S‌kydemon is great, runs brilliantly on my Android tablet most of the time
‌
‌‌I was hoping to use it analyse things and overlay a plot with my GoPro footage but I find that during aerobatics practice it gets very confused about speed/altitude and sometimes stops logging entirely.

Looking through the exports files there are various points where the GPX file has no <ele> tag for an entry, which corresponds to a line in the KML file where the elevation is shown as zero.

S‌imilarly, there are times when both the KML and GPX files show speed as 0.000000‌
‌‌
I assume that that the sudden height changes coupled with short periods of zero groundspeed during a vertical ascent are probably confusing things - I assume there's some sort of 'sanity check' applied which is deciding the data is invalid?

A‌fter three entries where Speed is 0.00000 and ele is recorded as -46.11624‌ (!!) it terminated the log and then restarted in a new log.
‌
I‌ understand that this isn't an issue for 99% of flights and that those checks are probably in there for good reason, but is there any chance of getting an aerobatics toggle so that the checks can be disabled and it just logs exactly what the GPS tells it?‌‌

‌‌

I see the same sort of thing during my aerobatics, but I suspect it's as much due to the structure of the aircraft getting between the GPS antenna and the satellites.

They're not our checks. We simply log whatever your GPS is giving us. Unless you have a really great GPS receiver with really great view of the sky all the time (unlikely when you're upside down) I would expect occasional duff data like this. It's not SkyDemon which is causing it though.

OK, thanks Tim, so you don't do any checks or validation, just grab the info, regardless of any accuracy or 'loss of signal' report from the GPS?

I‌ did some straight GPS logging on an aeros flight before I asked this and it didn't seem to record any of these strange values‌, certainly none of the below sea-level altitudes were there and no zero speeds - of course, the logger was probaly using a much more frequent GPS report time and so the comparison may be unfair
‌
‌I've not programmed GPS on Android, but wrote NMEA based tracking systems for Couriers and Yachts years back, so I'm used to pulling the data apart - I‌ also logged GPS on my iPhone and that seemed OK (for SkyDemon in flight I use a Samsung Galaxy Tab 3), so I assumed the GPS wasn't being affected by the a/c attitude and that it was Skydemon filtering/smoothing the data and getting it wrong

T‌he Inverted portion of any flight manoevre is short, and the Bulldog has a very open canopy - when I was first testing Skydemon to see if I'd buy it, I used it on trains with a very limited view of the sky and had no issues (though obviously they didn't change attitude)‌

I guess the original question remains the same - what's the criteria for stopping a flight log and could this be extended for aerobatics flights and/or can the frequency of position reports be tuned to minimise the number of 'bad' reports received.
‌
I also wonder if anyone has found any major differences in the loss/re-aquire performance of Android/iPhone/iPad built in GPS?
‌
Anyway, I‌'ll have another look‌ at what's going on when I'm next airborne and see if I can figure any more out.
‌‌

The system to determine when you've stopped flying consists of more than one check, including time, speed and altitude. There's not much sense documenting it here because it's subject to change, but you really do have to have landed (or be hovering very close to the ground) before SkyDemon deems you to have landed. Alternatively you can completely lose GPS signal for a period (maybe 30 seconds, certainly no less) and SkyDemon will stop logging then too.