Hello Paul,
The MESU200 is an excellent mount. I have no experiences with it but reading the docs it has servo motors with integral encoders. So it will have a kind of control loop where it moves to the required position. Maybe it has some initial overshoot or undershoot or delay and this maybe could cause some problems. Just an idea which only can be excluded by testing it with increased settle time.
Han
I’ve changed my settle time from 1 to 4 seconds Han. Not sure if I’ll get ban opportunity tonight.
Best
Paul
@han,
What qualifies astrometry.net as the reference? I’m am not sure that it should be the reference.
How many images did you compare?
What image scales did you use?
Thanks,
-Ray
Ray,
I did several comparisons in tha past years. The table earlier in this post was based on some images from a 10 micron user. There is no prove the Astronomy.net gives the most accurate position. It all relative. There is not absolute reference. You can see that ASTAP using the U16 (UCAC4) database based ASTAP and Astrometry.net are getting closer. This could be either by the stars selected or the catalog itself. The current database is based on Gaia which beats all other star catalogs with a factor 1000 or more in accuracy. You can also see that the local Astrometry.net and online version differ.
But all errors are super small less then one pixel. I assume from optical distortions. A solver works or it fails miserable. There are no small errors.
Han
Han,
I’m taking a different approach to testing accuracy. I’ll let you know what I find out.
-Ray
One thing I did not see mentioned is the pixel scale of the image to be solved. You have an accuracy limit of +/- one pixel with any plate solver. If your camera / OTA produce 2" pixels and you bin, your best possible accuracy is 4 arc seconds. If you tell SGP that centering is acceptable at 20 pixels, then you are centering somewhere inside an 80 arc second circle.
Charlie
If you share your approach Ray, we may be able to help out too.
best wishes
Paul
Charlie,
About star position accuracy.
It is the same as for guiding, you can measure a star position on sub pixel accuracy. The star HFD, half flux diameter diameter is maybe 2 pixels meaning the star flux is spread over something like 16 pixels (4x4). By measuring the flux in all illuminated pixels you can calculate a point called center-of-gravity. Then if you combine the X, Y position of several hundreds of stars the accuracy increases even more.
If you would have an under-sampled image where a star illuminates only one pixel you would have less accuracy. So the spread out of the star flux over several pixels helps with measuring accurate.
In ASTAP there is option demonstrating and proof the accuracy. It can plot circles (star database annotation) based on the solution . They will fit nicely around the imaged star. It can also be used to measure optical distortions indicated with arrows . Here a link to an example image with severe distortion:
https://www.astroforum.nl/filedata/fetch?id=1403775&d=1563702107
Back to SGP centering. I hope we can trace the cause soon.
Han
Good point about averaging over multiple pixels to get a statistical pixel that represents the center of the star image.
Bottom line, I have been using ASTAP plate solving for the better part of a year and have had virtually no plate solve failures in hundreds of runs. My system has a pixel scale of about 1" when binned 2x2, for reference.
Charlie