I will try GPG too but the weather here is going to be bad for at least a week.

When you try GPG, the parameters to think about are Estimate Period checkbox, and Major Period.
If you have no clue about the worm period of your mount the check Estimate Peiod (look here for clues: github-wiki-see.page/m/OpenPHDGuiding/ph...unt-Worm-Period-Info).
However, it's probably better to uncheck Estimate Period and enter the known worm period in "Major Period".

Also note, that GPG takes 2 full worm periods before it's fully in effect. So, if your worm period were 383s, like mine, then that's 13 minutes. (It'll be 50% in effect in 1 period). Every time you reset guiding (e.g. find a new guide star, that timer resets).

It is exactly the same thing as PHD2's PPEC (Predictive PEC) algorithm.

Hy

Hy, you mean that the GPG algorithm, the background is the same as PPEC ?

Cause if so, I believe I should have the same result as PHD2, but unfortunately, I don't - and truly I wish I could!

Already calibrate correctly, left it for more than 2 cycles, replicate all the values but the difference is huge.

Yes. I can't speak to why you have guiding issues, but GPG is exactly the same code as PPEC, or at least it was when I added it to Ekos a year or two ago. I got the open-source code from the person whose PhD work led to PHD2's PPEC, and it was the same code he added to PHD2. It works well for me, and I have gotten feedback from others that it has improved their RA guiding.

Hy

Great, good to know Hy!

I'm gonna revise all the parameters, cause literally my RA spikes like crazy while using GPG (+1,5) while with the same setup, night, settings etc in PHD2 with PPEC is well below 1.

First results after switching from traditional RA guiding to GPG RA guiding:

And after some more experiments, it turns out that a sloppy polar alignment (right side) produces a much nicer guiding than an almost perfect polar alignment (left side).

And finally the morning update after taking lights all night.

It is very noticeable that after the meridian flip the guiding results become increasingly worse in function of the elapsed time:

It's also possible that the sky conditions changed.
For sure, plot things like SNR or num-guiding-stars as well.
Look for all that in the Analyze tab.

Hy

FWIW, I just updated the Internal Guider section on the KStars Handbook.
You can find it online here:
docs.kde.org/trunk5/en/kstars/kstars/kstars.pdf#guide
Let me know if you see any issues.

Hy

Very impressive.

For me, on page 130. the link "this table" doesn't work. "The main settings to consider are Major Period and Estimate Period. If you know the worm period for your mount, perhaps by examining this table, then uncheck Estimate Period, and enter your known Major Period. If not, then check Estimate Period."

Thanks Jerry. I'll try and push this change invent.kde.org/education/kstars/-/merge_requests/964/diffs and it should find its way into the handbook in the next few days.