Live Stack Failure all of a sudden

[SkyShed Admin]

Hi Robin,

As you know I love SC and have used it for years. I use the same gear, same profile for DSOs every night. Last night and tonight I had a problem I couldn't figure out.

Using the latest version of SC. Then the Beta tonight when things weren't working.

When I fire up, I always load the same profile. (SC driver for) Altair 26C, 16 Bit (ULN), 200 gain, 120 secs, 10 dark offset, -20 temp. Then I adjust the time and gain as I'm focusing with a Bahtinov. After I finish focusing, I re-load my profile (to make sure I get the settings right) and wait a couple of 120 sec frames for the gain to come down before hitting Live Stack.

Last night and tonight, after I had it imaging for 5 secs, 4000 gain for focusing, then waiting 10 mins to watch if focusing stayed good, I did my usual and switched settings to DSO and began to Live Stack.

But both last night and tonight it wouldn't stack. Not even the first frame. it would over count to 10 (after 120 sec) then start again on 0-120 secs.

I thought maybe it's a cable, but there were not connection problems as it was in the same 16 bit mode during focusing. Receiving shorter images. I even hit Live Stack by mistake while it was at focusing settings and it began to stack.

Therefore I seems like it's not the cameras and it's not the cable? I would swap the cable out, but because it's connected to a Hyperstar cable-guide, it's not a quick thing to check.

I grabbed a couple of logs. Anything you can see which might point me toward problem? Btw I was imaging in Cygnus so there was no shortage of alignment stars.

Also I dl'd the SC Beta to see if it was version problem. Same thing happened.

Thanks!
Wayne

[admin]

Hi Wayne,

what I am seeing in the logs ties in with your description of the frame timer over-counting by 10 and then going back to the start. The logs are showing dropped frames of type 'AltairTimeout' - which is basically that SharpCap asks for an exposure and then waits for the expected duration plus an allowance for the transfer of the data to the camera. For a large resolution camera on USB3, SharpCap allows 10 seconds for transfer, so if a 120s exposure is not ready in 130s then SharpCap assumes it has failed and tries again (a better option than potentially waiting for ever for a frame that never arrives).

I think it is probably worth checking all the usual USB connection issues (including running the camera on a separate cable - even if not on the scope) to test. It sounds like the timeouts are only happening beyond a certain exposure length, so you could try testing with different lengths of exposure. I am also seeing disconnectes of the camera in the log, but it's hard to be sure if that is you pulling the plug and reconnecting or if it is dropping out due to a bad connection.

You might also try adjusting the USB speed setting for the camera to see if that helps.

cheers,

Robin

[SkyShed Admin]

Thanks Robin! I will try those things out tonight!

Wayne

[SkyShed Admin]

Hi Robin,

1st thing I did was change the USC cable (shorter). Same problem.

I tried taking shorter subs at my otherwise normal settings . 79 secs - worked, 90 secs worked, 110 secs worked, 115 worked. 120 secs didn't work. Set to 119 secs and it stacked 11 subs, no problem.

It seems there's something it doesn't like about 120 secs? I didn't try higher because I don't take subs higher than 120 s.

Going back out to image shortly. Will update.

Thanks again for your advice.

Wayne

[SkyShed Admin]

H Robin,

Replaced and re-routed the USB so it was shorter. Switched USB speed to 2 (from 1). Messed around with different sub time settings settings, and had stacking success at less than 120 secs. 70, 90, 100, 115 secs.

Tried my usual settings and it wouldn't stack. Tried switching to 16 RAW rather than 16 Bit RAW (ULN) at 115 secs instead of 120 and it worked. Got to 55 subs and it stopped stacking. Guiding was good. Focus was good. I'm pre-processing the subs now in PI. I got 45 good of 55 subs.

Thanks in advance for more help!

Log file attached.

Best!
Wayne

[admin]

Well, progress of a sort then...

I think I have an Altair 26 (mono maybe) here in the observatory at the moment, so I will test on 120s frames to see if I see anything similar.

There's nothing obvious in the log that points to why it stopped stacking - or at least I couldn't find anything. Can you estimate or work out the time that it stopped from the files that you have? Maybe if I had the time I might find something by being able to narrow down my search in the log file.

cheers,

Robin

[admin]

Hi Wayne,

just checked with my AA26C (turns out it was a C - I think I had an M here at one point but that one went on to someone else). Anyway, I get a similar issue with my observatory computer with longer exposures sometimes failing by going 10s over duration. However, when they do work they download to the PC within less than 0.5s, so it doesn't seem to be the download time.

I found that setting to 120s worked, but then it failed if I adjusted USB speed once set to 120s. It then kept failing and only worked again if I set the exposure to 60s and then back to 120s.

I have to do some more testing (another PC, with/without Pegasus Powerbox as a hub, etc, other models etc) to try to work out the exact conditions that cause this and then see if I can rig something to fix it. Presumably if I can replicate whatever the effect of the exposure change to 60s and back does at the start of each frame, I might be OK.

I'm not seeing quite the same as you - I can get failures at less than 120s too by adjusting USB speed, but it is definitely similar.

Might take a little while however - I am off again on holiday tomorrow, so will not be able to sort this until I get back.

cheers,

Robin

[SkyShed Admin]

Thanks for your quick attention Robin! It's prime time for imaging here and I only use SC. So I look forward to you returning and "digging in to it".

Safe travels in the meantime!

CS!
Wayne

[admin]

Hi Wayne,

I did have a bit more time to dig into this one...

I found the following

1) The only way I could make the 10s over time and dropped frames happen was to change the USB speed setting during an exposure of at least 10s (not sure of the exact cut off, 5s is fine)

2) Once the problem happens once due to a change of USB speed it keeps happening for every frame

3) The problem can be cleared by changing the exposure - for instance change from 120s to 60s and then straight back again - the next 120s exposure will be fine

4) Increasing the timout allowed for the frame to download doesn't help - I went up to 1 minute and you still get dropped frames.

5) I can work around the problem here by changing the code so that the update to USB speed is not sent to the camera during the frame - instead it is saved until the frame ends.

Now, this doesn't quite tie in with what you are seeing, except that you are using saved Capture Profiles to set your camera settings. Maybe your '120s exposure' profile has a different USB speed to your 'focus' profile, and maybe the code sets the USB speed after the exposure, so maybe it triggers the bug indirectly via that sort of route? Maybe even re-loading the profile tries to re-set the usb speed to the same value and causes the issue?

You could try editing the profile file with a text editor and removing the 'USB Speed = X' line - that should stop SharpCap trying to set that control during profile load, which might help.

Could you post a copy of the profiles that you are using? Perhaps I can recreate the problem using those profile files if it manages to happen some other way than setting the USB speed that I just haven't found yet?

cheers,

Robin

[SkyShed Admin]

Thanks very much Robin!

The interesting thing is that I didn't knowingly change the USB setting until I was trying different things to get it to stack. I will grab my profile tomorrow and attached, it would be easy enough re-create "fresh" if needed.

I will try editing as first.

Best!
Wayne