Just flown a mapping mission with the camera triggered using cam_trigg_dist and an IR trigger, but I'm getting a discrepancy between the number of images in the folder, and number of cam trigger events in the log.
Using AMSL Altitude True
Reading log for GPS Messages in order to get AMSL Altitude
Log Read for GPS Messages
Reading log for CAM Messages
Log Read with - 220 - CAM Messages found
Images read : 224
CAM Msgs and Files discrepancy. Check it!
The camera was triggering about every 2.8 seconds, which it is capable of doing, but there are more images than trigger events. How can this be? All of the images look to be spaced out an equal distance.
Running AC3.3-rc11 and MP 1.3.32.
We've seen this before and i was because the SD card couldn't keep up with the rate of logging. Maybe try reducing some of the logging to see if the problem persists. another option is to try replacing the SD card. I've had good luck with SanDisk.
Hmm yeah I don't know if the card not being able to keep up is the problem. It's a fast sandisk ultra 8gb card which I've benchmarked and the write speed is good. The log is only 10mb for a 15 minute flight so I think it's unlikely the card speed is the problem. I don't have full IMU logging or anything like that enabled.
Is there a counter in the log which will show if the log is missing samples? MSG/TimeUS gives me a straight line.
I've just done some more digging and there's something weird about how the geo ref tool is handling this log.
Within the geo ref tool, when I try and use a time offset then generate the location kml, it shows the flight path gets cut off about 40% of the way into the flight, and then jumps to the end position. Also the photo locations don't match for one of the legs of the flight.
If I use the 'create kml + gpx' option within MP, I get the full flight path displayed as expected.
Is this an issue with the geo ref tool that needs to be investigated, or is there something wrong with my log?
Edit: updated MP from 1.3.32 to 188.8.131.52 (beta updates) and now I can use the time offset without it cutting the log off half way. Still has the non-matching number of trigger events to images though.