Sometimes (not always) when I'm standing still the map orientation swings around wildly. Map orientation is set to 'Track up'. The map display should be stable with the map features matching what's around you. Sometimes they are but sometimes the map swings around like crazy. What is behind me shows as in front then swings around again. Any ideas on what is happening?
TL;DR - Most likely poor reception resulting in apparent large, fast shifts in position. The map view is trying to follow these shifts.
Most likely, this happens when you are in challenging conditions. Perhaps you are near a cliff or steep hill, or in the bottom of a gully. Or in an urban canyon. These situations obscure part of the sky and usually involve multi-path reception. Multi-path means that the signal from one or more satellites reaches the receiver directly AND by way of reflection off of a nearby object. In this situation, the unit tries to distinguish the real (direct) signal from the reflection. But its best guess tends to change with time. When this happens, the unit's idea of it's current position shifts "instantly", usually by a relatively large distance.
Garmin handhelds are notoriously bad at suppressing these shifts. If you look at the activity track, you will likely see a spiderweb pattern around the position at which you stopped.
It has been suggested that turning on auto-pause will suppress the spiderweb pattern. This causes the unit to stop tracking the activity when it senses that you have stopped. Might or might not work for the track. Probably won't help with the map rotation.
You can also use north up. I realize that this is not the way you want to operate. But it will prevent the heading-up rotation issues in these circumstances.
thanks for that great info. I was not near any cliffs or steep hills or in an urban area with any tall buildings but I was in a heavily treed area (big tall trees and lots of them). The track does show spider webs and the position was jumping around and the map was rotating as you described. At that point the satellite page was reporting an accuracy of 11m. However, I moved to an area of open sky (by a lake out of the trees). The accuracy improved to 3m and the spider webs stopped but the map kept rotating all over the place. Any further ideas? or suggestions of further testing I can do? I had WAAS/EGNOS enabled at the time with the satellite system set to "GPS"
No further suggestions. Typically, dry deciduous tree cover would not cause this sort of thing. Wet leaves or needled trees (wet or dry) might. OTOH, if that's what you saw, I'm sure it was a contributing factor.
It might be that the map orientation is more sensitive to the apparent motion than the track log. That is, maybe smaller (or slower, or both) deviations are sufficient to upset the map orientation but not sufficient to show spiderweb tracks.
It's also (remotely) possible that the compass needs to be calibrated. When the unit is moving slowly (typically below 2mph), the heading is supposed to be derived from the electronic compass. Above that speed, it is derived from the apparent motion of the device, based on GPS information. The large, fast excursions caused by multi-path make the unit think it is moving over 2mph, so it uses the false satellite-derived heading. If the compass is badly calibrated, or if you are using the unit in situations where the electronic compass is unreliable, that might also be the cause of the symptoms you see.
Examples of such situations: Inside your car, under high-tension power lines, on a metal surface (table, the hood of your car), or near some other large ferrous object (the example that sticks in everybody's head is that giant WWE belt buckle).
The reported accuracy does not have anything (directly) to do with the issue at hand. In fact, it's largely a marketing tool. The true meaning of 3m (for example) is something like "the probability that the position shown on the GPS is within 3m of your true position is x%". To put this another way, the reported accuracy of ym means that y is the smallest value that meets the x% probability critieria.
No manufacturer, including Garmin, will tell you what value of x% they use. There is clearly a huge difference between 50% and 95%. It's a marketing decision because it affects your perception of the device. If they set it too high, the device will report larger position accuracy values (worse accuracy) all the time. Leading you to believe that it's a poor device. If they set it too low, you'll eventually notice that it's reporting values that are clearly unrealistic. Also leading you to believe it's a poor device.
Best way to treat the accuracy value is that it is relative - "smaller is better". But not to actually believe the absolute distance shown.
so I did some more testing. I went out in a bushy area that had trails marked on the base map (garmin topoactive maps). The map was doing it's swinging all over the place. I changed Setup > Routing > Lock on Road to off (it was on) and the map rotation seems to have gone away. Could it be that positional inaccuracies were giving the unit a hard time trying to lock on to a trail?
No clue. I never route with lock on road with a handheld.
yes, the so called 'accuracy'
the 'accuracy' of positioning is AFAIK something designed by the smartphone industry, as there people are not able to understand what it means but want just have a number showing 'my one is better'.
The question of accuracy, well compared to what? What would mean my position is accurate to 3m and your one just 10m? Some people think this has to do with some 'real position', but they are not able to say what they mean by that, have no idea what does 'position' means at all.
Earlier, Garmin devices did show the more scientific value PDoP, the potential dilution of position. This is in fact a real statistical value. But it is in general much higher value, the area where the receiver might be is much bigger. Therefore it is of no value for advertising, customers would compare it to values given by the smartphone and will say the GPS device is junk. Marketing of GPS devices for non professional customers had to follow this new meaningless information.
The 'accuracy' expresses when someone will at this moment navigate to this position, he will probably find it within a 3m circle.
Note 'at this moment' and not tomorrow and not afternoon and not next day etc.
But not only smartphone users
some years ago, there was a bridge build over river Rhein between Germany and Switzerland.
carefully measured all on both sides with Leica differential GPS, the concrete anchors on both sides of the river were build. Then s special transport brought the new steel made bridge and big crane tried to install it. Bad luck, it did not fit there. Why? They forget that german altitude reference differs from the swiss one and thus the chart datum used during planning was wrong. Both sides were accurate, very accurate, but each to its own reference.
The reported "accuracy" on a GPS device is still based on vertical and horizontal DoP. However, modern devices typically do not expose those raw numbers. No point in showing things that the average user does not understand.
Note that the accuracy figure started with GPSrs, not cell phones. Although cell phones are much more prevalent than GPSrs. So cell phone usage certainly helped "popularize" the concept. One of the nastier things about cell phones is that they do not depend entirely on GPS for position. They use "soft" information such as cell tower triangulation and nearby WiFi networks. Makes it very hard to relate "the accuracy my phone shows" to the "accuracy my GPSr shows" even when they are sitting side by side.
Bottom line is that the accuracy numbers are relative. Smaller (on the same device) is better. Cross-device comparisons are not meaningful. And pay no attention to the absolute number.
I have simply problem with the word accuracy used in this context. The correct word here should be precision.
Accuracy means something completely different and can not be used to compare measurements between each other like PDoP does.
