Forum: General Topics

Forums / General Topics / Increasing density of track points - Solved

 

Subject:Increasing density of track points - Solved 

ranchodoug

20:08
Sunday
11-Mar-2012

Location:
San Diego, CA

Phone Model:
Blackberry 9930

Stephen,

I tried a couple suggestions I've seen in the FAQ/User's Guide/Forums to increase the density of track points, but the spacing from trackpoint to trackpoint while walking/meandering is still too high. In many cases, trackpoints are created a couple of meters apart. In other cases, points are 50 or so meters apart. I'd like them all to be just a few meters apart. Storage is not an issue.

The settings I'm using:

Web transmit interval: 15 secs
Web send track: on
Web Send live location: on
Web Auto-adjust Update Interval: on

Setting, activity profile:
Walking
Min track distance: 1m
min track time: 5 sec
New section time: 600 secs
Max track speed: Disabled
Unpause distance: 25m

What am I overlooking?

Thanks,

Doug
 

Stephen

12:03
Monday
12-Mar-2012

Location:
Surrey, UK

Phone Model:
BB 9800 Torch
BlackBerry 8900
SE W910i
Nokia 5800

Hi Doug,

The Min Track Distance and Min Track Time are the settings relevant to this, however your values of 1m and 5sec *should* result in quite a dense track as it is, so I'm not too sure why you're getting such large gaps. I use 12m and 5secs for my own walking profile, which works fine for me.

Are you in an area with good GPS reception - ie a clear open view of the sky, etc? It might be simply that the GPS is temporarily out of signal for a few seconds - I get this occasionally, though its generally not a big deal. One thing you could try is to enable the GPS Logging (set it to Raw GPX or Raw CSV/TXT). This will then log every point received from the GPS device (before being filtered by TMJ) to a file on the memory card. By comparing the points in the log to the recorded/filtered track this should at least confirm whether the issue is GPS- or TMJ-related.

Have you only noticed the problem in the last few days? I guess the recent solar storms might be relevant, though I didn't notice any effect on GPS here...

Cheers,
Stephen
 

ranchodoug

20:03
Monday
12-Mar-2012

Location:
San Diego, CA

Phone Model:
Blackberry 9930

Stephen,

I sent you an email with a small annotated segment from a walk yesterday. The segment is in an area which is essentially So. Calif. desert so there is mostly a clear, open view of the sky. There is a note where there was some tree coverage. I double checked the trackpoints in the online track and they are pretty consistently 30 secs apart. Unfortunately, GPS logging wasn't enabled. I thought there was a correlation with small tight turns but I think those points are where we paused.

I did a small experiment this morning while driving to work with GPS logging enabled. The trackpoint (TP) to TP time was consistently about 30 seconds until about 20km into the trip when abruptly the TP to TP delta time dropped to a consistent 3-4 seconds. About 20 minutes after arriving, the TP to TP delta time jumped back to 30+ secs, with a few isolated points at 200+ seconds.

Does TMJ add each GPS log point to the track?

Doug

 

ranchodoug

10:55
Tuesday
13-Mar-2012

Location:
San Diego, CA

Phone Model:
Blackberry 9930

Stephen,

I experimented again on the drive home with Min Time/Distance disabled. While driving (moving or stopped over a 20km, 30 min trip) and during a stop at a store, TP to TP delta time was consistently about 30 secs. My Blackberry 9930 was inside my car the whole time. The GPS log and TMJ track were identical over the entire trip.

Does TMJ influence the GPS update period or just respond to GPS interrupts? If the latter, is there any way of increasing the GPS update rate?

Thanks,

Doug
 

Stephen

17:53
Tuesday
13-Mar-2012

Location:
Surrey, UK

Phone Model:
BB 9800 Torch
BlackBerry 8900
SE W910i
Nokia 5800

Hi Doug,

I've been thinking about this, its a bit odd. TMJ essentially asks the phones GPS to provide it with regular updates, ideally at 1 second intervals, it is then entirely down to the phone to decide how often it actually sends a new update to TMJ. The Raw GPS Log includes every point that the phone has provided to TMJ, so since this seems to be the same as the filtered track (which enforces the Min Distance and Min Time options) this pretty much rules out the TMJ track filtering as the problem.

I wonder, could you check the Menu/GPS/GPS Options/Force 1 Sec Refresh option. I added this when BlackBerry released OS6 since this gave similar issues to yours (though I'm not sure identical). The option makes TMJ 'ask' the phone for updates in a different way. I'm wondering if maybe BlackBerry have changed again for OS7...

Whichever option it is currently on, could you toggle it, then restart TMJ and try again... does this make any difference?

Irrespective of the track filtering, on the Plan Screen you *should* see the location marker flash between red and blue roughly once a second if it is all working correctly (it changes colour each time it receives a valid GPS update), assuming you're in an area with adequate reception...

Any luck?

Cheers,
Stephen
 

ranchodoug

19:25
Tuesday
13-Mar-2012

Location:
San Diego, CA

Phone Model:
Blackberry 9930

Stephen,

Your suggestion solved the problem. The Force 1 Sec Refresh option was not on so I checked it. The resulting delta update times aren't every second, but good enough. Example delta update times just walking around the building:

1.907, 2.076, 1.997, 1.018, 1.994, 5.008, 3.009, 3.950, 3.934, 3.005, 1.990, 2.009, 1.997, 1.001, 1.017, 1.985, 1.004

That's good info on the blinking location marker on the plan screen. I've been looking at the 'Points' on the Track1 screen.

FYI, I ran another experiment while driving to work. The same thing happened as yesterday. After 55 points, 27 minutes, 17 miles/27 km, the GPS updates dropped from an average of 30 seconds down to an average of about 2 secs. After I stopped and was sitting the update times jumped back up to anywhere from 20 to 60 seconds. It looks like the Blackberry is slow to lower update times when going from stationary to moving but fast to increase intervals when going from movement to stationary. Probably a battery life issue.

Thanks for solving the problem. This will certainly improve the quality of twisty/turny tracks.

Doug
 

Stephen

20:54
Tuesday
13-Mar-2012

Location:
Surrey, UK

Phone Model:
BB 9800 Torch
BlackBerry 8900
SE W910i
Nokia 5800

Hi Doug,

Thats a relief - glad it worked!! I'll make sure the next release defaults to ticking that option in OS7 (the current one only ticks it for OS6)...

Those delta times look about right - they will probably improve further when outside with a better signal too.

Cheers,
Stephen
 

ranchodoug

1:45
Wednesday
14-Mar-2012

Location:
San Diego, CA

Phone Model:
Blackberry 9930

Stephen,

Another experiment on the way home yielded a *perfect* track (other than 'driving' on the wrong side of the road occasionally and 'running over' a few trees/bushes - easily attributed to small and acceptable GPS errors and my poor driving). I'd give TMJ's performance an A+. This is exactly the close fitting track I was hoping for.

I left a 5-star review on BBerry Apps today for you and TMJ, in large part due to your exceptional support.

You and TMJ are a Class Act.

Doug

p.s., Can you change the subject: to "Increasing density of track points - Solved"?
 
 

Stephen

18:49
Wednesday
14-Mar-2012

Location:
Surrey, UK

Phone Model:
BB 9800 Torch
BlackBerry 8900
SE W910i
Nokia 5800

Hi Doug,

Thanks!! :-)

Cheers,
Stephen
 

(You must be logged in to post a reply to this thread)