Poor GNSS accuracy? [Bug 8]

[Gadwin]

I marked the entrance and exit on the very right picture with two brown dots. The exit is on the very bottom of the picture and the entrance is a little above.

The cross section on the very right picture, which was marked with a yellow circle in the post before, is the place where I take and put back again my shopping cart. Even this was not accurate with the 67. But that is not really important in my view. More important is the behavior inside this building and when I got out.

I think it speaks for itself if a device can even handle very difficult settings, right? So I think the test is legit and shows clearly the better performance of the 66sr. Of course I would have to do more tests, but I never experienced such behavior with the 66sr.

Many people reported even back then, that they don't see any difference of the 66sr and 66s, when they have no obstacles and so on.

But that is not interesting and garmin explicit advertise, that their Multi-band devices can handle even in difficult environment. It is natural to expect, that a device which has also Multi-GNSS and Multi-Band and even one stallite system more on board, that it should at least perform the same like its previous model.

I think users who had 66i and go to 67i will really enjoy it. But about 66sr to 67 is really not so clear.

Spoiler

accuracy.png

[Nail]

But that is not interesting and garmin explicit advertise, that their Multi-band devices can handle even in difficult environment. It is natural to expect, that a device which has also Multi-GNSS and Multi-Band and even one stallite system more on board, that it should at least perform the same like its previous model.

In good conditions, the 67 is better than the 66sr.
See this test. The black circles are places where I was 100% following the same line.

Spoiler

67vs66sr8.jpg

[Gadwin]

But that is not interesting and garmin explicit advertise, that their Multi-band devices can handle even in difficult environment. It is natural to expect, that a device which has also Multi-GNSS and Multi-Band and even one stallite system more on board, that it should at least perform the same like its previous model.

In good conditions, the 67 is better than the 66sr.
See this test. The black circles are places where I was 100% following the same line.

Spoiler

67vs66sr8.jpg

This result is after the compass calibration, right? Before that you had a really bad track.

[Nail]

This result is after the compass calibration, right? Before that you had a really bad track.

I have not done such comparative tests before.
I've done tests similar to yours from the store before.
I noticed the error accidentally, turned on the GPS at home and after some time I took it outside and started recording. It turned out that the GPSMAP 67 did not catch the position after going outside. He does it much later. There is no such problem when you start the device outside or shortly before leaving the house.
Check yours, turn it on and leave it on the table. Watch if he is at your house or leaving? Record it, of course.

[Gadwin]

So I compared two other tracks, what I would call in a "normal" condition. I walk pass by two buildings, left is a building wth 8 floors at the right it is a building with 4 floors. I walked with the 66sr many times there and it was always on the path or nearby at the right side. I always walk at the right side of this path really close to the right edge.

The 67 was way too off compared to the 66sr, what I never experienced. I started the recordings outside.

I measured also the deviation of the point where I should be. The deviation to the 66sr was indeed around 1.8m and to the 67 it was 3.2m.

Since users reported here with the 67, that it could reach 1.6m accuracy, it should be indeed more accurate in super super good conditions. But when condition worsen a bit, it is way worse compared with the 66sr.

Spoiler

ac.png

[neilwoody]

I have been using my GPSMAP 67 outdoors with software 4.70/firmware 2.46 with GNSS/Multiband enabled. I have been getting 7 - 8 feet accuracy while walking. I got the same results testing under moderate conifer foliage. My GPSR went to 6 once when I was stopped. I use my GPS for bushwhacking off-trail so these results are acceptable for me. I have attached a Garmin Basecamp screenshot of my track which was alongside road. The track is displayed on a USGS 7.5 topographic map.

Spoiler

GPS67.png

[Nail]

I have been using my GPSMAP 67 outdoors with software 4.70/firmware 2.46 with GNSS/Multiband enabled. I have been getting 7 - 8 feet accuracy while walking. I got the same results testing under moderate conifer foliage. My GPSR went to 6 once when I was stopped. I use my GPS for bushwhacking off-trail so these results are acceptable for me. I have attached a Garmin Basecamp screenshot of my track which was alongside road. The track is displayed on a USGS 7.5 topographic map.

The magnification is not very large, so it is difficult to assess the quality of the trace. Without comparing with other GPS does not reflect the quality. Have you used a 66sr? If you switched from Montan or Oregon, you must be happy. Your feelings are subjective and good for you and will be good for most people. On the other hand, I want to change from 66sr to GPSMAP 67 and I assume that I will take a small step towards a better result. Garmin added a new team to the crew of other satellites, so I expect more accurate indications. Also got a new chipset for improvement. I'm still optimistic that there will also be a good result in the forest. For balance, below I show 66sr which is not ideal, both in the forest and on the street.

Spoiler

67vs66sr9.jpg

[Gadwin]

How much apart are the those lines? If it is within 1.8m, it is optimal. I guess even with the 67 under good conditions you can't expect an overlapping. More important is it to compare it with satellite images and look how accurate it was. Normal maps like topoactive are not accurate enough, that is the reason why I always showed my tracks on a satellite image.

[neilwoody]

On a recent walk, my GPSMAP 67 displayed an accuracy of 7 - 8 estimated accuracy. I have attached a screenshot of my track on Google Earth.

Spoiler

GPSMAP67test.jpg

[Nail]

On a recent walk, my GPSMAP 67 displayed an accuracy of 7 - 8 estimated accuracy. I have attached a screenshot of my track on Google Earth.

I take it you were on foot? The line floats quite a bit, if that's how you moved it's fine. All in all, without a comparative test with another GPS navigation or repeating the same route, we will not get an answer as to the accuracy.
I zoomed in on this area and it's not particularly strenuous for the GPS.

How much apart are the those lines? If it is within 1.8m, it is optimal. I guess even with the 67 under good conditions you can't expect an overlapping. More important is it to compare it with satellite images and look how accurate it was. Normal maps like topoactive are not accurate enough, that is the reason why I always showed my tracks on a satellite image.

Quite a distance between lines 5m.

I have such a reference path in the forest. As mappers in OSMap do, they collect GPS tracks and move the road where there are the most overlapping tracks. That's what I did with the path I walk most often. Soon I will take 67 there and show what it can do.

Spoiler

67vs66sr10.jpg

The paths from my tests I put on the OSMap map myself and they are set according to satellite maps. GPSMAP also helps me with mapping.

Spoiler

67vs66sr11.jpg