter) of over 6 pixels.  The critical setting in PHD2 to avoid guiding on sensor noise is the Min-HFD value, which is what I told you in my response.  Yours is set to 1.5 pixels.  That clearly doesn't make any sense - you're telling PHD2 that a 1.5 px HFD "object" on the sensor is probably a usable star even though your configuration could never produce a star that small.  You should set the Min-HFD value to something that aligns with the smallest legitimate stars your guiding system can produce - nothing smaller.  That will deal with your original complaint although doing unattended imaging without a session manager than can do cloud recovery is a futile exercise.

I think Michael was offering comments based on what he saw as whack-o guider settings - and he is right.  So here is a litany of things you're doing that don't make any sense, which of course you can choose to ignore if you're already completely happy with your results:

1.  You should bin the guide camera 2x2 to get the image scale above 0.5 arc-sec/px.

2.   You should start doing multi-star guiding and even more importantly, stop trying to choose a single guide star by clicking on it.  Use the auto-select mechanism in PHD2 which will let you take advantage of multi-star guiding. That will be even more effective if you do step 1.

3.  Run the Guiding Assistant and accept its recommendations for Min-Move values.  Yours are ridiculously small and result in guide corrections for nearly every guide camera exposure.  This, coupled with using only 1-second exposures, means you are chasing the seeing.

4.  Start using the Calibration Assistant and see if you can improve on the fairly poor results I see in the logs. Yes, it matters.

5.  Disable the star mass detection feature, it isn't suitable for your situation.

6.  Restore the Max-duration settings for both RA and Dec to their default values.  In fact, restore all the guider settings to their default values.  If you upgrade to the 2.6.13dev5 release and use the new-profile-wizard to create a new configuration profile, all of these things will be sorted out.

Regards,

Bruce

[Adina Fetche]

Point taken for the Min-HFD, I missed that and will sort it out. I will disable the star mass detection features as well, I suspect that with Min-HFD allows agglutinated hot pixels to mimic stars to prevent For the rest, I need explanation in terms of comparison of outcomes in between different value settings and FWHM. Most of the things you wrote there applies for your regular 500mm focal length nebula imaging session where faint details get lost anyway and FWHM in the main frame doesn't really matter.

I would go even further down with the exposure, real time, if could.  I saw the Best Practice comparison on the main page, and indeed maybe the 4 seconds graph looks better in PHD2, but I don't really care how my graph looks like, I care about the FWHM in the main frame.  As it is written there, in PHD2 the stars will average out, but the same will happen in the main frame at the cost of FWHM and image sharpness and that defeats part of the purpose of guiding. PE also averages itself out if you take long enough exposures. You get the elongated stars because the Dec doesn't move that much, but RA averages itself out. 

What I do probably doesn't make any sense to you because you do not know what the target is,  the target is not even visible in the main frame I posted at the beginning of the topic and that is an 1 hour subs. The apparent size of the target is 22 arcsec that is about the size of Saturn without rings and the apparent magnitude is 22 so it is well hidden in the light pollution.   

I have stretched this stack of 9 hours (-1h, 30 min and 15 min subs-, I left aside some additional 6 hours (I need to see if they are suitable enough for stacking). If you look closely, you might see an Einstein ring, (at least in the coloured version taken from Sky-map.org- mine is showing up a little if you put at least 50 cm from the display-still quite noisy)  Now, if you can give me some piece of advice that will lead to a direct decrease in my FWHM in the main frame, I am happy to take it, but otherwise, I do not want to change things for now as it seems to be ok from my point of view.

Best regards.  

[Brian Valente]

It seems like you are wanting to apply DSO-appropriate guiding (PHD) to basically planetary type of target? 

If that's the case, why not just switch to lucky imaging or use metaguide. From what you're describing (and your response to the suggestions) it's the wrong tool for what you are trying to accomplish

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/68fb2c22-cb54-442b-8f75-5bbe7726b6afn%40googlegroups.com.

--

Brian 

Brian Valente

astronomy portfolio https://www.astrobin.com/users/bvalente/ 

[Adina Fetche]

I don't really understand why you would not classify an 8 billion light year away, 22 apparent magnitude target as a DSO. I don't think you can get a more DSO than that.  I don't think I want to do a 1 hour exposure subs as a lucky imaging. I think PHD2 as a part of an imaging session is science and science has questions and reproducible results. I don't really understand why my question about FWHM  in the main frame and exposure times it is not regarded as a reasonable question. You guys have so much experience, I am sure there must be a way to compare the outcomes of 1 sec and 3 sec for example in FWHM results in the main frame. 

Best regards 

[Brian Valente]

>>>The apparent size of the target is 22 arcsec

What is that target, and what's the resolution in arcsec you need to resolve the detail you require? 

 

What is your imaging instrumentation and image scale you are using to achieve the above?

>>>I don't really understand why you would not classify an 8 billion light year away, 22 apparent magnitude target as a DSO. I don't think you can get a more DSO than that.

I did not say the target was not a DSO, I said you are attempting to employ DSO-style guiding (slower cadence focused on guiding out mount-related noise without chasing seeing) on a target that appears to be extremely small and require extremely fine resolution (although still waiting for that detail above). I was suggesting a different approach may lead you to the results. 

From your image example, I don't see the details you refer to re: Einstein Ring. So unless I misunderstand what your goal and results are so far, I don't think your current approach is working very well.

You turned down the recommendations from Bruce, citing a reluctance to do this in favor of more rapid corrections. Are you familiar with metaguide and have you considered using it here? 

>>> I am sure there must be a way to compare the outcomes of 1 sec and 3 sec for example in FWHM results in the main frame. 

That's an easy enough experiment, you can do it yourself. The results will be tied specifically to your conditions and equipment, so it will be far more accurate for you. 

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/4281284b-18a6-4049-8a13-477f0d4590ban%40googlegroups.com.

[Adina Fetche]

The reluctance in doing some of the advice given by Bruce derives from the lack of purpose. He said to bin guiding the camera. I am happy to do it but why? As it is the star I use for guiding has at 1 sec a SNR of 14-16. Is that not enough to guide?  Will binning the camera show me the target star through the clouds? No way, so to me Advice 1 seems a bit useless to my problem.  

Advice 2:   You should start doing multi-star guiding and even more importantly, stop trying to choose a single guide star by clicking on it.  Use the auto-select mechanism in PHD2 which will let you take advantage of multi-star guiding. That will be even more effective if you do step 1. -This can be useful if I am willing to increase the exposure times, which I am not, not for such a dim target and -again- it is not why I started the topic. I had a very specific problem, the drift during the cloud cover, I do not think doing advice 2 will help me in my problem.- I would rate this a rather useless advice. 

Advice 3:    Run the Guiding Assistant and accept its recommendations for Min-Move values.  Yours are ridiculously small and result in guide corrections for nearly every guide camera exposure.  This, coupled with using only 1-second exposures, means you are chasing the seeing.- another advice which will not prevent the drift during the cloud cover. And yes, I am chasing the seeing for the reasons I have already discussed. Unfortunately, I regard this advice again useless-definitely not helping my cloud problem, but I am happy to be proven wrong. 

Advice 4:  Start using the Calibration Assistant and see if you can improve on the fairly poor results I see in the logs. Yes, it matters. Now, this advice is something else. What is fairly poor result? I am asking now: what is a poor result? What RMSE is too much on 10mph wind? How about 7 mph wind? 80% humidity, 90% humidity? What is a poor result?  How can someone say I have a poor result if they didn't see the FWHM in the main frame? How can someone say I have a poor result if they didn't see the atmospheric conditions, winds and so on?  This is just sad coming from an expert. Even my kid does a more thorough analysis of the results than that. I am being kind grading this advice as useless, again not helping with my problem. 

Advice 5:  Disable the star mass detection feature, it isn't suitable for your situation. Good advice. I will either do that or at least decrease the star tolerance by less, lets say 20% or so( will see how it behaves) 

Advice 6:   Restore the Max-duration settings for both RA and Dec to their default values.  In fact, restore all the guider settings to their default values.  If you upgrade to the 2.6.13dev5 release and use the new-profile-wizard to create a new configuration profile, all of these things will be sorted out.- maybe, I need to check it out first- but not helping my problem 

7. The main advice regarding Min-HFD seems to be good advice, will do that too

Now that's a problem because I got 7 advices and only 2 are pertinent to my problem and that makes me lose confidence in the advice. For example, If you come in my clinic, yes?, with back pain and I am speaking about your blood pressure and your eye glasses, you would lose confidence, trust me. Anyway if you see a problem in my guiding graph which corrected will result in an improvement in FWHM in the main frame, I am happy to make adjustment. The 1,2,3,4 and 6 will not have that outcome, possibly 3 and 6 have counter effect. 

Bruce said to use a software that can recover from drift. Well, that is just not going to happen automatically. When I set the scope on the target, I go to a bright star in the vicinity then I move the scope where I know the target is to have that bright star for guiding. I can practice during a full moon period with setting up Phd2 automatically from SharpCap- but I don't have too much experience with that.

The target is SDSS J2238+1319. Has about 22 arcsecs. I have a 12 inch SW scope with f/4.9 reduced to f/4.4 by a field flattener coupled with a 2.4 mcm pixel size camera( about 0.37 arcsec/pixel)

I know about Metaguide, I did not try it, maybe I should, but I just don't have the time to do it.  Look, I live in the UK, getting 15 hours of fairly good sky with no Moon in a month is an exception. It takes me 20 minutes to focus after I started PHD2 before starting the session. I wish I had  more time to try more of the settings, but for now I just need to do what I know it works. If you guys come with convincing evidence that your advice works better for FWHM I am happy to copy paste the settings you propose, but you expert opinion is not good enough. I am sure you saw so many graphs and know all of this, but I am not sure if you coupled the graphs with the main frame outcomes. 

I have outlined the shape of the ring. It is still a work in progress, but my kids recognised the features. Put a meter in between you and the display sort of to reduce the noise impact, and looked at the native image I posted initially not at the rainbow one :)

Best regards.  

 

[Aris Pope]

Adina I'm glad to see Bruce was able to help. I dig the star spikes by the way! Love my Newts and RC's!

Aris

--

You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/d74910d5-9bd9-4f7d-bb9e-780db355d10fn%40googlegroups.com.

[Adina Fetche]

Yes, me too and thank you for that. I like them too, the big star near the ring is only 12.8 magnitude, but the ones above in the full frame are in the range of 8-9 magnitude.

Best regards

[mj.w...@gmail.com]

To use your doctor analogy:

You keep visiting the doctor because your temperature has risen 0.1degree. 

Which is smaller than the accuracy range of your thermometer and is not enough of a rise to be of concern.

Binning will lessen the chances of Star Lost events.

Raised Minimum Move and exposure will mean the mount is not constantly being guided when it doesn't need to be.

And an unacceptable Calibration is not a good baseline for an evening of good guiding.

Michael

Wiltshire UK

[wes McDonald]

Adina 

In my view your tracking requirements are outside of what you should expect from guiding.  In my experience and the experience of the hundreds of users posts I have read, guiding at best gets you to about .3”.  At best.   There are , as I am sure you know, many factors that influence how well one might expect to guide, and these include the obvious perturbations from the atmosphere, mount dynamics, mount driver implantation of pulse guiding, PHD algorithm and settings, guide camera pixel scale.  In the end one has to fiddle the knobs that exist to get to some combination of all changeable factors that yield sufficiently good guiding for the specific imaging requirement.   And your imaging requirements are tough, especially if you are using non research grade mount equipment.  I presume your observatory mount is not such a mount, as you must guide out its periodic error.  In principle there is no need to guide unless you have periodic error, assuming your equipment and software are up to the task.   Most amateur kits are not and thus we use the excellent additional solutions such as PHD to augment our hardware.

Given guiding is not the correct solution, I would suggest you look into a Telescope Driver Master.  The TDM provides a high resolution encoder for the RA axis with which the software measures the actual rate of the axis directly.  It provides ST4 guide pulses to the mount controller to adjust the rate 5 times per second.  This is obviously far faster than one could reasonably do by star tracking guiding.  The TDM of course has its own limitations, but it does close the RA rate loop which allows it to compensate for a variety of mount and mount driver related deficiencies.  The chief issue the TDM has is that it is open loop relative to star position on your image plane.  Noise in the rate measurement and imprecision in the mount ST4 response can lead to drift, as can slight timing error in the TDM due to crystal oscillator drift.  Having said that, the maker of the thing advises that 4 minute or less exposures be made. Obviously, depending on your image scale, your mileage may vary.

But the experience of one user I know of, who has used the TDM to manage a mount being used for exoplanet research, is that the TDM is very very good.   My own experience with it has been good, but my requirements are pedestrian relative to his and yours.  My lab testing of it (I used it to establish the precision of my mounts ST4 guiding) demonstrated the correctness of a 4 minute kind of time frame.  Unfortunately my setup has various uncontrolled factors, for example I was judging drift relative to my computer clock rather than a time standard.  But still in all my experience with such errors in actual application against the important time piece (earths rotation rate) leads me to accept the 4 minute result as a useful one.

Obviously, if you are not going to close the tracking loop on a star image, you will need precise polar alignment. Nowadays this is not particularly difficult using Sharpcap and other tools (as you do). But at the same time the mount limitations come into play here.  The TDM does allow ST4 Dec guide pulses however , so one can use PHD at a slow cadence to correct DEC drift.  Pulse guiding for the RA axis however cannot be used unless there are specific provisions in the mount controller, because the TDM will quickly adjust the RA rate to compensate for the pulse guide commanded rate changes.   It’s possible to envision how a mount controller could disable ST4 inputs when pulse guide commands were being issued, but I doubt any do that.

One further benefit of the TDM is that since it does not employ star images to establish an accurate RA sidereal rate, it is of course unperturbed by clouds.  Which was one of your chief complaints I guess.

Based only on your image in the thumbnail, I think I am old enough to give you a word of advice.  When people try to help you, especially people who know what they are talking about, and who selflessly spend hours each day responding to folks to help them with their issues in this hobby, it is best to not attack them.  It’s a turn off, and it reflects badly on you.   You should probably chill.  You don’t have to prove to the world how smart you are.  We get it.  But this is not Facebook or X.  Civility is preferred.

Now go ahead and flame me.  I am made of asbestos.

So that’s my 2 cents for you.  I wish you happiness and success.  

Wes

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6934844c-b35d-4bb8-acfd-1d904a6e28bdn%40googlegroups.com.

[Adina Fetche]

Wes and Michael,

I am sorry, but I do not see the discussion as you are presenting it. Wes- I did not flame anyone I was responding to a free criticism. Michael, to keep it on the doctor analogy , I came to see the doctors with my back pain and I did not mention anything about anything else, but I am being criticised because I use to much sugar in diet, for example, and tobacco products. But, I came with a totally different thing,  my cloud drift( back pain). Why do I need to be criticised for my exposure times and correction thresholds? They are not related with the drift during cloud cover, everybody knows that.  Returning to the doctor analogy, I can accept that critic because I know, tobacco and sugar are bad. But here, I am not given any proof that by doing what I do, I am either doing something bad which can lead to, for example:  my mount coming to harm or my FWHM are worse than it should be  or, at least, a  kind of evidence showing that guiding with 1 or 3 s is the same-which btw would make pointless to change the setting to begin with.

Returning to the doctor analogy, I would not go to have a surgery as long as I am fine and it is not foreseeable that I will get worse. Is the same thing here. If anyone wants to criticise my guiding style at least ask for a main frame to measure a FWHM. Ultimately, I do not use PHD2 to get nice graphs, I use it to get nice images. 

Wes, I do not want to invest in any fancy equipment, this is a hobby and it will remain like that. All I want, is to get the best out of my equipment, and, in this case, I think, nobody has a proven answer as there are to many factors that come at play. But, saying that, an expert should know how to handle challenges, I mean, this is why they are an expert.  

When I started the topic, I had a fair problem, which hopefully will be fixed. I did not flame, I did not criticise PHD2 or any features of it- which btw is an amazing software, I have no words to describe it, but at the end: We are doing scientific measurements here with our scientific equipment, when we take these picture, and  science needs to be proven otherwise it is not science.  

Best regards

[Bruce Waddington]

I think we're done with this thread.  If you had carefully read my response, you would have seen this disclaimer:

So here is a litany of things you're doing that don't make any sense, which of course you can choose to ignore if you're already completely happy with your results:

Most users appreciate being told about mistakes that may seem unrelated to their original question instead of having to learn the hard way and come back to get help one problem at a time.  Obviously, you don't appreciate that, which is fine.  I was intending to provide explanations behind the various pieces of advice but after reading your messages and noting their tone, I'm not going to waste my time doing that.  You can probably figure them out yourself, they are tied to basic principles of astronomical seeing, statistical sampling, mechanics, and optics.  In the meantime, by all means, murder ahead with what you're doing - just don't plan on seeing any further comments from me.

Bruce