Tried the new beta last night and it had a few autofocus issues while I was watching it early in the evening.
Finally, it seemed to be OK so I let it run unattended but then then after a flip it could not focus and eventually crashed with what looks like an out of memory issue.
I have gone back to the most recent full release as of this morning as I have objects to complete before our clouds return for the winter.
Log at:
There are all sorts of problems going on here…
First your camera starts returning what appear to be completely black
images… SGPro tries again, but gets the same result 3 times in a row anf
gives up:
[10/10/2015 1:28:06 AM] [DEBUG] [Camera Thread] Find Stars: 0 found…
[10/10/2015 1:28:06 AM] [DEBUG] [Camera Thread] Auto focus HFR calculated
at: 0
[10/10/2015 1:28:07 AM] [DEBUG] [Camera Thread] HFR value from auto focuser
is 0!, trying again…
Then near the end of the sequence you mount is failing repeatedly with out
of memory errors.
Well, I have no clue then. I just checked the camera with some various level flats and the earlier stable SGP release and it is responding normally.
The mount connects and slews with no issues at all. It responds to commands from both SGP and TheSky.
The logfile (see below) seems normal to me given that I was not running a sequence. There does seem to be some can’t get pierside issue, not sure why as the driver is set to provide this and it does seem to work as far as flipping, even with the current beta.
To be fair, you also indicated that the current beta responded the same way during the beginning of a sequence.
Quite true. I was just being sure there were no hardware problems like a stuck shutter or mount issues. These seem to be behaving normally. It just seems odd that both the camera and mount, two largely unrelated items, would have issues at the same time…
Just as an aside, the can get pierside error that shows up also shows up in older logs where everything ran smoothly so that is probably not an issue. Also, all the images to that point show normal star fields and my all-sky camera and cloud sensor both confirm the skies did not cloud up anywhere near that time (just some high clouds before dawn, several hours after the crash).
Here is where it seemed to go off the rails. In fact I have seen recent versions do something similar while I was watching. It seems to think that focus is going bad and in trying to fix that it makes it even worse, getting way out of position. I suspect that it got so far off it was seeing no stars at all and that caused it to give up.
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] → Find stars, try 1
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Generating new star list…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Find Stars: 11 found…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] → Find stars, try 2
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Generating new star list…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Find Stars: 11 found…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] → Find stars, try 3
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Generating new star list…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Find Stars: 11 found…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] → Find stars, try 4
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Generating new star list…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Find Stars: 11 found…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] → Find stars, try 5
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Generating new star list…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Find Stars: 11 found…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Star list contains 11 stars…
[10/10/2015 2:14:56 AM] [DEBUG] [Main Thread] Calculating HFR with sample size: 100
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] → Find stars, try 1
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Generating new star list…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Find Stars: 11 found…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] → Find stars, try 2
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Generating new star list…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Find Stars: 11 found…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] → Find stars, try 3
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Generating new star list…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Find Stars: 11 found…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] → Find stars, try 4
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Generating new star list…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Find Stars: 11 found…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] → Find stars, try 5
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Generating new star list…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Find Stars: 11 found…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Star list contains 11 stars…
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Calculating HFR with sample size: 100
[10/10/2015 2:14:57 AM] [DEBUG] [Camera Thread] Auto focus HFR calculated at: 363
[10/10/2015 2:14:58 AM] [DEBUG] [Camera Thread] Detected focus getting worse! Shifting auto focus range
[10/10/2015 2:14:58 AM] [DEBUG] [Camera Thread] Focuser moving to 20786
[10/10/2015 2:14:58 AM] [DEBUG] [Camera Thread] Focuser move call complete
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] MoveFocuserAbsBlocking: Focuser position 20786 matches requested position 20786
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] Handling monitoring event (Good Night System, Warning).
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] GNS: Sent warning message to GNS ((SH2-132 and Fish RGB Plus 10-5-15) Cannot find focus point, assuming focus is outward, adjusting smart focus range…)…
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] Auto focus data
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] - Data Points: 11
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] - Step Size: 102
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] - Current Position: 20786
[10/10/2015 2:14:59 AM] [DEBUG] [Camera Thread] - Initial Move Position: 21296
I also would ask what part of the SW causes the focus to assume things have gone bad and start looking for a new position? How new is that feature and can I turn it off? My best guess is that that is what caused this issue.
Since I have a very well characterized system and a high end focuser, I can deal with focus shifts w/o those features and would rather disable them it they might cause problems.
I guess that I have to assume that this is due to some change in the SW since nothing has changed in my system for well over a year and as far as I know nothing is broken as it responds exactly as asked to when operated manually either thru SGP or the focuser SW.
Finally, the only thing that did happen about that time was that the seeing started to deteriorate a bit:
Even so, it did not get that bad until well after the crash (I have used SGP in the past with far worse seeing), but maybe that could have caused SGP go go into it’s “can’t find focus” routine.
This feature was introduced about 1 year ago in 2.3. It is called smart focus and can be turned off in the AF settings. Other than that AF has not changed in very long time (because we are doing that in 2.5).
Yeah, that is probably not it as I have had good luck with versions much later than 2.3 and smart focus has pretty much always been engaged. Very odd that I have only seen this recently (last couple of betas).
Also interesting is that I did not catch this when it happened because I had stopped using GNS due to it’s issuing a lot of false alarms when guiding took longer than normal to settle and it had gotten quite annoying. I need to ask about this on the PhD forum…
That leads me to ask if an built-in web monitor is coming in 2.5?
You quoted something I deleted. I deleted it because I can now see clearly where it ran and after it ran it started producing images with 0 stars and 0 HFR. This is almost certainly what caused the issue. Is your imaging region pretty star-poor?
No plans for this at the moment.
Not especially. It is the fish head (IC 1795) part of the heart nebula.
OK. Well, I would turn the option off and increase your exposure time by 50% (9s?) maybe and see how that works
OK. Just for completeness, here is the “GNS issue” (not really a GNS issue perse) that I posted over on PhD. It looks like something they would need to deal with:
I have been using PhD/SGP as well as GNS for some time but lately have stopped using GNS since it is waking me up a lot for no reason since if PhD (as reported to SGP) takes longer to settle that normal it will trigger an alert.
This leads me to ask just how error reported to SGP is calculated? It appears to be a fairly large number of guide iterations RMS. Is that correct?
If so, it might be nice to optionally report (for dither settling) a much smaller set of RMS, maybe just the last two or three average errors. I very often see that even though guiding has actually settled, SGP takes a long to time “realize” that and start an exposure and I suspect that may be because the dither is being used as part of the error sent to SGP and until this “scrolls away” from the RMS, it thinks that guiding is worse than it is so holds off on the exposure start.
Do I have this right or is there something else going on?
OK, I will try that. I never had any problem w/o it as I always rough pre-focus manually and then run a manual autofocus just before sequence start and have a temp triggered focus during the sequence so it never really gets far enough off to need smart focus.
Thanks!
We fail if SGPro cannot settle in 15 minutes.
2.4.3.8 Has some problems with Nebulosity rejection and HFR sample size sliders, make sure you reset those.
Interesting. It seems that GNS is reporting much sooner than that. Maybe I am mistaken.
Yes, I noted that and as I recall I did that.
OK. I would even go so far as to set them, save, close and reopen to make sure they stuck. The bug was a little nasty (it has been fixed, but not released).
For reference… seems like the same issue. Nebulosity rejection was maxed out, AF got confused started to reposition the focuser…
Ah, yes, it does. I did not check to see where that slider was at - I will do that from now on.
Can I assume that if smart focus is disabled it would not do that or would it anyway if the nebulosity setting got that high?