Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Recovery configurations for PHD2 and SGPro?

  • Please log in to reply
9 replies to this topic

#1 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 8944
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 14 October 2019 - 05:10 AM

I'm not having much luck with guiding recovery - yet another failure last night due to a brief loss of a guide star in the early hours.

The conditions that I'm hoping to get resumption working with are temporary loss of guide star due to stray clouds OR a remote facility temporary roof closure due to either wind gusts or short term high humidity (their rulez). In the latter case the roof may reopen 15+ minutes later (no earlier).

I don't seem to be able get the two software components to work together well. Going on last nights logs the guide star was lost for 40 - 50 seconds (4 cycles of 10 sec guide exposures). During this period it seems from the PHD2 logs that it was commanded (by SGPro?) to stop guiding. Meanwhile in SGPro's logs it spent the next 4 hours at 1 minute intervals asking for the PHD2 status and getting back "stopped" - it was at some sort of impasse regarding who makes the first recovery move. I suspect I have something miss set.

I think I need a fresh slate - and start over again with all the recovery settings - so if anyone can share what PHD2 and SGPro settings that they know works - that would be great. TIA.


As a foot note - very occasionally my Lodedstar X2 guide camera errors via its ASCOM link - a manual disconnect/reconnect of just the camera in PHD2 resolves this and guiding can be restarted. Can this solution be automated in PHD2? I.e. is there a recovery-on-camera-error feature hidden away somewhere? 

 


Edited by Tonk, 14 October 2019 - 08:31 AM.


#2 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 8944
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 15 October 2019 - 12:41 PM

Does anyone use the SGPro guiding recovery tools? Do they work?



#3 spokeshave

spokeshave

    Vanguard

  • *****
  • Posts: 2149
  • Joined: 08 Apr 2015

Posted 15 October 2019 - 05:55 PM

It would help to be able to see your logs. Something may have happened that prohibited recovery from working. Setting up recovery is fairly simple. Tools-->Sequence Generator Options-->Sequence Options

 

Then make sure that "Attempt to automatically recover the sequence" is checked and pick how often and for how long SGP should attempt recovery. Additionally, you want "Center when target starts" to be checked in your sequence settings if you want SGP to recenter when it attempts recovery.

 

When SGP tries to recover a sequence, it will first start PHD2 looping and ask PHD2 to auto-select a star. If a star is selected, one of two things will happen next depending on whether you have centering selected. If so, SGP will pause PHD2 and attempt a center. If the center is successful, it will again start PHD2 looping and tell it to select a star. If a star is selected, SGP will tell PHD2 to start guiding. If that is successful, the sequence is recovered and SGP resumes normal operation. If any of the above steps fail, SGP enters the wait period and will try again after the wait period is over (provided that the total recovery time has not elapsed). 

 

There are some things that SGP cannot recover from. If PHD2 loses the camera or mount connection, the sequence will not recover. If PHD2 crashes, SGP cannot recover. 

 

Recovery works very well for me. It is fairly common for clouds to move through during an imaging session and SGP handles that quite well for me.

 

Tim



#4 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 8944
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 15 October 2019 - 06:17 PM

Thank you Tim :)

Can testing of this be simulated? - its hard to conjure up clouds on demand and impossible to control the roof to run tests

 

Does the centering action need a plate solver? Could I avoid the centering part? as experience with my 10Micron based rig is even after 4 hours the image drift is only a small number of pixels at most (1" drift per 15 minutes typical - so around 16" max in 4 hours => ~10 pixels @1.55" per pixel).

I guide because I'm doing 60 minute exposures. If images are 10 minutes or under I normally don't use the guider



#5 Alex McConahay

Alex McConahay

    Cosmos

  • *****
  • Posts: 8245
  • Joined: 11 Aug 2008
  • Loc: Moreno Valley, CA

Posted 15 October 2019 - 07:42 PM

>>>>> Can testing of this be simulated? - its hard to conjure up clouds on demand and impossible to control the roof to run tests

 

Obviously, I do not know the specifics of your setup, but have you tried throwing a blanket over the front of your scope to simulate very dense cloud?

 

Alex



#6 don314

don314

    Lift Off

  • -----
  • Posts: 15
  • Joined: 10 Apr 2017

Posted 15 October 2019 - 08:12 PM

I have is set, and I have had it work.  I knew a bank of clouds would be passing while I slept, and in the morning I was missing data from about 2am to 3am, and in the images just before and after that window, you could see image quality degrading as the clouds approached/passed. 

 

To enable it just double click the recovery Icon at the bottom of the screen, select the "Sequence Generator Options" tab, and set the delay and duration.  I'm set at 5 minutes and 6 hours.  There is nothing to do in PHD2.

 

I have no problems with my X2.


  • Tonk likes this

#7 Tonk

Tonk

    Cosmos

  • *****
  • topic starter
  • Posts: 8944
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 16 October 2019 - 01:40 AM

Ok cheers. I'll have another go and take care with settings. 

 

The rig is 1500km away from me so blanket chucking is off the menu - but i do have a robotic lens cap and an OAG so i can simulate a blanket :) cheers for the idea Alex.



#8 spokeshave

spokeshave

    Vanguard

  • *****
  • Posts: 2149
  • Joined: 08 Apr 2015

Posted 16 October 2019 - 06:07 AM

Thank you Tim smile.gif

Can testing of this be simulated? - its hard to conjure up clouds on demand and impossible to control the roof to run tests

 

Does the centering action need a plate solver? Could I avoid the centering part? as experience with my 10Micron based rig is even after 4 hours the image drift is only a small number of pixels at most (1" drift per 15 minutes typical - so around 16" max in 4 hours => ~10 pixels @1.55" per pixel).

I guide because I'm doing 60 minute exposures. If images are 10 minutes or under I normally don't use the guider

You can simulate it by just stopping guiding in the middle of a frame. SGP should enter recovery mode. Since you have a robotic lens cover, though, that is clearly a more realistic simulation. Centering does require a plate solver and you can certainly avoid it by turning off centering in your target settings. However, centering on recovery is not so much to compensate for drift as it is to compensate for something going horribly wrong in guiding. As an example (one that I have seen happen), if PHD2 loses the guide star, it will keep trying to reacquire the same star in the search region. If there happens to be a warm or hot pixel in that same region, it can choose that pixel thinking it's a star. It will then try to bring that pixel back to center, which, of course, it can never do. It will eventually realize that the "star" isn't moving and give up, but in the mean time, it has move the mount quite a bit trying to center the hot pixel. The result being that the mount is now pointed somewhere else. When SGP recovers, it will happily resume taking pictures, oblivious to the fact that it is now pointed in the wrong direction. That is admittedly an unlikely event - though it has happened to me on more than one occasion. Careful settings in PHD2 and proper use of guide camera darks can make it even more unlikely. 

 

Tim


  • RossW likes this

#9 Alex McConahay

Alex McConahay

    Cosmos

  • *****
  • Posts: 8245
  • Joined: 11 Aug 2008
  • Loc: Moreno Valley, CA

Posted 16 October 2019 - 08:25 AM

>>>>>>  i do have a robotic lens cap and an OAG so i can simulate a blanket  cheers for the idea Alex

 

A robo-lens cap to simulate a blanket that is supposed to be simulating a cloud.

 

Wow, simulating a simulation of an error. 

 

Sounds like my daily encounter with reality. 

 

Alex



#10 nimitz69

nimitz69

    Surveyor 1

  • *****
  • Posts: 1545
  • Joined: 21 Apr 2017
  • Loc: A barrier island 18 miles south of Cocoa Beach

Posted 17 October 2019 - 01:51 PM

There is no reality..... only our flawed perception of it dalek12.gif dalek12.gif dalek12.gif




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics