invalid GPS log

Hello Washichi,

Your GPX file was corrupt. There were multiple identical timestamps with different coordinates??? The software did not know what to do and which coordinate to do it with. ;)

I removed the duplicates and also some spurious trackpoints that gave speeds up to 61mph.

So....here ya go. Fixed with the Fit File Repair Tool. Maybe it will now work for you.

PS. I am going to try a different fix to see if I can repair more points.

Hello Washichi,

Your GPX file was corrupt. There were multiple identical timestamps with different coordinates??? The software did not know what to do and which coordinate to do it with. ;)

I removed the duplicates and also some spurious trackpoints that gave speeds up to 61mph.

So....here ya go. Fixed with the Fit File Repair Tool. Maybe it will now work for you.

PS. I am going to try a different fix to see if I can repair more points.

Wow Datezz, thanks! I can load it now.
Didn't think it was that simple.

- my video and gps are now synced, but my km/h gauge gives mostly values from 0-20 , not higher.
more realistic km/h values would be 100 km/h when driving straight.
--> it seems that Virb reads the <ele> (elevation of the waypoint) as speed, and ignores these tags: <gte:gps speed="90.1" azimuth="102.8"/>

and If I can load the .gpx file, does that means there aren't any corrupted points anymore?

Thank you for your help, if you know any method / fix to make my GPX files 100% good please let me know.
I'm gonna take a new film + gps in daylight, so I can test around and do some visual sync :p.
I will post my results here.

Hello Washichi,

Okay I took the first file I sent you and replaced all the timestamps with 1 second recording.

It has the speeds you wanted.

Whatever is recording the GPX track is getting the coordinates okay but the timestamps are a mess.

So....here ya go. Another file fixed with the Fit File Repair Tool with a little help from my favorite Word Pad.

Your GPX file was corrupt. There were multiple identical timestamps with different coordinates??? The software did not know what to do and which coordinate to do it with.

No, not really. The problem is that the tracklog has a very high sample rate; 5 points per second. All timestamps are different, but Virb Edit apparently cannot handle timestamps with fractions of seconds and rounds or truncates them (looks like FFRT is doing the same)

Hello Javawa,

Good catch. I will filter that track again and see if the data makes better sense.

There are no Garmin units that record at 5 trackpoints per second so does it have to do with the camera frames per second?

The metadata from the file indicate that it is created by a smartphone app called "Trace Snow"...

I will filter that track again and see if the data makes better sense.

Filtering alone isn't sufficient because the timestamps will still be truncated, resulting in speed-spikes all over. I did some cleaning actions using Excel and a finishing touch with BaseCamp, resulting in this:


The cleaning process is too complicated to be practical; better use another app (or better yet: a proper GPS device ;) ) to record your track log.

Hello Javawa,

Not my track. I was just trying to get the guy a usable GPX track from his data. :)

Filtering alone isn't sufficient because the timestamps will still be truncated, resulting in speed-spikes all over. I did some cleaning actions using Excel and a finishing touch with BaseCamp, resulting in this:


The cleaning process is too complicated to be practical; better use another app (or better yet: a proper GPS device ;) ) to record your track log.

Hello Javawa,

I looked at your file. Nice job. Your speed changes look smoother and probably more like how the OP actually drove.

Not my track. I was just trying to get the guy a usable GPX track from his data. :)

I know, my advice was meant for washichi... ;)