Ascent way different

[skinnie]

Hello all,
First of all thanks for this great community and site. Very helpful.
Yesterday decide to test for the first time my Gpsmap 65s on a hike in Andorra (firmware 4.0 patched, topo hispânica v7 pro , battery save mode activated and “clamped “ on my backpack. Record method auto, altimeter auto calibration on)
I also recorded the hike with my garmin 945 lte (firmware 4.07 beta, gps + glonass, altimeter calibration on start only).
This hike was imported from Komoot where it stated to have about 780m ascent https://www.komoot.com/tour/506929947?r ... mFQQfdVUyk

Importing the hike from Komoot to garmin explore gave me 1061m of ascent.
I did the hike (with slightly different start position (less than on km way and flat ) and even shortcut a bit.
945 lte gave me 1085m of ascent
https://connect.garmin.com/modern/activity/7735569264

And to my surprise the gpsmap 65s gave 1649m of ascent!
I had multi gnss and multi band activated.
https://mega.nz/file/gsFBBQSL#5wb46ZXXV ... wvvYP8iqcg

What can explain such big difference?
Any hint ?

[0x01]

I have very good experience with the 65s altimeter:
Known ascend 100m+ measured accurate. Absolute altitude deviation usually within +-2m (autocalibration on). And I have identified several marked altitudes which are present on hiking guideposts as wrong (explaining would explode this post).
I think the data from Komoot and Garmin Explore are from map and therefore you have the deviation.

[skinnie]

I have very good experience with the 65s altimeter:
Known ascend 100m+ measured accurate. Absolute altitude deviation usually within +-2m (autocalibration on). And I have identified several marked altitudes which are present on hiking guideposts as wrong (explaining would explode this post).
I think the data from Komoot and Garmin Explore are from map and therefore you have the deviation.

Thanks, makes sense.
On a hiking post it only deviated 2m from written altitude as you said.

Edit: just uploaded the gpx to connect and it gave 900m d+, and to Komoot 790 d+ , so apparently each platform “reads” it according to the map.
How can i know which d+ did the gpsmap 65s measured?

[mimichris]

We are fortunate in France to have free access to DEMs from all over France in 25m, 5m and 1m meshes (Lidar and radar), so the Twonav Land mapping software can display these DEMs in superimposition of a topo map and I can calculate for each point of a track made with a Garmin, the exact altitude of each point. Which gives me a D + with precision.

[GPSrChive]

skinnie,

I have looked at the information provided, the original course on Komoot webpage, your 945lte statistics on Garmin Connect webpage, and the GPX file downloaded from your GPSMAP 65.

I could provide better details if I had the GPX file from Komoot and the 945lte. Only having 'raw' data from one of the three devices makes a detailed comparison impossible.

However, I can say that if each device were to record their own elevation values for each track point along their respective recordings, the odds of any two (much less three) devices agreeing 100% of the time are very minute.

While most modern GPSr devices are relatively accurate with latitude and longitude, they are also not very accurate when recording elevation values. One device may be recording GPS elevation while the other is using atmospheric pressure to calculate elevation values, which will always result in some ambiguity when comparing their results.

Many file sharing services will use third party software to match the recording to a known DEM (Digital Elevation Model), which results in all track points having their respective elevation values rewritten. This can be necessary when the source data is clearly quite poor, or does not include elevation data to begin with.

The frequency of recorded track points in the original raw data also play a huge role in the overall statistical value of the final product. For example, if we have two recordings with the same start and finish points, one with 1000 track points between and the other only 500 track points between, the 'more detailed' recording will always provide greater statistical values than the other.

We do not know what device the Komoot recording was created with, or if the data has been 'filtered' or 'processed' before publishing (this is why the original raw data files are so important), so we only have the values they provide as 'estimates'.

We also do not know how many track points your 945lte recorded for the same journey again leaving us with only the 'processed' values displayed.

The GPX file from the GPSMAP 65s does include a total of 3300 track points for the journey, each with it's own atmospheric pressure derived elevation value attached.

I suspect if you were to have access to the 'raw' data from the original Komoot course and your 945lte, you would find they each have fewer saved track points for the same journey, resulting in the lower statistical values displayed.

Even if you could configure all devices to record track points at the exact same rate, and start and stop them all at the exact same moments, the resulting recordings would still have differences that would lead to 'discrepancies' when compared to each other. But one must remember, these are not 'real' or 'true' discrepancies, these are simply a 'fingerprint' for each unique recording. None of the recordings are incorrect, they are each simply unique to their host devices.

[skinnie]

skinnie,

I have looked at the information provided, the original course on Komoot webpage, your 945lte statistics on Garmin Connect webpage, and the GPX file downloaded from your GPSMAP 65.

I could provide better details if I had the GPX file from Komoot and the 945lte. Only having 'raw' data from one of the three devices makes a detailed comparison impossible.

However, I can say that if each device were to record their own elevation values for each track point along their respective recordings, the odds of any two (much less three) devices agreeing 100% of the time are very minute.

While most modern GPSr devices are relatively accurate with latitude and longitude, they are also not very accurate when recording elevation values. One device may be recording GPS elevation while the other is using atmospheric pressure to calculate elevation values, which will always result in some ambiguity when comparing their results.

Many file sharing services will use third party software to match the recording to a known DEM (Digital Elevation Model), which results in all track points having their respective elevation values rewritten. This can be necessary when the source data is clearly quite poor, or does not include elevation data to begin with.

The frequency of recorded track points in the original raw data also play a huge role in the overall statistical value of the final product. For example, if we have two recordings with the same start and finish points, one with 1000 track points between and the other only 500 track points between, the 'more detailed' recording will always provide greater statistical values than the other.

We do not know what device the Komoot recording was created with, or if the data has been 'filtered' or 'processed' before publishing (this is why the original raw data files are so important), so we only have the values they provide as 'estimates'.

We also do not know how many track points your 945lte recorded for the same journey again leaving us with only the 'processed' values displayed.

The GPX file from the GPSMAP 65s does include a total of 3300 track points for the journey, each with it's own atmospheric pressure derived elevation value attached.

I suspect if you were to have access to the 'raw' data from the original Komoot course and your 945lte, you would find they each have fewer saved track points for the same journey, resulting in the lower statistical values displayed.

Even if you could configure all devices to record track points at the exact same rate, and start and stop them all at the exact same moments, the resulting recordings would still have differences that would lead to 'discrepancies' when compared to each other. But one must remember, these are not 'real' or 'true' discrepancies, these are simply a 'fingerprint' for each unique recording. None of the recordings are incorrect, they are each simply unique to their host devices.

First thanks a lot to take your time to explain me so well.
When I imported to garmin explore, I was convinced that explore would "read" directly the values recorded from the gpsmap 65s
I was astonished by such difference between garmin connect 945 lte and garmin 65s garmin explore.
If the difference was like 100-200m I think it could be ok, but not almost the double.
After all explanations given I realized the "problem" was garmin express and not the device itself.

I was away with computer, but if anyone wants here are the raw files directly from the 945 lte and the GPSMAP 65s:

2021-10-29-ra 945lte.zip

Track_2021-10-29 raw gpsmap 65s.zip

Have a great week!

[GPSrChive]

Interestingly, while the track recording from the GPSMAP 65s was 3300 points in total, the 945lte track is over 24000 points total!

The 945lte was recording data once every second, while the GPSMAP 65s was set to record at a variable rate. This alone meant these two recordings were never going to agree.

Also, it appears the altimeter on the GPSMAP 65s was not properly calibrated, as the GPSMAP 65s GPX file indicates the starting point elevation was 5954 feet above sea level, while the 945lte FIT file indicates the same location was only 94 feet above sea level.

They didn't agree to begin with, and they were not configured to record at the same rate, so they were never going to match each other at the end of the recording.

[skinnie]

Interestingly, while the track recording from the GPSMAP 65s was 3300 points in total, the 945lte track is over 24000 points total!

The 945lte was recording data once every second, while the GPSMAP 65s was set to record at a variable rate. This alone meant these two recordings were never going to agree.

Also, it appears the altimeter on the GPSMAP 65s was not properly calibrated, as the GPSMAP 65s GPX file indicates the starting point elevation was 5954 feet above sea level, while the 945lte FIT file indicates the same location was only 94 feet above sea level.

They didn't agree to begin with, and they were not configured to record at the same rate, so they were never going to match each other at the end of the recording.

Thanks for your inputs.
Regarding to the gpsmap 65s I am new to it (although I had an etrex 30x).
Is there an option to increase the record rate on the gpsmap?
Regarding the altimeter of the 945 lte, at the begining it is wrong for sure, and I will report it to garmin, maybe due to the beta firmware.
I have it to "calibrate on the begining of activiy" so it should have been ok.

[GPSrChive]

The GPSMAP 65 series track log recording method and interval can be adjusted as desired in Setup > Tracks on the device.

Please see GPSrChive > GPSMAP 65 for additional information.

[GPSrChive]

The GPSMAP 65 series track log recording method and interval can be adjusted as desired in Setup > Tracks on the device.

Please see GPSrChive > GPSMAP 65 for additional information.

What is the correct elevation value for the starting point of the tracks you shared?