Too many focus runs

Description

I have run two sequences tonight using different rigs. At the start of teh sequence, SGP slews to the target. It then focusses. It then does a Centering routine (using ASTAP). Then it focusses again. Either the first or teh second focus seems superfluous. In fact I had already focussed and so I cancelled both runs on the sequence now running.

Do I have something set incorrectly, or is this a bug?

Ta

Link to Logs

Useful Info

OS: Microsoft Windows 10 Home
Ver: 3.1.0.435
.NET: 4.8

It is possible that in your equipment profile you have under the ā€œUse Auto Focusā€ set icon a checkmark "auto focus after automatic centering action. So SGP slew to target, will auto focus, then centers, and due to the checkmark refocuses after centering. It is worth a look to see if that is the case.

I do indeed have that checked. I have done that because I want to refocus after a meridian flip. I;ve had it set that way for quite some time. But I donā€™t ever recall seeing the double focus run before. I have 4 rigs so I have watched SGP starting up quite frequently.

1 Like

I think that during all the recent updates, the developers changed the auto focus to now focus prior to centering. I too had to uncheck the ā€œfocus after centeringā€ for the same redundancy.

You can check the version update logs on the website. Hopefully this helped.

Understood. That would explain the ā€˜newā€™ behaviour. And it would explain why we donā€™t see it in Spain where we have not updated given all of the issues with the new versions.

Iā€™m on the last but one update at home (I think - itā€™s difficult to keep up). But my question still stands: Iā€™d like to trigger autofocus after a meridian flip (just in case anything has moved/slipped). How do I do that without double focussing before sequence start?

This happened to me last night for the first time as well. Is the solution to uncheck the box that auto-focuses on sequence start, i.e., will it still autofocus at the commencement of the sequence due to the centering associated with the plate solve? But, doesnā€™t that eliminate the autofocus prior to the plate solve? And, if you uncheck the box to autofocus after centering, will there be no autofocus after a meridian flip?

Yes, I noticed this last night too. The second autofocus seems superfluous.

Yes, this is intentional (at the moment)ā€¦ it was pointed out to us why folks with big heavy mirrors need it. We will probably modify the option ā€œAuto focuser after centeringā€ with a little extra checkbox for ā€œunless focus was triggered at sequence startā€.

For now, most should be fine not requesting AF at sequence startā€¦ it is only really required to get you to a point where you can plate solve properlyā€¦ other than that, it doesnā€™t have much use (when paired with the other trigger).

The whole country? Your power is vastā€¦ :wink:

Ha ha. Very good, Ken. We have two rigs in Spain. Because we operate them remotely, we will not use Betas since we donā€™t want to run into any problems. Unfortunately, the latest ā€˜stableā€™ release does not seem to be all that reliable, as evidenced by the need for very frequent new releases of the ā€˜stableā€™ version. So my imaging buddy and I have not upgraded these rigs. I am not able to follow all the threads on a daily basis, so I was quite mystified when the second focus routine began (and you will no doubt have seen the other thread). I donā€™t know when you introduced this new way of doing things, but how well was the change of behaviour publicised? It appears to have discombobulated a few of us - and it is very frustrating to be having unexpected behaviour at sequence start up - especially when so many of us get so few decent inaging nights.

Itā€™s a great pity because the stable version of SGP once seemed to be quite reliable. I appreciate that the mirror guys have specific needs and that you are keen to address their requirements. But a lot of us old guys still use refractors. I used to be a great advocate of SGP. However, as it is, Iā€™m not entirely sure that I would recommend SGP to newcomers to the hobby as things now stand. It seems to have been a little bit too flaky, for too long now.

1 Like

Wellā€¦ I fear your expectations of SGPro and us are simply too high. We do this in our spare time (as I sit here on Sunday and do support)ā€¦ it doesnā€™t pay the bills and we donā€™t have time to do all the things you are suggesting. I have explained this from time to time in the past (e.g.):

Feel free to to use whatever version of SGPro you are comfortable with or even to delay using a release for 6 months after release. The version you refer to as stable also went through a stabilization periodā€¦ This stabilization cannot occur in beta because the number of permutations of equipment and sequences it is exposed to is just too small.

2 Likes

Didnā€™t intend to upset you, Ken. I am well aware that your have limited resources. In my view, youā€™ve tried to introduce too many innovations too quickly in an attempt to satisy demands from users. Perhaps users need to have their expectations about the possible rate of change lowered given that you are a modest operation.

I noticed this double focus myself last night.

I think the amount of changes is great, small changes are done, released and tested and if any problems are found they are fixed very fast. I do however understand what gnomus means and i wonder if there could be some ā€œlong time tested and verifiedā€ version that people worrying about bugs could use?

I control 2 remote setups myself, 1 in Spain and 1 here in Norway and i usually just upgrade to the latest beta version at least every 2-3 days, i havenā€™t really had much big problems except during the time when SGP kept resetting the target settings last fall.

I would like to thank Ken for the great work he has done the last few months in really making SGP great, it runs better than ever, especially the changes done to autofocus with the help of the community is a great addition, before these changes autofocus would often get lost and ruin the rest of the night.

The release is generally stable. I would argue that the double AF is not going to destroy your night. It may be annoying and it will get addressed but itā€™s unlikely to be a reason to revert back to a previous version.

The best approach to making the release stable would be more involvement in the betas. The more feedback we can get on the beta the better it will be when we hit release. We hit the release button once we are no longer finding issues with the beta and issue reports stop coming in. At that point we release and substantially more users begin using it and, not surprisingly, finding issues.

Thank you,
Jared

This is version 3.0. It is still available for download. Or do you mean something else? We donā€™t have time to make adjustments to it, but we do still support it in terms of help.

I donā€™t think Iā€™ve claimed that my night was destroyed. In fact, I saidā€¦

The issues Iā€™ve had - be it the ā€˜slow plate solve timesā€™, the ā€˜deletion of targetsā€™, or the ā€˜autofocus-reset-if-one-starts-too-far-outside-of-good-focusā€™ - are, as you say, irritations. But then I came to SGP because it had less irritations than the then alternatives.

I agree that most issues tend to get sorted, and I appreciate the effort that this takes on your part. But there is an impression that you seem to have bitten off a bit more than you can chew with this latest release. Responses such as, ā€œWellā€¦ I fear your expectations of SGPro and us are simply too high.ā€ tend to reinforce this perception.

I am not sure that I have suggested any new features - if anything Iā€™ve tended to argue for status quo until such time as an ā€˜improvementā€™ is a genuine and stable improvement. The program is a commercial piece of software: its profitability is not the issue.

I have a suggestionā€¦

Could the software easily produce a simple xml file or flow chart, showing each sequence step once the seqence is ready to run?

That way, the sequece could be looked at as a, well - sequence, a bit like an if/then flow chart? People could then scan this and see where events are taking place in an easy to understand way, and go back and edit within the main GUI if they want to remove a certain action, in this case the double focusing when not needed.

I appreciate this may not be easy but if it is, perhaps it coud be looked at?

Or, maybe this already exisits and I havenā€™t found it yet?

This is no different than any other stability period for any other minor point release that weā€™ve done. Whenever weā€™ve transitioned from beta to release there is always a 1-2 month period where things are getting found and fixed rather rapidly. Maybe it seems more chaotic now as weā€™ve changed our build process to be nearly hands off so deploying a new release is trivial for usā€¦so we do it more, lots more. Where in the past we would batch up fixes into a new release every week or two.

I wish we had the resources to have a QA team but sadly that isnā€™t the case. We take a fair amount of effort to make sure that there are no major issues prior to a release but escapes certainly happen.

Weā€™ve worked on something like this and it always gets difficult to go much beyond the first couple of events as we canā€™t predict whatā€™s going to happen with focus runs, guiding, etc that will eat into time predictions. So you basically end up with something that is almost certainly incorrect. I think if we were going to do something like this it would need to be more real time and adjust as the conditions changed.

Thanks,
Jared

Ok, then consider this a feature request!

I also have noticed this double autofocus and it is indeed annoying when slewing to multiple targets.

Again, the only change here is that on sequence start, auto focus will trigger before centering and then again after centering. It does this because you have asked SGPro to focus on sequence start AND after a centering action. The ā€œdouble focusā€ will not occur on subsequent targets. You can eliminate this behavior by unchecking the ā€œFocus at sequence startā€ option.

1 Like