SharpCap 4.1 Beta Update - 13th December 2022

[admin]

Hi folks,

was hoping to get this version uploaded yesterday, but got stuck tracking down a couple of last minute bugs.

This week's changes include

Better handling of devices that disconnect from USB

SharpCap can now automatically detect and handle cases where a camera temporarily disconnects from USB. This is currently availble for ZWO and Altair cameras (that are being used by the direct SDK connection - not ASCOM or DirectShow). More brands of camera will be added soon. You will need to enable it in the Camera settings too, since at the moment it is off by default.

If the camera that is in use disconnects from USB then SharpCap will show a warning and wait up to 20 seconds to see if the camera reconnects. If the camera does reconnect in that time, SharpCap will re-initialize the camera, set all the camera controls to the values you had set before the disconnection and resume whatever it was doing (including capturing videos, live stacking, etc).

You should be able to test this by pulling the plug on the camera while it is running, waiting a few seconds and then plugging back in. Not advised as part of your normal imaging procedure, but it may help prevent issues if you have a temporary glitch either due to cable issues or a camera hardware gremlin.

For SharpCap Pro users, the reconnection is seamless (no prompt). For free users, there is a prompt to confirm reconnection.

I can imagine that there may be cases where the physical connection to the camera is fine but the camera shows as having disconnected from USB - perhaps due to some software bug on the camera itself. In those cases the camera may not reconnect to USB after glitching, but if you happen to spot it happening, pulling the plug and reconnecting while SharpCap is still waiting for a possible reconnect may be a solution - I would be interested to find out, as if that works at all I will make the 'waiting' time longer.

Other Stuff

* Fix bugs in feature tracking
* SDK updates from ZWO, SVBony and Altair
* Add initial Italian translation (created by Google Translate - corrections welcome)
* Improvements to ghosted image alignment, including ability to rotate ghost image through 180 degrees
* When calculating histograms, ignore area of image that contains timestamp (including QHY GPS data timestamps)
* Fix issues with keyboard zoom shortcuts
* Fix crash with CPWI not handling full disconnection gracefully
* Add 'repeat until time' step to sequencer

And, before I forget...

https://d.sharpcap.co.uk/download.html? ... 0&arch=x64

cheers,

Robin

[celkins]

Hi, Robin,
a rare clear night, here in Cheshire, so I'm out under the stars, with the latest beta...

I recall seeing in the daily digest, someone raised strange live-stacking behaviour - frames not being stacked, when PHD is clearly guiding: I've just had this happen in both running instances: I cleared it by clicking on pause, then resume - unfortunately, too quickly to gather additional logging at INFO level.
The following is a snippet from the log window:-

Code: Select all

03:39:37 INFO  Live Stacking is paused.
03:39:37 INFO  Frame not stacked because : Live Stacking is paused.
03:39:48 NOTE  Live stacking deactivated
03:39:48 INFO  Guiding disconnected due to live stack close/camera close
03:50:05 NOTE  Live stacking activated
03:50:06 INFO  Guiding reports that guiding has resumed.
03:50:18 NOTE  Clearing stack
03:50:19 INFO  Guiding reports that guiding has resumed.
03:54:53 INFO  Live Stacking is paused - Guiding is not active, currently Looping
03:54:53 WARN  Frame not stacked because : Live Stacking is paused - Guiding is not active, currently Looping
03:59:57 INFO  0.0000% of pixels (0) not stacked due to sigma limits

This last line shows something to be obviously wrong, because you wouldn't know no pixels were rejected by sigma limits unless you'd stacked the frame, which had indeed happened! Full log available if useful, but I only turned additional logging on at this point - sorry...
Looks like a state machine issue to me - if I were to hazard a guess

Cheers,
Carl

[admin]

Hi Carl,

yes, there was something similar a few days ago, but that was with SC4.0, and I have rewritten the bulk of the guiding handling in live stacking between 4.0 and 4.1, so it's a surprise that it is happening in both.

I would appreciate the full logs - I suspect you are right that it is either failing to track the guiding state properly or there is some sort of race condition going on

cheers,

Robin

[celkins]

Robin,
another one for you: on the right, everything is fine: on the left, no histogram!

Closing, then re-opening the histogram from the toolbar fixes this...

[admin]

Hi,

interesting, I've seen that once myself too, but so far it hasn't happened when I've had the opportunity to try to dig into *why*... Nothing bad in the log that would point me in the right direction, so I will have to put my thinking hat on as to what could be happening. Interesting that the top smart histogram bars are there...

cheers,

Robin