Plate-solve-goto-center just perfect !

Image processing, astrometry, photometry, etc.
Post Reply
brujo
Posts: 6
Joined: 24 Jan 2019, 20:41

Plate-solve-goto-center just perfect !

Post by brujo » 24 Jan 2019, 21:03

Hello.
I have just finished my first goto and center with plate solve session on AstroArt 7 and I am admired how good it works.
My instrumentation:
Skywatcher EQ6-R mount,
Celestron 11 and TS RC6" syde-by-side on the mount
QHY174 Mono Cooled camera
Pegasus MotoFocuser
AltairAstro GP130M guide camera

All this (except guiding on PHD) connected to AstroArt.
I was able to center every galaxy, nebula I have selected all over the sky.
AstroArt is a wonderful application !

AstroBernd
Posts: 4
Joined: 17 Mar 2019, 16:23

Re: Plate-solve-goto-center just perfect !

Post by AstroBernd » 17 Mar 2019, 16:43

Hello Brujo,
Never dared to try to center this way. Will try it out next time.
Thanks for sharing!

Dunmunro
Posts: 58
Joined: 10 Jan 2019, 01:11

Re: Plate-solve-goto-center just perfect !

Post by Dunmunro » 17 Mar 2019, 19:28

AstroBernd wrote:
17 Mar 2019, 16:43
Hello Brujo,
Never dared to try to center this way. Will try it out next time.
Thanks for sharing!
I have also been using this feature successfully. I am now imaging remotely using Chrome Remote Desktop to control two observatory computers, each controlling a mount, camera, autoguider and focuser:

https://photos.app.goo.gl/btxzzxVWzjX9Qcb19

Plate solve assisted centring is absolutely critical for remote control and this feature gives me great capability and flexibility. I now no longer bother trying to use a multistar alignment to train the mount, rather I simply begin a session with the mount in the Zero or 'home' position (aimed north, CW shaft down) and after the mount initializes with the date and time, I locate the object on the Star Atlas and command the mount to goto my first target object. The mount will typically be ~1 or 2 degrees away from the target; I then take a ~10 second exposure and then use the telescope control-> Center->Goto plate solve, I click on 'center' and it plate solves the current image, determines the current location and the offset from the target location and then slews the mount to the target. Typically the Goto is accurate to within an arc minute or so. Syncing the mount to the target location seems to aid in accurate Gotos to the next target.

It is important to previously do some plate solves on your images using the 'find coordinates' icon and fine tune the plate solving engine. Use the stats command to ensure that the plate scale is correct,enter the horizontal field as accurately as possible and then reduce the % to 1 or 2 (this will speed up solves and prevent false solves) and then click on 'save parameters'. the telescope control-> Center->Goto plate solve will then use these parameters for solving.

einari
Posts: 17
Joined: 28 Jan 2019, 10:46

Re: Plate-solve-goto-center just perfect !

Post by einari » 20 Mar 2019, 20:41

I just tried like Dunmunro with my C8@1390mm and iOptron CEM60-EC mount.
Find coordinates worked but the rest - Center (plate solve) did not.

Dunmunro
Posts: 58
Joined: 10 Jan 2019, 01:11

Re: Plate-solve-goto-center just perfect !

Post by Dunmunro » 21 Mar 2019, 17:58

It should either work correctly, goto to a wrong location, or do nothing and report an error message.

If it goes to the wrong location check to make sure that you did not get a false solve on the image. I have found that occasionally the goto will be inaccurate for very short slews and then it might be necessary to repeat the process.

einari
Posts: 17
Joined: 28 Jan 2019, 10:46

Re: Plate-solve-goto-center just perfect !

Post by einari » 08 Apr 2019, 17:40

Back to this topic now that I have used AA's plate solving more.
Mixed feelings about it.
It doesn't always center object using plate solve (even if it has found coordinates).
And in my last session I first used Find coordinates, then plate solve center, not center, then again Find coordinates and now it refused to find coordinates.

Luckily manual centering is working better.

Dunmunro
Posts: 58
Joined: 10 Jan 2019, 01:11

Re: Plate-solve-goto-center just perfect !

Post by Dunmunro » 08 Apr 2019, 19:03

einari wrote:
08 Apr 2019, 17:40
Back to this topic now that I have used AA's plate solving more.
Mixed feelings about it.
It doesn't always center object using plate solve (even if it has found coordinates).
And in my last session I first used Find coordinates, then plate solve center, not center, then again Find coordinates and now it refused to find coordinates.

Luckily manual centering is working better.
Can you provide more details of your telescope+reducer and your imaging and guiding cameras?

If it refuses to find coordinates it might mean that the mount moved outside the target range for platesolving. What settings are you using in the "find coordinates" dialog window?

einari
Posts: 17
Joined: 28 Jan 2019, 10:46

Re: Plate-solve-goto-center just perfect !

Post by einari » 08 Apr 2019, 19:54

I was using C8 @ 1390mm, SXV-694 Pixel size: 4.54uM x 4.54uM
Field size 30.9 x 24.7 arcmin 0.674 arcsec/pixel.
Don't know what guiding camera has to do with this but it's a ASI120mm and 50mm finder.

Strange thing is like I wrote first Find coordinates worked then tried Plate solve center, it moved a bit but not much (maybe 10' max), but then when I tried Find coordinates again it almost immediately said could not find.
Used default settings, size 4x3, 5 ref stars, horizontal field 31, tolerance set to 50 or 60.

Dunmunro
Posts: 58
Joined: 10 Jan 2019, 01:11

Re: Plate-solve-goto-center just perfect !

Post by Dunmunro » 08 Apr 2019, 21:01

einari wrote:
08 Apr 2019, 19:54
I was using C8 @ 1390mm, SXV-694 Pixel size: 4.54uM x 4.54uM
Field size 30.9 x 24.7 arcmin 0.674 arcsec/pixel.
Don't know what guiding camera has to do with this but it's a ASI120mm and 50mm finder.

Strange thing is like I wrote first Find coordinates worked then tried Plate solve center, it moved a bit but not much (maybe 10' max), but then when I tried Find coordinates again it almost immediately said could not find.
Used default settings, size 4x3, 5 ref stars, horizontal field 31, tolerance set to 50 or 60.
It's possible to use the guidescope and camera for platesolving, which is why I asked about that.
Your focal reducer may have off axis distortion which can fool the plate solving engine.

It's probably best to set your field size to 30.9 and reduce the error tolerance to 2% along with limiting the rotation range. I find that false matches are possible with tolerance set too high, and there's no reason to have it above 2% once you know the correct image scale. I try to remember the orientation of the N-E orientation mark and if it changes then it indicates a false solve, which in turn will cause the mount to slew to the wrong target area and subsequent solves will probably fail. I also check the stats for platesolved images periodically, just to ensure that the measured pixel scale is correct, and if not then it indicates a false solve, but with the error tolerance at 2% I don't recall any false solves.

when you issue any movement command you must take another image and use that for plate solving. It is best to delete any prior images so that you don't inadvertently use the wrong image to solve on.

iOptron uses the mount's max slew speed for gotos, even over very small distances. I have written to them to allow for reduced slew speeds which should make short slews more accurate. I have found that in some cases I can sync the mount onto the current image coordinates and then issue the goto command within the iOptron commander interface, and this seems to work in some cases. However, I found that about 80% of the time, starting from the zero position, that two plate solves will place the object dead centre in the 3rd goto's frame with my CEM60, non EC mount.

fabdev
Posts: 101
Joined: 03 Dec 2018, 21:43

Re: Plate-solve-goto-center just perfect !

Post by fabdev » 08 Apr 2019, 21:35

A further detail: The "centering" via plate solve is limited to a 2x2° zone, so a maximum error of 1 degrees is corrected. This is was done to further reduce the risk of false positives and because the intended behaviour was just for "centering", e.g. when the object is already visible or just outside the field of view.
Inside Service Pack 1 there will be the option to increase the search zone to 3x3 and 4x4 degrees.

For larger errors, for example at the beginning of a session, the correct procedure is to plate solve then telescope sync (with the option: Last image).
Clear skies,
Fabio.

Post Reply