Feature Request: Define region within which auto-select will attempt to find star

graemecoates commented 4 years ago

Couldn't see a request for a similar feature - would it be possible to have a feature to define an area within which the auto-select routines attempt to find a star?

I'm thinking here of a couple of use cases. One might be if you have a fairly short focal length guidescope, where it's possible to choose a guide star that's someway outside of the image area. Another similar use case might occur if you have less than perfect alignment between the guidescope and the imaging scope.

In both cases, selection of a star in the wrong area of the autoguiding frame can result in an increased chance of field rotation - if I was able to define an area (eg x,y bounds) within which auto-select guide star would work (or at least try to find one), then this might be avoided for unattended guiding.

Thanks!

agalasso commented 4 years ago

I'm not sure if this is useful to you or not, but region of interest for star auto-selection is already available through the server API. https://github.com/OpenPHDGuiding/phd2/wiki/EventMonitoring#available-methods note the roi optional parameter to the guide method

graemecoates commented 4 years ago

Ah - that's interesting to know. I'll have to check out if it's possible to pass that from SGP over to PHD then. Thanks for the pointer! (would be good to have it in the GUI as well though).

bwdev01 commented 4 years ago

Hi Graeme. Has this been a problem you've actually encountered? Could you maybe give us some numerical examples where it was a practical problem? I'm asking only because I can't recall ever hearing of this sort of problem being reported on the forum.

Thanks, Bruce

From: graemecoates [mailto:notifications@github.com] Sent: Monday, February 03, 2020 3:12 AM To: OpenPHDGuiding/phd2 Cc: Subscribed Subject: Re: [OpenPHDGuiding/phd2] Feature Request: Define region within which auto-select will attempt to find star (#835)

Ah - that's interesting to know. I'll have to check out if it's possible to pass that from SGP over to PHD then. Thanks for the pointer! (would be good to have it in the GUI as well though).

You are receiving this because you are subscribed to this thread. Reply to this email directly, view https://github.com/OpenPHDGuiding/phd2/issues/835?email_source=notification s&email_token=ADDHSVYZYPZT6FYNUQVY6JDRA73ZXA5CNFSM4KOWUSDKYY3PNVWWK3TUL52HS4 DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKTN5YA#issuecomment-581361376 it on GitHub, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADDHSVZ2I52HX4E2KKQO6MTRA 73ZXANCNFSM4KOWUSDA . https://github.com/notifications/beacon/ADDHSV4TB5E46CSQL3TEXYLRA73ZXA5CNFS M4KOWUSDKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKTN5YA .gif

graemecoates commented 4 years ago

The example I can give is as below. [I'm fully aware that, the "correct" solution is to get better alignment, but I hope this is a reasonable example - this is fairly close to what I think I'm seeing. (another use case might be from an off axis guider that has poor enough star shapes off axis that we should restrict PHD to searching in only one area I suppose)]

Example:

I have a dual rig, one with a moderate length refractor (say 800mm), and the other with a widefield telephoto lens based system. FoV are 1x0.7deg and 3x2deg respectively. I can alternate between using either the refractor or the telephoto system as the guider/imager.
Assume polar misalignment of say 4 arcmin (OK - this could probably be improved, but for a portable rig, this might be somewhere near expected).
Using a the telephoto as the guider for the refractor, if PHD2 picks a guide star off near a corner, then I'll get a much worse result in terms of field rotation than if it picks a star near the centre of the field.

In theory, it could pick a guide start that is somewhere around 2.5 degrees away from the far side of the field being imaged, assuming perfect field alignment - if there's a bit more misalignment between the two systems, the result could be worse.

With these in mind, using narrowband exposures of 20min (fairly reasonable for ensuring read noise is swamped), then the calculated field rotation distances by declination at higher decs are as follows (see: http://celestialwonders.com/tools/rotationMaxErrorCalc.html):

Dec Rotation distance (microns) 40 4.6 50 5.5 60 7.1 70 10.4 80 20.5

If I could pin down the guide star to within the field of the refractor, then the rotation distance for furthest regions of the imaging field and the guider halves in this case above.

Ultimately, the answer is to get better polar alignment as rotation goes fairly linearly with polar misalignment. And of course, the effect is worse at higher declination. But, I am reasonably sure I have seen this take place for a system I use for some objects in Cepheus - probably exacerbated due to a slight misalignment between guider and imager.

Just wonder if it would be helpful, given we have the option via the server API, but not in the GUI.

Cheers

Graeme

OpenPHDGuiding / phd2

Feature Request: Define region within which auto-select will attempt to find star #835

Ah - that's interesting to know. I'll have to check out if it's possible to pass that from SGP over to PHD then. Thanks for the pointer! (would be good to have it in the GUI as well though).