When automatically running sequences, is it possible (or have an option) to run autofocus before starting up the autoguider?
using an OAG which requires reasonably good focus to acquire star for autoguiding.
however, i often have a couple hours waiting until the target starts, and by then it’s pretty far out of focus.
The auto sequence now is start up autoguider followed by autofocus. I’m wondering if that order can easily be switched, or have an option to autofocus prior to autoguider starting up
If we get a good focus before PHD2 auto-selects its guide star, we are less likely to end up with a saturated star and more likely to end up with the “best” guide star. This applies to changing targets too: any time a new guide star needs to be auto-selects and focus needs to be done, please do the focus first!
It amazes me how well PS2 is sometimes able to work with de-focused stars at all. But other times it just cannot cope. De-focused stars is often a problem for me if it has been a long time (ie. weeks) since last session - an all too common occurrence here. Probably because seasonal changes means diff temp etc.
If running unattended (as I often try to do) this stops the sequence before it really gets going.
I am currently running a sequence that waits 2 hours for the target to be up (and for it to get dark). By the time it does so, the temperature has dropped 15 degrees and the scope is not is good focus. PHD2 picks the best star, then autofocus runs and that star ends up being very saturated. That star would not have been picked if it was already in focus. My normal 0.5 arc-second guiding turns into almost 1 arc-second.
I’ll add my voice to this request as well. I’d already mentioned the need to “slew - focus - center” a target rather than the current “center - focus” approach which causes plate solving problems.
I should probably start another thread, but what I’ve been thinking would be very helpful (and hopefully not too major of a task to accomplish) would be to have the ability to add ANY event to the sequencer, just like an exposure. There are certain events tied to sequencer items already, so some of that could be accessed just with a “null” event that doesn’t actually take any images. Anyway, just a thought. I realize there are challenges to using something like this in addition to any coding challenges.
I requested the same several years ago. In the meantime, I have a workaround. I put in a dummy target leading up to the real target start time. This does not have centering enabled and just takes heavily binned images every 10 minutes or so. The AF triggers are common for the sequence and it has the effect of nailing the focus right up to the real target. It fixes another issue for me too, with my RCT, I get more accurate focus if the focus point is within a click either side. In this way after a few AF runs, it has it zeroed in.
I allowed this into 3.1 because it is not really a feature request. The auto focus trigger that forces a run at sequence start is supposed to run before centering or guiding, but it was broken. The only gear related activity that should run first is slaving the telescope to the observatory so auto focus is not trying to focus on a wall.
This has been addressed in any version of SGPro > 3.1.0.283
Thank you Ken.I think i know what happened in my case. If you don’t let SP pro open PHD2 the pause command does not go out or is not received by phd2. Tested several times and the only solution is to let sg pro connect to all equipment. If i load phd2 first to calibrate or do other maintenance it won’t pause before focus. Sorry for my lack of fitness in explaining this.
i’m not sure i understand how it’s supposed to work?
What I always envisioned is that it slews first, then autofocuses, then plate solves. This avoids the issue where it slews->failed plate solve due to lack of focus
right now I have two problems on 327
it autofocuses before slewing, so if I’m in the trees it will fail. prefer to slew first
if I start a sequence early (while it’s light) so it runs remotely, it tries to autofocus immediately.