AutoFocus with OAG and Guiding

I am using a Lakeside motorfocuser on a WO Star 71 with a QSI 683-WSG. PHD2 is my guide software. I wonder if I am using SGP correctly to autofocus… It works, but it can take quite a while and has occasionally failed me. Currently I manually start an autofocus run before starting my sequence. That way I am in focus, start the sequence and the first target is centred, PHD2 starts guiding and off we go.

The problems occur when a focus run is required later during the sequence. I have the autoguide settings so that guiding continues during autofocus and have ‘Settle auto guider before auto focus frames’ selected. That way I stay on target and there aren’t large jumps in framing after a focus run, which I was getting when I stopped guiding during autofocus. However, it can take quite a while for the guiding to settle, especially at the more out of focus ends of the routine.

What I would like to be able to do when an autofocus run is required is to stop guiding, run the autofocus, re centre on my target, restart guiding and off we go with the next sub.

Is there any way to do that? Or is there a better way to use autofocus and stay well on target?

my PA could probably use a touchup since i do sometimes see large star displacements when PhD starts back up - i also pause guiding during focus, also because i have an OAG. and since so far i’ve insisted upon focusing thru my narrowband filters, an autofocus run can take 10 minutes sometimes.

however, almost every time PhD finds the guidestar again and recenters it. i’m thinking that if you increased the size of the search box in PhD that you might be able to recover from this situation?


I have a need for this also - but unfortunately I also need centering to be accurate in the first place or I can’t use it.

I’m not sure how many people are doing automated imaging - especially unattended - with OAG and long focal length. There is a fundamental problem that if an adjacent guidestar is pulled into the scene by slight movement - it could be grabbed and result in an offset.

This could be fixed by doing a plate solve after the autofocus and confirming it is still within X arc-seconds of the target - and if not trigger a re-center.

This is obviously added complexity - but I think it is needed for automated oag runs. And it does require accurate centering.


1 Like

fair enough, though i’ve had no problem at 1100mm, which might be considered halfway to long focal length.

The excess movement to take the guidestar outside of the green box in PHD2 and therefore PHD2 deciding to guide on a new star that I have seen seems to come from a slight total shift in the image due to the movements of the focuser. I haven’t even tried this with my Celestron EdgeHD 800 yet, which will no doubt throw up a whole new plethora of autofocus vs guiding issues!

I had some poorly registered subs after a recovery operation by SGP due to a cloud interruption. And, of course, PHD2 grabbed the brightest star in the field when it resumed, not my choice.
Kind of makes recovery useless, at least on tightly framed targets at long FLs. Worse was that I did not notice it happening even though I was “in attendance.”

You should probably look into why framing changes so much when guiding stops during focus. Autofocus works much better when PHD2 isn’t fighting against it. I always have guiding turned off during autofocus. I do get a slight amount of image shift, but I set the search region in PHD2 to be the largest allowed and the guide star always stays inside the box during AF, and when guiding resumes, PHD2 will bring the guide star back to its original position.


In my case I am using EdgeHD11 at f/10 with 8300 ccd and 0.4" per pixel. The guide camera is a qhy5iii-174, which is fairly large and sensitive, and I am guiding with MetaGuide. For optimal framing I may need to put the guidestar somewhat close to the edge - and it may move off during autofocus or pull another one in. If it happens to be well centered and there are no other guidestars around then it is less of an issue.

I have let it run, doing automatic autofocus periodically while I sleep, with exposures up to 30 minutes - and it has worked well. It is doing all this stuff about unlocking the guidestar and stopping guiding - then running autofocus - then re-locking the guidestar and starting guiding - and it works well.

But sometimes another star does creep in and things go wrong.

So although this is a fairly challenging imaging situation - the equipment isn’t the limiting factor since it is mechanically capable of centering on the arc-second scale - and the autofocus gives very repeatable curves. The main thing I need is accurate software centering so the guidestar lands where it should be and the framing doesn’t change during an imaging session. If I had that and a check on the framing after autofocus - it should all work much more reliably.


1 Like

For what it’s worth, I am imaging with an Edge 9.25 with an OAG at f10 and 2350mm. Pixel scale is .57" per pixel, and my total RMS error averages about the same - somewhere around .5 to .6 arc seconds. I do not guide during auto focus, because it is obvious to me that moving the star around with PhD while trying to measure HFR for auto focus is likely to cause inaccurate readings, both for PhD and for the auto focus. So you get neither good focus, nor good guiding. I have never had any problem with PhD not getting right back on the same guide star after the auto focus run. If polar alignment is good, there just should not be that much drift of the guide star in the few minutes the auto focus run takes.


Dean, I completely agree with your thoughts and I would definitely like to let autofocus happen without guiding running. The problem I have encountered is that the whole image can move significantly when moving the focuser. So, if I run autofocus without guiding, PA is good and the telescope is pointing in near enough the same place, but the optics have made the star shift too much for PHD2 to cope when it restarts…


Do you know what is causing the image to shift? I’m surprised that focusing a refractor would shift the image. My Edge 9.25 does shift the image slightly while focusing due to the moving mirror, but it returns close enough that I never have a problem getting back on the same guide star. My refractor has no shift during focusing though.


Hi Dean,
Sorry for taking an age to reply. Thanks to your response I started to investigate the scope and the focuser and discovered that the focuser was a bit wobbly. Not what it’s meant to be like. A bit of research and I have tightened up the correct screws and it is now solid. Hopefully that will stop the star movement during focusing and it will be possible to pause guiding during autofocus without losing the guide star. I look forward to trying it out.
Thank you for your help,

1 Like

I never had this problem when I was imaging with a C8 Edge at 2000mm FL last year… the problem was to find a guide star with the framing I desired… imaging remotely without a motorized rotator, this is a real challenge…

Well that’s good and bad news! I found an issue with the focuser on the scope that was giving me problems - it was loose and wobbling a bit, enough to mess things up. All fixed and working properly now. I haven’t tried autofocus with my EdgeHD 800 yet, but it’s good to hear that you had it working well. Fingers crossed that it all behaves for me. As for finding guide stars… My Lodestar X2 has been amazing so far. What guidecam do you use?

Glad you were able to fix the focuser.

I was using a QHY5L-II. Just purchased a QHY174III. This new camera sees 3-4 times more area and it is much more sensitive so I’m quite hopeful that finding a guide star will be much easier.