|
The pdf that has the Tpoint tutorial and example goes over the number of points you need before you can introduce extra terms.
From memory I think that was about 50. There is a maximum number of terms suggested you can add with the lower number of map points.
You also need to fit the data and remove outlier to make the model more accurate.
Only then it says is the PA model accurate.
So the idea you can do 6-20 may work but it isn't what is in this tutorial.
In my experience with my setup (limited experience) I manually did models of up to 30 points. I found the PA suggestions sometimes made the guiding worse so I therefore assume further away from accurate PA.
I did 200 point model (it only took about an hour) and used those suggestions. They improved things. Mind you it is easy to adjust the elevation the wrong way so if there is any doubt about which way to turn that check the manual before you do it.
I have also done a few 50 - 100 point models and after fitting the data, getting rid of outliers, adding terms the PA adjustments seemed to improve things. I do an autoguiding straight after an adjustment as final proof I did improve things and not worsen them.
If you aren't doing large enough models, and removing outliers then adding terms then the PA adjustments may not be correct as they will change when you do this.
I use 3x3 binning when using my Proline 16803 or 2x2 binning with the ML8300, 10 second exposures, slew time seems to be about 5-6 seconds, download time is 1 second so the whole thing from slewing to the point, taking the exposure, plate solving it takes 19 seconds on my setup. I could probably reduce that to 15 seconds by using 6 second exposures if it were a moonless clear night.
So try the above and see how it goes.
It does seem a bit elusive to get it exactly perfect. My setup is very close (the PA adjustments said zero in one axes) yet I sometimes get slightly eggy stars still. I am wondering if I am getting flexure in my guide camera in the MMOAG.
Greg.
|