Never happens when testing the gear on less than perfect evenings of course…!: funny once in a while behavior, sometimes the auto center function is stuck in the same error, say (55 pixels or any other value) and keeps getting the same exact values after every centering cycle (instead of converging into a lower error)…I’d very much appreciate if anyone knows a potential reason for this.
90% of the time auto center works perfectly and gets it within a few pixels of the reference image.
I have a 25 pixels max centering error threshold, I use pinpoint solver.
Last night auto center just “didn’t want to converge” …no changes from prior nights other than covering and uncovering the scope to continue an imaging run on the Rosette.
thanks!
Without a log, there is a lot of guessing. Anyway:
There is an issue with the Centering by Sync when the mount has a pointing model. This applies to different mounts and mount software. The solution is not to use a pointing model or to wait for the soon to come Centering by Correction method.
What Pixel scale are you using? A 25 pixel centering error might be a very low value when expressed in arcsec, lower than what your mount is able to do.
I guess I’ll be forever newbie, somehow looks like these “troubled times” I force- loaded the pointing model with EQMOD…I just tried tonight making sure there is no pointing model at all and it slewed and centered like a charm within a few pixels in the first try…THANKS!!
For the record, I think the experiences vary by user. I had trouble myself when no model was used (EQMOD, EQ8, PlateSolve2). Seemed to work best for me when I left the model intact (no more than 7-8 points) BUT did not allow syncs from the plate solving to add to the model (in EQMOD Alignment/Sync / User Interface = “Dialogue Based”). This way, each subsequent platesolve would adjust the sync (hopefully get closer to desired pixel error) without modifying the model itself.
Based on past postings (in SGP Forum and CN) there have been others with similar experience. Not sure why users with similar setups would experience different results. So stick with whatever works.
I have been using 50 pixel tolerance. Not quite brave enough to try 25 :=)
Dave, last night I had more than 7-8 reference points in EDMOD, it was like 27, so what you say makes. sense…it may work or not depending on the # of references. For the record, when I didn’t use the reference model tonight, the scope had been already sync and pointed from prior nights and unmoved from its location…
