I reported this issue on the PHD forum and Bruce has acknowledged it is a bug in this initial release of mutli-star guiding and a fix is on the way. On my rig I had the settling limits set to 0.7px and 7 seconds and I observed in the logs that when the bug occurred the reported settle distance was bouncing around between 1 and 1.5 pixels. So as a temporary workaround I upped my settle limits to 1.5px and 15 seconds and tested that last night and it seemed to avoid the settling issue. I’m confident that my mount is settling to less than 0.7px within 15 seconds so I’m betting a few extra seconds settling time on every image to avoid the risk of a 15 minute time out when the bug kicks in.
2 Likes
Following earlier advice I downloaded the multi star version of PHD. It all works fine, but with and intial run af 4 hours I didn’t see much in the way of improvement. My rig usually runs with a best Total RMS of slightly under 0.5 arc sec with the multi start version of PHD the best I am getting is around this figure.
I didn’t run a new calibration nor use Guiding assistant to see if any settings required adjustment. Would either of these make a difference?
That’s a good question and one I wondered about as well — is recalibration necessary?
Not sure if it is recommended by the PHD2 folks, but I certainly did not need to do one (recalibration). And unexpectedly, noticed an immediate improvement. This was with a 6" APO 1050mm focal length on an MX+ w/ 60mm guide scope (225mm focal length).
Results from others seemed to be mixed, but many do seem to notice a significant difference. I am still curious how/why the multistar could make such a difference.
DaveNL
I think it does two things:
- Better centroid calculation by averaging across multiple stars. Useful if (like me) your guide stars are not perfectly round due to being at the very edge of my reducer/flattner
- Better averaging of seeing within the guide scope / OAG’s view, again by averaging across multiple stars.
I’ve personally seen a clear improvement in the guiding numbers and some small improvement in final star fwhms.
Bug fix has been implemented and beta testers are now trialling it prior to Bruce releasing an update. Best to just loosen your settling criteria slightly in the interim.
1 Like