Instinct 3 amoled 50 fake elevation gain

Example of the issue?

Throughout the entire activity, there are constant small spikes up and down, which end up adding elevation gain even on flat terrain – which is unrealistic.

The graph is missing the scale on both axes, so we do not see how much the elevation changed, and how fast, but if there was a weather change (dropping atmospheric pressure, or a bigger change of temperature) towards the final part of the activity, then it would explain the increase of the elevation.

not the same climbs, but for example climb soft 8.09 climb soft 7.08 

It is impossible to compare the two graphs without seing the scale. Please use the "Compare" function in GC Web, and post a screenshot of the two profiles overlayed

Please post the graphs from the two activities (before and after the update) in the following form (select the two activities, and hit the Compare button). The example below shows two of my own activities (before and after 8.09) where I run 5 rounds of a hill climb. Not only the two profiles are practically identical, but also the 5 rounds in each activity are accurate and respecting the reality.

I get what you're asking for, but I don't have two identical activities recorded on different software versions. The only thing I’ll be able to provide in a few days is a comparison between two devices (Edge vs Instinct). The graph I shared was just to illustrate the issue.

The core of the problem is also in the actual elevation values measured by the devices compared to another device or even to maps with known elevation data. I'm not saying it's definitely caused by the update — but I only started noticing it right after the update. Maybe it’s a device issue, but if that’s the case, it started acting up pretty soon.

5746043 said:

I don't have two identical activities recorded on different software versions.

And what about the two you posted above?:
   

BTW, in that case, there are many "dents" in the ascent of the left activity, increasing the elevation gain - that's typical when the pressure sensor port is exposed to the air flow or to the wind, or when it is wet or dirty, because in such cases the wind/airflow/water/sweat/dirt can easily cause slight pressure changes altering the elevation. When the watch is protected with a sleeve, or the direction of the wind is different, the elevation profile is smoother (like on the right side).

Yeah, I’ve read about the various factors that can affect elevation readings, but even before, the watch wasn’t protected from wind or anything — and it measured just fine, even better than the old Edge 830. And now suddenly it’s way off, and we’re not talking small differences — it’s off by hundreds of meters.

Anyway, I’ll try that experiment next time and cover the watch with something to see if it makes a difference.

Just to give you an idea, I'm attaching a screenshot from the Strava app — one shows the elevation data as recorded by the device, and the other shows the corrected elevation data based on Strava's own adjustments.