Total Ascent Values Exaggerated by 50% [Bug 29]

[art3k]

Todays hike, computer trip:

Ascent/Descent 558/651 m

Ascent/Descent 558/651 m

558.bmp

But on Save Activity something different:

Ascent/Descent 449/450 m

Ascent/Descent 449/450 m

588.bmp

Also, screen from offline track analyser:

Ascent/Descent 445/439 m

Ascent/Descent 445/439 m

track analyser.png

Well, it's pretty much the same ascent value as calculated on my fenix 6: 445 m


So, what goes up? Both garmins calculates correct, calculated from track is correct (continous recording 2sec interval), but there are something strange with Total Ascent/Descent values on Trip Computer datafields. Maybe it's not Total Ascent/Descent but something else? Mistake on display only?

[RickN55]

What does it say on your Garmin Connect? Mine are always at least 3x more than my Epix and 65s and what the Elevation Corrections say when you turn them on,

I'm think I'm going to sell on the 67, way to many bugs on it, first bad Garmin buy I've ever done tbf, had Garmins since the 60CSx era never had no issues until the 67!

[art3k]

a/d 449.0/450.0 on GPSMAP67 (min/max: 205,8/582,2)
a/d 444,7/445,5 on fenix (min/max: 204,4/578,8)

I'm sure it's an error in the data field values on Trip Computer, not in proper calculations.

Latest firmware on both (last beta on fenix)
Well, i doesn't really matter, but I performed calibrations knowing start point elevation value
Elevation value taken from elevation map, easy find on the internet. Autocalibration off (not once, not continous, never use it, especially in the mountains).
Statistics on Trip Computer were cleared just before activity start.

[RickN55]

Had another reply from Garmin, they've finally admitted there is an issue and they know what is causing it :

Thanks for bearing with me. Sorry for the delay here.

I have reviewed your experince with the total ascent concerns and found the cause here.  Please let me apologise for the amount of text here. 
 
What is happening here is there is a elevation overcalculation happening in the activity itself.

The readings of elevation are correct which is showing the barometer is working. What is breaking is the activity file the 67 is generating. Within that, it is messing with the end calculation causing the total ascent calculation to be doubled and in some cases I've checked over with other test files tripled or more. This is not a device level issue as I have replicated this. Apologies for the situation previous. The information provided at the time is still correct in the total ascent and GPS calculation methods across other units. But this device at present is having this model is having this unique scenario. 

What does that mean for now?

I have raised an internal investigation ticket and we have been investigating this experince with another user however, we are not seeing reports in mass which we would expect but as a whole others may not look into elevation comparatively or as often. 

But can I please ask that if you know anyone who uses a 67 and is experiencing the same experince here you have them contact us directly about it. We can then raise the investigations we need to get as many people reporting this experince as possible. 

That is the reply I have just had from Garmin, at last things are moving in the right direction! So get reporting guys!

[pedigree1]

I have been following this thread with interest. I had a battle with Garmin over a similar issue on my 66s which was never resolved, in effect I gave up and just learned to live with inaccurate ascent figures. Garmin suggested/concluded that me carrying my unit in my pocket rather than having it exposed to the atmospphere on e.g. a belt loop or rucksack lanyard was the cause.

Looking at data from my 67i I have been pleased with ascent/descent data from it; it is showing a relatively good match with the expected figures as far as I can see. E.g. latest hike: Garmin says (on the recorded activity screen on the unit) 564m ascent, digital mapping software had the route at 541m. So, my point is that not every unit is seeing this issue, which is surprising given that Garmin appear to now admit that there is a software fault. Oh, and I still carry my unit in my pocket.

Or am I missing something? Can someone briefly summarise this issue rather than having to re-read 15 pages of posts?

[RickN55]

I've had 3 67 units from Garmin now, and they've all done it.

[dhg]

Hi All --

I wonder if anyone has any further news on this from Garmin? My attempt to raise the issue with them just resulted in general denial and a desire to close the ticket as soon as possible. They kept on asking for irrelevent info like error messages, when my initial description of the problem was not one that involved error messages.

Regarding some of the other posts here, please note the issue I originally posted on was that the gpsmap 67 screen count of ascent and descent was exagerrated and not that the ascent and descent were mismatched.

The issue affects the displayed counts but strangely not the values that get recorded on the gpx file.

eg on a circular walk I find that the display ascent / descent count is about 50% more than it should be during activity. By the end of say a 1200 feet gain walk the device will show an ascent/descent of approx 1700 feet, but when I look at the recorded gpx after it will correctly reflect around 1200 feet.

Please note I'm now running the recently released 7.60 firmware and that hasn't fixed the issue.

Cheers.

Don.

[JungleJim]

I wonder if anyone has any further news on this from Garmin? My attempt to raise the issue with them just resulted in general denial and a desire to close the ticket as soon as possible. They kept on asking for irrelevent info like error messages, when my initial description of the problem was not one that involved error messages.

Regarding some of the other posts here, please note the issue I originally posted on was that the gpsmap 67 screen count of ascent and descent was exagerrated and not that the ascent and descent were mismatched.

The issue affects the displayed counts but strangely not the values that get recorded on the gpx file.

eg on a circular walk I find that the display ascent / descent count is about 50% more than it should be during activity. By the end of say a 1200 feet gain walk the device will show an ascent/descent of approx 1700 feet, but when I look at the recorded gpx after it will correctly reflect around 1200 feet.

Please note I'm now running the recently released 7.60 firmware and that hasn't fixed the issue.

I have been in contact with Garmin Support several times on this issue. Initially they weren't really helpful but now they seem to be a bit more willing to look into this. So I'm collecting some data for them to help investigating.

My observations so far (with device software version 7.60):

• it seems the total ascent / total descent fields do not apply any filtering or smoothing. So any change in elevation is counted towards ascent (if positive) or descent (if negative)
• the (absolute) elevation is quite accurate but not very stable. In other words it fluctuates quite a bit with a small amount. So in my area, which is quite flat and at sea level, it will often go from 0m to a couple of meters above or below even though there's no change in actual elevation. Even when the weather is stable.
• the auto calibration feature is weird. Auto Calibrate set to Once should calibrate to GPS elevation at start of activity recording, but this does not seem to happen. Auto Calibration set to Continuous sometimes result in the elevation steadily increasing or decreasing. For example one time it started at 0m and ended at -20m while the walk was mostly flat. But it also happened that the elevation kept increasing steadily.
• when manually calibrating to known elevation and then setting auto calibration seems to work better, as in elevation is quite accurate but it still fluctuates a lot resulting in too high ascent and descent. I did not test with Auto Calibration set to Off yet.
• Total Ascent and Total Descent also change in value when no activity recording is active. This is not consistent with other data fields that show data related to an activity (like Trip Time). I have set my device to only record trip data during an activity BTW.

So in a nutshell, absolute elevation readings are quite good (in my tests), but due to the fluctuation in elevation value the total ascent and descent are often (but not always) way too high.

[Przekątny]

[*] the auto calibration feature is weird. Auto Calibrate set to Once should calibrate to GPS elevation at start of activity recording, but this does not seem to happen. Auto Calibration set to Continuous sometimes result in the elevation steadily increasing or decreasing. For example one time it started at 0m and ended at -20m while the walk was mostly flat. But it also happened that the elevation kept increasing steadily.

I don't know how automatic calibration (Once) can affect constant increases or constant decreases in altitude. If we test two identical receivers, in which we calibrate the altimeter at 100 m above sea level and in the second one at 500 m above sea level, the ascents and descents in both receivers should be similar. The ascent and descent are influenced by changes in atmospheric pressure; the pressure changes regardless of whether the altimeter has been properly calibrated (automatic or manual).

Automatic calibration (Once) affects the correct altitude readings, the device should be turned on in an open space, wait for full satellite acquisition and then turn on activity recording. If the device shows an incorrect altitude, the altimeter will be incorrectly calibrated, but this will not affect the ascent and descent values.

[JungleJim]

I don't know how automatic calibration (Once) can affect constant increases or constant decreases in altitude. If we test two identical receivers, in which we calibrate the altimeter at 100 m above sea level and in the second one at 500 m above sea level, the ascents and descents in both receivers should be similar. The ascent and descent are influenced by changes in atmospheric pressure; the pressure changes regardless of whether the altimeter has been properly calibrated (automatic or manual).

The constant increase/decrease in the value of the Elevation data field happened with auto calibration set to Continuous, not when set to Once.

I don't know how Total Ascent/Descent and Elevation are determined exactly, but what I observe is that when the Elevation field changes value (for example from 0m to 1m), the Total Ascent and Total Descent fields are also updated (plus 1m in this example). So the fact that elevation determined by the device fluctuates frequently even though the actual elevation I'm at stays the same, makes that Total Ascent and Descent values are exaggerated.

Automatic calibration (Once) affects the correct altitude readings, the device should be turned on in an open space, wait for full satellite acquisition and then turn on activity recording. If the device shows an incorrect altitude, the altimeter will be incorrectly calibrated, but this will not affect the ascent and descent values.

In my tests I always make sure that satellite reception is good and accuracy has stabilized to 1.8m. I did tests with either calibrating to known elevation before starting activity recording, or not calibrating to known elevation. From what I've seen total ascent and descent indeed seem to be calculated only by changes in elevation, not absolute elevation values (which makes sense of course).