Session Ending Abruptly

[johndias]

Running 3.1.5 and I keep having an unpredictable issue where my sessions will just time out and stop.  I'm running Windows 10 and my camera is a 60Da.  Sometimes the session will time out after 9 or so subs and I'll restart only to have it time out after 2.  Then I can restart again and it will complete.  It's really random behavior.

 

Sending my log files to support and will reference this thread.

[admin]

Guylain, I ran six tests today and had no timeouts with the 150ms delay.

 

Let me know if there's any other info you want or testing done, happy to help.

 

No other info required.  If you are happy I'm happy.

 

I will leave the delay in there so it continues to work in future releases as well

 

Regards,

[johndias]

Hi John,

 

I just sent you an email with a download link.

 

Regards,

 

Thanks!  Testing now - I started with my Windows 8.1 laptop as it exhibits this problem less frequently than the Windows 10 system (tends to get a lot further into the session before timing out).  My first test ran through almost 48 frames (of 50) before timing out.  The second test ran all 50 no problem.  I've attached those logs.

 

The Windows 10 laptop is a little better but still timing out regularly.  I've tried both cables on that system as well with the test build you provided.  Those logs are also attached.

Windows 8-1_logfile-20160125-08h29m34s365-backgroundworker-5536-2016-01-25.zip

Windows-10_logfile-20160125-08h45m29s632-backgroundworker-5880-2016-01-25.zip

[johndias]

Please use Canon215 with the 60Da and keep me posted with the new build I sent you via email.

 

Thanks

 

Ah, OK - switching to Canon215 and will test again.

[astroman133]

John,

 

What other "solution" did you use without error?

 

Having the name of this program could provide valuable information to Guylain.  In addition it may be useful to know if it uses the Canon SDK and if so, what version of the SDK works.

[johndias]

I really hope I didn't waste your time from using the wrong drivers this whole time!  The Windows 10 system powered right through 50 sub session no problem (using Canon215 with the test build).

[johndias]

John,

 

What other "solution" did you use without error?

 

Having the name of this program could provide valuable information to Guylain.  In addition it may be useful to know if it uses the Canon SDK and if so, what version of the SDK works.

 

I'll let him know via email, I wouldn't want to be rude and name a competitor on his board - I really like BYE and have no intention of buying the other one.

[johndias]

I would like to rule out (or not) the newer canon sdk files in 3.1.4 and 3.1.6 as maybe being the cause.

 

Copy the content of the \canon folder from 3.1.0 into the \canon215 folder of 3.1.6.  Select the canon215 driver in bye once you did the copy.

 

If this solves the issue then the new canon sdk does not play nice with your camera model.  I don't think this is the cause but it will rule out the newer canon sdk as maybe being the culprit.

 

Regards,

 

I totally missed this message, my fault.

[admin]

I'll let him know via email, I wouldn't want to be rude and name a competitor on his board - I really like BYE and have no intention of buying the other one.

 

Absolutely no issues in you saying what other product you used here on the forum.  This is all part of troubleshooting and I don't mind at all.

 

All is good at my end.

[johndias]

Absolutely no issues in you saying what other product you used here on the forum.  This is all part of troubleshooting and I don't mind at all.

 

All is good at my end.

 

Cool - it was MaximDL.

[admin]

You've marked this thread as solved with the Canon215 drivers in BYE.  Is it really solved?

[johndias]

Yes, after I changed to Canon215 my Windows 10 laptop ran through a 50-sub session no problem.  I can try this with the GA build of 3.1.6 if you want me to confirm?

[astroman133]

I asked about which SDK (Canon or Canon215) was used when the error was encountered.  This was because the 60Da should work with either, but some people had experienced issues with the newer version.

 

This could have been addressed a week ago and without nearly as much effort if you had simply answered the questions that I asked in post #3.

[admin]

Yes, after I changed to Canon215 my Windows 10 laptop ran through a 50-sub session no problem.  I can try this with the GA build of 3.1.6 if you want me to confirm?

 

Yes, please try that too... but since it is working with the Canon215 it will/should with the link I sent you too.  The only difference between the released version of 3.1.6 and the link I send you is a 100 millisecond pause before the BULB command is sent.

 

The fact that it did work with Canon215 as-is tells me than the inner-workings of the latest SDK has changed and does not necessarily play nice (or at the very least differently) than it did with the SDK versions up to SDK 2.15 with the 60D/60Da.

[johndias]

I asked about which SDK (Canon or Canon215) was used when the error was encountered.  This was because the 60Da should work with either, but some people had experienced issues with the newer version.

 

This could have been addressed a week ago and without nearly as much effort if you had simply answered the questions that I asked in post #3.

 

Thanks for the help, I did answer that in post #6.  I would have been glad to try that earlier.

 

However, some additional testing I am doing is showing that it may be a combination of the driver and the timeout Guylain added.  I'll post my results later.

[admin]

One thing I do find odd in all of this is that the SDK is reporting a success on the 2nd try trying to send the bulb command when in fact it is not.  I'm almost willing to bet that when this occurs the shutter does not even open... which leads to the image ready for download event never being fired because there is none, the camera never took a picture in the first place

[johndias]

OK, I did some more testing of this.  I used ONLY my Windows 10 Surface Pro for these tests.

 

First tested with 3.1.6 GA installed and the factory USB cable.  I did three 50-sub sessions and shut down BYE between each session, all sessions used Canon 215 driver.

 Session 1 - 30 subs before timeout

 Session 2 - 9 subs before timeout

 Session 3 - 7 subs before timeout

 

Next, I remove 3.1.6 GA and installed the 100ms bulb delay version you sent me.  Same USB cable and Canon 215 driver.

  Session 1 - 50 subs, no timeout

  Session 2 - 50 subs, no timeout

  Session 3 - 1 sub and timed out

  Session 4 - 50 subs, no timeout

 

At this point, I switched cables to a longer USB cable I recently bought.  Canon 215 drivers again.

  Session 1 - 1 sub and timeout

  Session 2 - 50 subs, no timeout

  Session 3 - 50 subs, no timeout

  

Finally, I ran two more sessions with the newer Canon driver (not the Canon215), this is still with the 100ms bulb delay.

  Session 1 - 8 subs, timed out (new USB cable)

  Session 2 - 7 subs, timed out (factory USB cable)

 

I am thinking that a combo of driver and the timeout are in play.  My best results are clearly with the 100ms delay using Canon215 drivers. 

 

I don't understand the timeout after the first sub in a couple of those situations, but this is a HUGE improvement.

 

If you want the logs from this let me know?

[admin]

Send me the log file where it timed out after 1 shot using the 3.1.6 with 100ms fix, I want to make sure this is the same error.  If yes I'll increase it 150ms instead of 100ms.

 

Thank you for running all those tests.

[johndias]

Here you go Guylain, and thank you for working with me on this issue!

logfile-20160125-14h24m55s743-3376-2016-01-25.txt

[admin]

Yep same error.  Please download again from the same link I sent you yesterday, I've increased to delay by a few ms again.

 

I really don't like this fix,  when a delay fixes an issue it is usually the wrong fix and in my books I barely consider it a quick fix patch.  I'm happy it addresses the issue somewhat but I'm going to continue thinking about this one hoping to find a permanent solution without having to rely on a delay.

 

Thanks again for all those tests, it really helps.

[johndias]

I'll consider this a workaround  I will grab that download and get back to you by tomorrow with the results.

[johndias]

Guylain, I ran six tests today and had no timeouts with the 150ms delay.

 

Let me know if there's any other info you want or testing done, happy to help.

[johndias]

Bump?

[astroman133]

I did not respond to your previous note because you sent your log files privately to Guylain.

 

Did you post this same issue previously, or was that someone else?

 

I seem to remember an issue that caused disconnects after 9 or so images had been captured, but I could not find it by searching the archives?

 

Disconnects like this are usually caused by cabling issues. 

 

Have you tried a different, shorter USB cable? 

Can you re-create your issue in an indoor setting where the camera is stationary relative to the PC?

Are you running off a battery or an A/C adapter?

Is the battery a Canon battery or a 3rd party battery?

Is the battery fully charged?

What version of the Canon SDK are you choosing when you connect? (Canon or Canon215)?

What is the sensor temperature of the image that was captured before the shutdown?

[admin]

As Rick said, change cable and start with a full battery.

 

Keep us posted

[admin]

99% of all cases where you get a timeout after 120'ish seconds is due to the camera not firing the image ready for download event.  Without this event BYN can not download the image and after 120'ish seconds of waiting it gives up and end the imaging session. 

 

I don't recall getting your log files though, can you resend them.

 

regards,

[johndias]

I sent them to support@otelescope.com - if this is not the correct email let me know.  I'm sending again, with a fresh test I just ran indoors (not mounted to my equipment, etc), with a different cable than I've been using.  This also happens on both of my laptops.

 

For the indoor test I used my Canon camera battery (which is nearly fully charged).  But normally when I am imaging I use the AC adapter from Canon.

 

It's using the Canon drivers (not the Canon215).

[admin]

I sent them to support@otelescope.com - if this is not the correct email let me know.  I'm sending again, with a fresh test I just ran indoors (not mounted to my equipment, etc), with a different cable than I've been using.  This also happens on both of my laptops.

 

For the indoor test I used my Canon camera battery (which is nearly fully charged).  But normally when I am imaging I use the AC adapter from Canon.

 

It's using the Canon drivers (not the Canon215).

 

Thanks, I'll look for email.  This is the correct email address.

 

Regards

[admin]

Hi John,

 

You sent me all of your log files dating all the way back to 2014

 

Can you narrow it down for me and perhaps let me know on which nights the issue occurred? 

 

Thanks

[johndias]

Oh, yes, sorry.  If you look at today (the test I ran), you can also see it happened on Jan 18th and 19th during my imaging session that night.

[admin]

Oh, yes, sorry.  If you look at today (the test I ran), you can also see it happened on Jan 18th and 19th during my imaging session that night.

 

Cool, thanks.