I focus manually. Never implemented FocusMax although started to at one point.
But I am interested in hearing about autofocus software. I know FocusMax is used a lot although its an older product and the idea of having to slew to a suitable star run the routine and then plate solve back is not appealling. It just sounds like something that will go wrong at some point and waste a night's imaging. It would be better if it worked on the current field and not have to move the scope. After all that's what I do manually.
I have been using @focus3 with the sky x and have had good success with focusing. It focuses in place and almost always succeeds. It takes longer to focus with the ultra narrow nb filters but gets there once it extends the image time for each focus shot.
Once this rain and cloud goes I am aiming to get robofocus working on Voyager and it can be set to focus in place or slew to a star.
I'm in the SGP 3.x camp. Once you get steps and step size correct for your setup it's great. (There is a specific formula in the doco you need to follow).
I'm running at ~F3.9 with a Newt and even 5nm narrow band focusing works reliably.
There is also a great third party tool called "SGP Autofocus Viewer" that allows you to see exactly what is going on, both graphically as well as tabular. See attachments.
SGP 4.x now uses ASTAP to focus (already uses it for plate solving in 3.x) which is supposedly another step up.
The big question though is do you want to spend money heading this direction, or would Nina for free be a better area to dip you toe into?
Greg,
Imaging with newts eliminates the need for me to use a bahtinov mask
I just slew to a mag 2.0 to mag 2.5 star in the region of my target or its mid elevation path, take 20sec exposures in APT ( adjusting the histogram accordingly) and tweak my ZWO EAF electronic focuser until I get super tight sharp diffraction spikes on my Star.
Usually only requires 3 or 4 tweaks until I’m happy or around 5 to 10 minutes
So basically software assisted eye balling
Works for me every time unless atmospheric conditions deteriorate or are crap then focus usually drifts anyway whichever method you use
Martin
I understand that the voyager robostar focus software works very well.
It initially characterises your optical system by running V curves, and on subsequent focuses during an imaging run, it will use the characterisation and previous focus runs to converge at the best focus. So it tries to improve focus every time.
The auto-focus routine in the ASIair runs an initial v curve routine which takes a few mins but after that can be set to refocus every hour, every change of degree in temp or both as I use. The refocusing takes far less time than the initial run. I find it works perfectly but obviously you’re not just buying a way to focus by going down that avenue.
Voyager uses one of two internal focus methods (Or you can set it up to call Focus Max, Maxim DL or one of the TheSky versions)
The internal versions first do a "First light" run which characterises your optics by slewing to a suitable focus star and running through the focus point via many small steps from well out of focus, through focus and out the other side. Then it can either use it's "robostar" routine which slews to a suitable star (And plate solves to center it) and focuses on that, but it does not do the SGP-like step off focus and run an number of data points, it steps slightly off focus and given the mag of the star is known it knows what HFR to expect at what point in the focus curve it characterised before using a given exposure it adjusts exposure time and focuser position to fit to the expected HFR, then exposes a number of "integration" images of the star, then moves the focuser to what should be the equivalent point on the other side of focus and does the integration routine again. From that it can work out where each side was compared to best focus. The benefit as well as what I have found to be really good focus is that it focuses really quickly with NB filters, it slews to brighter stars for NB so the focus exposure time stays about the same, usually two or three seconds at most. It generally only takes a couple of minutes to slew to a focus star, focus, and slew back.
The first light wizard also populates the step size for the SGP-like full field focus setup (Which focuses in place) but I never use that. If your optics are reasonably flat I think you are better off with the robostar routine, particularly for NB.
Last edited by The_bluester; 04-07-2021 at 10:39 AM.
i've managed to get my autofocus working recently after some ascom issues. but i have been using NINA and have found it works well. I am liking it more than SG PRO esp the autofocus. it is easy to use and free so perhaps worth a try Greg.
i've managed to get my autofocus working recently after some ascom issues. but i have been using NINA and have found it works well. I am liking it more than SG PRO esp the autofocus. it is easy to use and free so perhaps worth a try Greg.
Another vote for "I used NINA and it works" - just make sure to set your AF to overshoot backlash compensation and it will make even the jankiest focusser nail focus.
My vote is for SGP Pro. SGP Pro averages the stars in the field that you are imaging so you don't have to find a particular star to focus on. Once set up with the correct parameters it works well. It can be configured to refocus based on a number of different criteria including time, temperature, filter change etc.
Ive used the AF routines in SGP, Voyager and NINA.
The results are pretty comparable, and they all work well, and sometimes all have issues from time to time.
I used to be a huge fan of Voyager's process, but had all sorts of issues when I started using the RASA 8 and Celestron focus motor. I found using the local field didnt work well on the RASA for some reason and sometimes got results that were clearly OOF. On my other rig (moonlite focuser, and Skywatcher 190MN) its been very good though, and much faster than NINA or SGP.
At the moment I have settled on NINA as my astronomy tool of choice, and so far its AF routines have worked well with the RASA8. It's simple, reliable and easy to understand. As long as you can dial in your step size, it works.
Focusmax Acquire Star works beautifully and does it all hands off - picks a nearby focus star, slew, focus, return slew and platesolve - but you need a full Pinpoint licence.
Voyager Robofocus does the same and seems to use a similar methodology to determine focus.
Actually, the Celestron motor any my C925 was tricky to get focus going on even in Voyager, once I did it generally worked flawlessly though. It needed the motor movement reversed and to make sure backlash compensation was going the right way or the mirror would still be moving as it took up the slack in the system. Voyager was the only program I found that would reliably focus that scope.
Greg, this is what the Voyager local field focus interface looks like when it is in action. It measures HFD for stars in the field you are currently imaging and fits a spline curve to find the best focus.
Last edited by peter_4059; 17-07-2021 at 06:04 PM.