Camera Stops Triggering During Mission But I Have all the Trig Messages

[GFlight]

Hi,

I have been using the Arducopter system for a few years now but first time posting an issue I can’t seem to figure out so I am reaching out to you guys.

During a mapping (grid) mission with a camera being triggered based on distance the camera seems to stop triggering in the middle of a flight line without any reason (no commands to do so). Sometimes it starts back up and sometimes it doesn’t. If I fly the same mission twice it happens in the same spots in each mission. I have a camera trigger cable and a camera shutter feedback cable and I can see that the pixhawk told it to trigger but it didn’t. I would think it was a camera issue but it is repeatable mission after mission. This is a recent issue in the past few months and I have not had this in the past. The only difference is that it is winter where I am at.

I am using a hexacopter and a plane and having the same issue on both all different equipment except the camera and mission planner so it might be a camera issue but wanted to make sure nobody saw anything in the logs first before I buy a new camera.

I loaded 3 logs here https://www.dropbox.com/sh/jugqsuv4divvflt/AABhebQLMK89RmzkZH38ogIca?dl=0 ,
1 from my copter today and another from my copter and plane a bit ago.

Thank you for your time.

[Naterater (Nathan E)]

Definitely odd. I looked through your first log, and everything you are telling me looks right. My first guess would be a focus issue with your camera with the low contrast background (snow). I don’t see antyhing in your mission that would’ve made it stop mid-way through flight lines unless there’s some program overload, and I can’t diagnose that from logs; you’d have to ask a developer. Maybe @WickedShell might have some input.

[WickedShell]

The log contains 162 TRIG events, 54 CAM events. Based on this we can conclude that the autopilot was attempting to trigger the camera correctly.

As @Naterater has noted if you are using auto focus on the camera this can result in the camera not solving the focus problem and then not taking the photo. This is particularly true if you are using a servo to actually depress a shutter (as the camera may not have solved focus once it’s drifted far enough within the time period). In general mapping with manual focus produces much more reliable results. (Fewer error results, as well as better for photogrammatery).

Other things to look at is if the camera was off when you landed, had a menu open, or was reporting SD card errors. Basically this is a list of internal camera errors that can cause problems. But from the autopilot logs it was still attempting to take photos so I’d recommend looking at the camera setup in detail.

[GFlight]

@Naterater and @WickedShell I appreciate your quick and informative responses. I flew again today with the same hexacopter but with a few different setups. I used a different camera than yesterdays post and was set to manual focus and I am seeing the same results. Please see the logs from todays flights here labeled 3_14_18_flight_1 and 3_14_18_Flight_6

Ardupilot

For flight 1 I had 164 trig messages, 164 cam messages, and 164 images. It was triggered using a modification to a Sony QX1 that adds a microcontroller to it and is connected directly to the pixhawk Ch7.

For flight 6 I had 164 trig messages and 164 images (for some reason my shutter feedback didn’t work right but I will figure that out). This time I used a Seagul #REC to trigger the camera also connected to CH7 of the pixhawk.

The images below are of flight 1 and flights 6. You can see they look identical. The flight plan is a rectangle and I have waypoints 15m beyond the flight polygon but it seems to turn well before it gets to them (see Flight plan image from flight 3_14_18.png) in the link above, it was too large to attach)

This one is a little different because it looks like it just stopped triggering well before the end of the flight line.

I have my WPNAV_SPEED set to 800(m/s) which I originally and normally have at 12m/s but I thought that might be the problem
I have WPNAV_Radius at 400(4m)

I also added two flight logs from January from my plane. labeled 1_5_2018_Plane_flight_1, and 1_5_2018_Plane_flight_2. You can see whole flight lines were missed but others were perfect and they are both identical. Please note that these flights were with autofocus with the seagul trigger.

I can obviously overfly to get around this but this has not been an issue in the past and would rather not overfly more than I need.

Thank you for your help.

[WickedShell]

Only speaking about flight 6.

mavflightview won’t work with the log for me for some reason, and I can’t recall where the tracking of waypoint numbers are but looking at just the status text messages the vehicle only reached waypoint 48 and started flying to waypoint 49, it never flew the last line you had planned.

MSG, 232549723, Reached command #1
MSG, 238611435, Reached command #2
MSG, 241658712, Camera: setup fast trigger capture
MSG, 273406719, Reached command #4
MSG, 275306134, Reached command #6
MSG, 280640137, Reached command #7
MSG, 287125847, Reached command #8
MSG, 323160245, Reached command #10 
MSG, 328713829, Reached command #12 
MSG, 333288862, Reached command #13 
MSG, 338813290, Reached command #14 
MSG, 374111393, Reached command #16 
MSG, 375993475, Reached command #18 
MSG, 380727438, Reached command #19 
MSG, 387688547, Reached command #20 
MSG, 423527195, Reached command #22 
MSG, 425413932, Reached command #24 
MSG, 430856819, Reached command #25 
MSG, 437949974, Reached command #26 
MSG, 473179506, Reached command #28 
MSG, 475064089, Reached command #30 
MSG, 479879909, Reached command #31 
MSG, 486433556, Reached command #32 
MSG, 522275758, Reached command #34 
MSG, 524214446, Reached command #36 
MSG, 528903970, Reached command #37 
MSG, 536529546, Reached command #38 
MSG, 572035348, Reached command #40 
MSG, 574461257, Reached command #42 
MSG, 579363266, Reached command #43 
MSG, 586468455, Reached command #44 
MSG, 622188832, Reached command #46 
MSG, 624049517, Reached command #48 

[GFlight]

Thank you for looking again. I removed flight 6 and re-uploaded it just in case something happened on my end. I intentially brought it home and did not do the last flighe line as planned. I was just doing tests and did not need the last line. Thanks for seeing that though.

[WickedShell]

@GFlight I’m not following. Are you still having an issue that needs to be looked at? It didn’t map the last line because you didn’t fly it, you have a matching number of trigger events and photos, so I’m not sure what we are supposed to be looking at here.

[WickedShell]

Ah okay, sorry I didn’t realize that was what you were asking about before, I’ve been very confused till now.

This looks to me like it’s due to the leash hitting the waypoint and moving on to the next waypoint which promptly turns the trigger distance off for the turn around. If you didn’t have all the waypoints that were attempting to turn the camera off before doing the turn you wouldn’t be seeing the problem.

[Naterater (Nathan E)]

I understand the problem with the first flights, but your last post and images doesn’t appear to have any problems. If you’re concerned about the offset in the flight lines, that is because of the way you are programming the mission in the “survey (grid)” tool. http://ardupilot.org/copter/docs/common-camera-control-and-auto-missions-in-mission-planner.html.

[WickedShell]

@Naterater the mission is supposed to be a grid, and with the plotted cam on/off waypoints he was expecting a perfect symmetry between left/right lines. If you look at the planed mission below it was expected to have photos between waypoints 2 and 4, and then 8 to 10. This would have drawn a perfect pair of lines, except the leash behavior caused the navigation controller to move on before the full line was imaged.

[Caca_Nakagi (Caca N)]

I’m having the same kind of problem…
I’m testing small missions Droping small balls using trigger/distance… My camera is setting to trigger by pwm using 1.5 seconds (to push servo) and i have 5 seconds from one trigger to another… And sometimes the servo didn’t push on trigger message…