App to track running

MeanderingMammal

Jthanmyfitnesspal wrote: »

MeanderingMammal wrote: »

Jthanmyfitnesspal wrote: »

I don't believe that the GPS is any more accurate on a watch than on a phone.

I'm a mobile phone the tracking app is in contention with a number of other services for the attention of the GPS, which may not be a dedicated function on the sensor set. As a result is getting less data than in a dedicated device like a watch.

That may not be hugely important, but in general the pool of error for a phone is going to be bigger than for a watch. The cumulative effect of that has the potential to be significant.

Not buying it. I have a GPS app that tracks both accuracy as well as location on the phone ("US Topo Maps"). The biggest factor affecting accuracy is sky-view. In other words, GPS sucks under foliage, in canyons (natural or artificial), in heavy rain, and also when your phone or watch is blocked by your body. When the sky-view is good, the accuracy goes below 10m (absolute), which is about as good as is theoretically possible. In the past, I found that the phone was as good or better than a Garmin hiker's GPS, whose sole purpose was GPS tracking.

Constellation visibility is certainly important. The more space vehicles that the device can see the greater the accuracy in three dimensions. Within the GPS constellation the maximum visibility is 12, and most devices cap out at 12 channels. Most of the time only 9 SVs are visible, hence the value in accessing GLONASS.

Physical obstruction of the signal is easier for the processing to deal with as the signal is just absent. If it's attenuated by water, either rain, foliage or human tissue, the change for the processor is the effect on the timing component of the message content.

Notwithstanding all of that, given that diff-GPS is no longer necessary then a 10m pool of error is quite poor.