When using autorun, the Asiair stops shooting after about 1 hour, and freezes for the rest of the night, even though it says "shooting" on the bottom right. I'm using version 1.4.3 on android. Is this a known issue?
C6 - Celestron AVX
ASIair - ASI533 and ASI1600 with zwo mini filter wheel - ZWO mini guide scope - ASI120mm mini
This morning I was doing some flats and it would get stuck after 2 of 5 captures. The Autorun log for the short session is below:
Log enabled at 2020/05/01 07:33:31
2020/05/01 07:33:31: Start Autorun
2020/05/01 07:33:36: Shooting picture 1#
2020/05/01 07:33:38: Download
2020/05/01 07:33:38: Shooting picture 2#
2020/05/01 07:33:40: Download
2020/05/01 07:33:41: Dither
2020/05/01 07:33:41: Settle
--- Capture progress got stuck here---
2020/05/01 07:34:04: Autorun Stop (I manually stopped it because the progress was stuck on 2 out of 5 captures)
Log disabled at 2020/05/01 07:34:04
Log closed at 2020/05/01 07:34:04
Any ideas why it gets stuck?
Short answer: turn off dithering when taking flats and darks.
What astro hobbyist calls dither is the process of moving the guide star by N pixels every now and then in the X and in the Y directions. (Different from dithering that we learn in college signal processing classes.)
This will cause all stars in the plate to land in a different pixel location to get around anomalies of CMOS sensors. When the stacking program runs, it corrects for the offsets. Once the mount settles down after the dither, the exposure in the main camera starts, and no further dither is applied until at least the end of that plate.
When you are taking darks and flats, there is no guide star. Or you are guiding on a stuck pixel. Either way, dither will never settle down.
When taking Lights, you have to make sure you understand the parameters in Dither Settings. Read up on Dithering in PHD2.
In short, do not exceed your mount's capability when you set the Stability and Settling time. If your stability selection is too small, a cheaper mount may not ever get there (so dither never completes). This can be cause by poor torque from the motors and/or poor backlash in the gears. This why legacy German mounts need to be carefully balanced and carefully biased. Balancing is required for good guiding too. This requirement is not needed for the mounts that use Strain Wave gears (the reason why I left a Takahashi mount for a RainbowAstro RST-135). The RST-135 has so much torque that you do not need a counterweight.
In v1.4.4 we set the dither settles down timeout to 60 sec. This would fix the issue.
2020/05/18 23:10:19: Shooting picture 20#. (the last photo that succeeded)
2020/05/18 23:14:21: Download
2020/05/18 23:14:31: Wait for Meridian Flip 1# (there were four more meridian flip sequences like these next few lines)
2020/05/18 23:14:31: Stop Guiding
2020/05/18 23:14:31: Stop Tracking
2020/05/18 23:21:06: Auto-Center Start
2020/05/18 23:21:13: Download
2020/05/18 23:21:23: Auto-Center 1#
2020/05/18 23:21:23: Mount slews to target position
2020/05/18 23:21:27: Expose
2020/05/18 23:21:34: Download
2020/05/18 23:21:35: Plate Solve
2020/05/18 23:21:37: Auto-Center Failed merid flip failed
The final meridian flip related log entry does something different
2020/05/18 23:25:41: Wait for Meridian Flip 5#
2020/05/18 23:26:41: Auto-Center Start
2020/05/18 23:26:48: Download
2020/05/18 23:26:52: Auto-Center 1#. (these last lines repeated over and over 1,200 times until I stopped Autorun.)
2020/05/18 23:26:52: Mount slews to target position
2020/05/18 23:26:55: Expose
2020/05/18 23:27:02: Download
2020/05/18 23:27:02: Plate Solve
I have come to think that maybe the Meridian Flip settings of the mount could interfere with things - what if it did a meridian flip before the ASIAir expected it, etc? Tonight I made change to my mount's settings (Ioptron CEM60) so that the ASIair should have a chance to manage the meridian flip before the mount is set to do it; and therefore should operate normally.
I finally caught this in the act!
I verified that the ASIair was set to perform the meridian flip before the CEM60 was. I assume that if the CEM does it independent of the ASIair,then the ASIair would not know about the flip, and would therefore a) not plate solve to confirm it and b) later on it would attempt its own flip messing things up even worse. Better to let ASIair do it so it's workflow has a chance - i.e. plate solving upon completion to center the target again.
I think a point of failure I experienced last time - if clouds have come out or the focus of the primary or guider are affected by the flip, then ASIair cannot correctly verify the object is centered after the flip and therefore imaging workflow does not resume.